Skip to main content
SearchLogin or Signup

The Value of Understanding the Human Side of Precautionary Behavior for Criminology and Crime Control Policy: The 2021 Bruce Smith Sr. Award Address to the Academy of Criminal Justice Sciences

Forthcoming in Justice Evaluation Journal

Published onMay 26, 2021
The Value of Understanding the Human Side of Precautionary Behavior for Criminology and Crime Control Policy: The 2021 Bruce Smith Sr. Award Address to the Academy of Criminal Justice Sciences
key-enterThis Pub is a Supplement to

This address is an accessible introduction to topics presented in greater depth in Schreck and Berg (2021) and Berg and Schreck (forthcoming). I am deeply grateful to Joan Antunes, Mark Berg, Michael Gottfredson, John McCluskey, and Travis Pratt for their kindness in reading and commenting on earlier drafts.

Abstract: Criminology has traditionally emphasized the offender, relegating theory, research, and policy on target decision-making to a place of relative unimportance. I advocate a different view. In choice theory, offenders prefer quick and easy reward at little risk, acting with target vulnerability very much in mind. I use choice theory concepts and assumptions to account for target precautionary behavior. The principles of certainty, swiftness, and severity can be developed to explain not only when people will act to limit vulnerability, but also suggest the kinds of precautions that tend to be chosen reliably. Low self-control, a source of victimization risk, lessens feelings of certainty of victimization and fosters a preference for the easiest and most immediately gratifying methods of protection. In addition to stimulating theorizing and research to improve the usage and consistency of precautionary behavior, this theory also offers insights about controversial choices to acquire defensive weapons. 

Keywords: Victimization, Criminological Theory, Precautionary Behavior, Self-Control Theory, Crime Control Policy

I once joked on Twitter that, unlike for the many in our field who study policing, domestic violence, or race, I have a rather difficult task finding news articles that would get people excited about the work I do. Journalists do not become famous writing stories along the lines of “would-be mugger sees a group of potential targets leaving a bar...and chickens out because robbing a group is a stupid idea.” If our precautions against crime work, usually nothing happens that would rise to the level of news. A burglar tries a door handle, finds it locked, and, seeing everything else secured, walks away. Yet, what ought to hold our fascination is an awareness that each time a decision to safeguard something makes the offender go do something else, we have witnessed the defeat of a lifetime’s accumulation of whatever it is that we think causes crime, whether socialization, inequality, psychological maladjustment, or biological predisposition. Moreover, neither incarceration nor rehabilitation was necessary to effect this defeat. Nevertheless, just like with journalists, we criminologists tend to be unexcited about the crimes that might have happened but did not. Were it otherwise, the development of theory and research about the choices people make to protect themselves and their belongings would be among the clear priorities of the field, and reflected at this late date in an abundance of scientifically supported policies directly benefitting potential crime targets. Because the present state of criminology is clearly not otherwise, I must instead begin this address with a quote:

It is the peculiar and perpetual error of the human understanding to be more moved and excited by affirmatives than by negatives. The root of all superstition is that [we] observe when things hit but not when they miss; and commit to memory the one and forget and pass over the other.

Those are the words of Francis Bacon, whom we credit with having invented the scientific method 400 years ago.

Bacon is describing a logical fallacy that we today call “survivorship bias.” The term comes to us from a World War 2 study of battle-damaged aircraft returning from combat missions (Casselman, 2016). The results showed that the enemy was very effective hitting the wings, tail, and fuselage. The enemy, however, appeared unable to register hits on the engines or cockpit. The United States military concluded from this that the wings, tail, and fuselage had the greatest need for additional armor protection. Abraham Wald, a statistician at Columbia University assigned to review the damage data, reached a very different conclusion. Since the engines or cockpit sections were as exposed as the other parts, he reasoned that chance alone ought to have produced hits to them. That the expected damage was absent in the data could mean only one thing: any damage to these critical areas resulted in the plane being shot down and thus failing to return to base. The inspectors in fact were studying only the survivors, and the data collected with such meticulous care merely described the damage an aircraft could sustain and remain able to return to friendly territory. The dangerousness of survivorship bias becomes very evident when thinking about what the policy repercussions might have been. If Wald had not questioned the military’s conclusions, if he instead accepted the evidence at face value, warplanes built later would have had vital components that were just as unprotected as before. However, the newer aircraft would have carried the weight of unneeded armor, making them slower and worse performing against the enemy. Let that sink in. A data-informed effort to save lives would have wasted resources and killed people. Reviewing only the affirmative evidence while ignoring the failures, as Francis Bacon warned, would have had disastrous results.

In criminology, we focus on offenders and their successes, and not their failures. Which means that much of our knowledge about crime and its control embodies this bias—or, as Bacon might say, too much of what we claim to be true about crime is evidence-based superstition rather than science. We fixate on the hits and hitters, and pass over and forget all the times they missed. And, if we care little for the offender’s misses, it becomes an easy matter for us to pass over and forget as uninteresting the choices targets made to cause the offender to miss. Survivorship bias, ignoring failures, gives the impression that criminals inevitably succeed, just as ignoring the failed college dropouts suggests that anyone can become a billionaire like Bill Gates just by leaving school and believing in the power of their dreams. I think this is why ideas like contagion effects, copycat crimes, superpredators, and crime specialization are rampant—it is not difficult to come up with such stuff if we assume targets are unable to do anything to make offenders fail. We see this in our prominent theories of offending (Schreck & Berg, 2021; Berg & Schreck, Forthcoming). Too few of them describe someone who acts in response to the vulnerability of the intended target, or who could be deterred if targets are not sufficiently vulnerable. Only P-O Wikström’s situational action theory (Wikström P. , 2006; Wikström, Oberwittler, Treiber, & Hardie, 2012) and Gottfredson and Hirschi’s (1990, 2019) self-control theory are exceptions. The first of these theories has not received anything like proper attention on this side of the Atlantic Ocean. The other theory, as confirmed unscientifically by anonymous reviews of my work, appears to be widely hated. This is unfortunate, because we may be missing a lot that is important for the understanding of crime and its control. Some colleagues and I recently published a paper investigating violence within relational dyads, meaning we had data from both potential offenders and their actual and potential targets (Schreck, Berg, & Rogers, 2021). We found that the actor’s low self-control increased the likelihood of their initiating a physical attack—no surprise there—but the target’s low self-control independently influenced whether an attack happened. Clearly, an actor is deciding to attack (or not) to some degree because of the other person’s background attributes and decisions. These results mean, contrary to what most of our crime theories would imply, that potential targets have a degree of control of the situation. If targets are not fundamentally passive and powerless, it is no longer reasonable or useful for criminology to continue giving a disproportionate share of its attention to offenders and their successes, while quietly consigning theory and research on the explanation of target behavior—plausibly the biggest reason would-be offenders fail—to the intellectual backwaters.

This matter is not merely of academic interest, because survivorship bias bleeds from our theories into practice. Our policy recommendations, like our theories, center exclusively on addressing the offender and thus elevate the importance of the institutions designed to control them (Berg & Schreck, Forthcoming). The message from academic criminology to the policymaker and public is clear: only these institutions and related offender-management programs can protect us from crime. The benefits of this arrangement (if any) to the potential crime target are indirect. Put another way, much like the relationship between “trickle-down economics” and the poor, the policies we advocate leave the human crime target to fend for itself. Who among our leading public criminologists recommends finding constructive ways to assist target decision-making, or providing those most at risk of victimization with effective and reasonable options, so that offenders abandon their plains or fail? Because of the weakness of our theories, and because we do not appreciate survivorship bias and its effects, our discipline instead urges policymakers to put more armor and resources where they are not needed, with the result of making our society worse performing (something obvious from watching the news) and leaving us just as vulnerable.

My life’s goal as a criminologist has been to try to move us past this. I offer a theory about the decision to engage in precautionary behavior, a model that suggests a new direction for policy, one sensitive to the needs and natural tendencies of the targets. Where Robert Bursik (2009) once, in an address to the American Society of Criminology, criticized us for repeating ourselves, treating old theories and ideas as if they were new, I think we have an opportunity for something genuinely original. And if we want to devise new crime control policy that relies less on a problematic criminal justice system, it would seem that improving the ability and options of people to protect themselves has an awful lot to recommend it.

I start with the basic assumptions, to describe the structure guiding my thinking about offenders and targets. I then move on to how the principles of certainty, swiftness, and severity shape target behavior. Finally, I discuss the role of self-control in all this, and conclude with policy and research implications.

A Choice Theory of Precautionary Behavior

To begin, choice theories were intended to be general theories of human behavior, and were not focused narrowly on the explanation of crime (Gottfredson M. R., 2011). Their assumptions and principles thus apply to the actions of everyone, meaning offenders as well as their targets. We all make decisions in our own interest. We are, all of us, governed by pain and pleasure. One can cobble a definition of victimization from these assumptions. If victimization is something with distinctive and universal qualities, it means that everywhere people will respond to it in a way consistent with their natural self-interest. In a paper published with Mark Berg, we defined criminal victimization as all pain, inflicted by someone else, without consent or legitimate cause, and returning no benefit to the victim (Schreck & Berg, 2021). So recall the pleasure-pain principle. Given how all of us want to avoid pain if we gain nothing from it, and how we are all hedonistic, we can expect that people everywhere, if they realize victimization is a possibility, will want nothing to do with it and will try to manage their risk unless otherwise prevented from doing so. Choice theory has no problem with the idea that self-preservation is natural to us and advantageous. The desire to avoid pain is the source of precautionary behavior.

Except in criminology we treat criminal motivation as all-powerful, not self-preservation. One might say that there is no place for cowardice in our theories. Edwin Sutherland (1956) could not have been more explicit, saying that self-preservation is present in some cultures and not others. I believe that there is a lot of evidence that self-preservation out-competes criminal motivation everywhere we should care to look. The benefits of committing crime are pretty meager and short-lived for the offender. Take murder, for instance. The murderer benefits in getting rid of a bothersome but often momentary pest, one quickly replaced by other pests, and usually has little else to show for it (Gottfredson & Hirschi, 1990). The costs to the victim are unpleasant and much more permanent. The broader costs of victimization—the lost property, injuries, trauma, inconvenience—are likely longer lasting than any benefits the offender receives. Given this disparity in consequences, and that offenders no less than their targets want to avoid pain, we can expect that people will tend to put more effort into not getting themselves hurt than into committing crime. Not surprisingly, data shows that the prevalence of precautionary behavior of some sort—like locking doors—is widespread anywhere there are such things as doors, door locks, and appreciable levels of crime (Rountree & Land, 1996; Wilcox, Jordan, & Pritchard, 2007). By contrast, crime data shows that very few of us hit others or steal. To me, if criminal opportunity is everywhere, as Gottfredson and Hirschi (2003) say, this asymmetry between self-preservation and criminal motivation says that it does not take much to get someone thinking about committing a crime to go have a beer instead. It does not take the direct involvement of the police. Although I could not locate any relevant research, I do not believe I am going out on a limb by saying that police are almost never on the scene of a crime in progress. Even hot-spot policing produces limited and only brief deterrent effects (Kopor, 1995; Sherman, 1990). Such findings, combined with the evidence that faster response times do not noticeably affect the probability of an arrest (Braga, 2015), should make us seriously question the belief that refinements in policing practice will deliver reductions in crime worthwhile enough to offset the problems that policing tends to create. It seems more plausible that it is the target’s immediate ability to fight or flee, or else complicate or prolong the completion of the crime, which brings home to offenders the possibility that they will face sanctions. We see this in Richard Wright and Scott Decker’s (1996) book on burglars, for instance. Burglars size up whether prospective targets have made their task easy and risk free enough, before ever walking up to the house. Maurice Cusson (1993) found that during their crimes criminals are terrified, being anxiously mindful of what their target may do. The presence of the police can matter, but the offender’s thoughts and decision-making inevitably center on the target.

Switching from the offender’s perspective back to the target’s, self-preservation does not mean we (as targets) can do anything we want. Among victim-blamers, there is this idea that targets can take any safety-minded action, however drastic, and if the victim failed to act then they were not really victims, which is rubbish. Choice theories are restraint theories, which tell us why people do not make what appears superficially to be the optimal choice. We have to make choices constrained by what threats we see as most likely and by the options that are actually feasible, given the resources, time, or energy available. It is not realistic to expect an abused and trauma-bonded romantic partner to find it easy to leave that relationship, especially if they are financially dependent. Similarly, youth living in high-crime neighborhoods who are trying to avoid victimization cannot simply leave for a safer neighborhood. Circumstances matter.

At other times, however, precautionary behavior and thus vulnerability is a choice, but even so, a decision to neglect precautions is not an automatic sign of irresponsibility. My theory recognizes that people have to manage competing interests and obligations in order to survive. We have to work, eat, sleep, socialize, or relax. Even the act of leaving the house to go to work, so we can keep paying for that house, reconfigures what is vulnerable and what is not. While we can agree that avoiding victimization is always to our advantage, we can expect there are going to be many times where the supposed optimal precautionary choice creates unacceptable amounts of pain, danger, inconvenience, or sacrifice to the person who has to implement it. Choice theory thus predicts that people compromise. We simply have to, in order to live. Ideally, what good crime control policy has to do is align proven effective precautionary action in specific risky situations with our basic tendencies and needs. This is why there is scientific value to understanding the human side of precautionary behavior.

So what do we have to do to get people to use precautions when they should? Travis Hirschi (1986) once wrote that, as far as he was concerned, the only important variables in criminology were certainty, swiftness and severity. Since these variables were developed under a general theory of human behavior, it is perfectly appropriate to use them to theorize the circumstances where people protect themselves against crime.


The first principle, certainty, is that the more certain we are of the possibility of our own victimization, the stronger our desire to take some sort of precautionary action. Offenders, if not criminologists, are aware of this concept as it pertains to their victims. We know that they try to manipulate the target’s sense of certainty by concealing themselves, or at least their intentions, until they feel that risk has reached an acceptable level. This means that if people feel safe, where the certainty of victimization is zero, it is going to be difficult to persuade them to go through the hassle and expense of buying an alarm, adding locks, or updating computer security. If the certainty of victimization is excessive, people will squander resources and, so to speak, put armor where it is not needed.

A salesperson once urged me to buy his company’s American-made smoke alarms. When he quoted the $5,000 cost to equip the house, he said, “How much is the life of your family worth to you?” Yet I rated the risk of a house fire as unlikely, and he could not explain why it would be a worse decision to use my limited resources instead to buy a car that was safer. (Our house, almost 20 years later, has not yet burned down; our car, however, has been hit multiple times.) Not everyone has a clear or proportionate view of risk, however. From a policy standpoint, we need people to feel a certainty of victimization any time the situation calls for it. The work of Chris Melde and his colleagues (2016) shows that teenagers who felt more certain crime would happen to them reported less victimization. This is indirect evidence, but the results suggest these teens did something in response to their fear to manage their risk successfully.

The question turns to what influences perceptions of certainty of victimization. Researchers have found evidence that situational characteristics matter for shaping certainty. Imagine, for instance, a person sitting on a park bench reading a book in broad daylight. It is unlikely that we will run for our lives. On the other hand, we probably will run in a situation where we see someone appearing suddenly out of the fog, approaching with a knife. The fear of crime literature—Kenneth Ferraro’s (1995) classic work on incivilities, Pamela Wilcox’s, David May’s, just to name a few—is full of evidence that environmental cues (like poor lighting, people loitering) tend to make people see victimization as a possibility.

At the same time, it seems to me that a person’s fear is partly a characteristic of their backgrounds. Some, when I described that person sitting on the bench reading a book, may think, “hmmm, I might be suspicious.” Since there was no objective reason for it, that suspicion is a reflection of person rather than the situation. Prior experiences with victimization, for example, make people more likely to be afraid of crime and engage in precautionary behavior. Machi Tseloni and Christina Zarafonitou (2008), using data from Greece, reported that direct and indirect experiences with victimization resulted in greater levels of fear. The point is that certainty is something that varies in intensity over time, by crime type, and by immediate circumstances. Some element of certainty may be a characteristic of the person, like prior experiences; however, this may not be as important as the cues noticed in the situation.


The second principle, of swiftness, refers to the ease and simplicity of the precautionary behavior. Everything else being equal, the easiest choice of all is to do nothing. We are more likely to use or tolerate simple precautions in preference to complex or difficult precautions. We prefer convenience to inconvenience. Which means that every extra bit of work, every minute lost getting things up and running, every dollar in cost, and each increase in the possibility of injuring oneself or others (and having to take appropriate safeguards), deters anyone from using the precaution consistently. Those of us from an older generation, before automatic car door locks and remote fobs, may remember that locking up the car could take up to four painful steps, eight if we had to roll up windows. As those features became standard, more sophisticated, and executable with a single push of a button, the Uniform Crime Report shows that car theft has fallen dramatically.

Many people perceive weapons as an obvious choice as part of their repertoire of precautions against crime, but consider the role of swiftness as a deterrent to acquiring them. For a weapon to be effective at protection, it has to be immediately available and prepared for use—i.e., easy and simple. A weapon cannot be in a secured and unloaded condition at that awkward moment when someone is coming through the front door with an axe; however, keeping a weapon readied for immediate use is inherently dangerous to the owner and anyone else nearby. Weapons thus have significant hidden costs that diminish their utility, at least for owners who intend to be responsible. In one of my papers, we reported descriptive data for the prevalence of alarms, leaving lights on, installing extra locks, getting a dog, getting neighbors to watch one’s home, and weapons (Schreck, Berg, Fisher, & Wilcox, 2018). There are substantial differences. Most people reported using precautions that cost nothing—leaving lights on and enlisting neighbors to watch the house. Only one in five obtained a defensive weapon, with slightly more buying an alarm or getting a dog—each of those has in common the necessity of imposing on their user some combination of time, money, and risk.

Swiftness is a characteristic of people and structural conditions as well. The theory acknowledges that we have to anticipate how disability or incapacity imposes limits on what is realistic for certain people, like with abuse victimization of young children or the elderly. The poor and disadvantaged likely have fewer options, though we know from Wes Skogan and Michael Maxfield’s (1981) work that they certainly do whatever they can to avoid being targeted. The state manipulates swiftness, deliberately increasing the difficulty and risk of taking precautions by those who are in legal custody. Which has the sideline of making it impossible for them to avoid brutal and inhumane treatment if they fall into the hands of authorities who abuse their power.


The third principle, severity, works in a couple of ways. First, people seem more likely to take precautions for those possessions and people where they might feel the loss most severely. I published a study a long while back in the Journal of Interpersonal Violence showing that adolescents who had parents who cared about them experienced less violent victimization (Schreck & Fisher, 2004). The idea of severity means we will sacrifice to guard the things that we treasure, irrespective of objective levels of risk.

But another way of looking at severity is in terms of how precautionary behavior is more likely to be used if it is gratifying (which is the complement of severity, in Cesare Beccaria’s model). People are more likely to be enthusiastic and consistent about precautions that are immediately rewarding. Perloff’s (1983) classic work shows that safety-minded behavior reduces anxiety, so precautionary action is pleasurable in the sense of alleviating distress brought on from a feeling of vulnerability. Many other actions that are precautionary, or that incidentally reduce vulnerability, are inherently gratifying. Robbers, for example, dislike targeting groups. Work that I have published (and Marcus Felson and Finn Esbensen, among others) found that people who have strong bonds of affection to their friends and parents tend to be less likely to experience victimization (Schreck & Fisher, 2004; Schreck, Fisher, & Miller, 2004; Schreck, Wright, & Miller, 2002). Which makes sense—if we care about someone then it is enjoyable just being in company with them. A group is the spontaneous and happy result. The fact that it makes the offender’s job more complicated and riskier is still a real fringe benefit, even if not necessarily planned.

Defensive weapons and firearms, too, may be attractive to some because they are inherently gratifying. We know that even firearms intended to serve for personal defense can be used for pleasure and recreation. One cannot say quite the same for house alarms, unless a neighborhood has contests for who has the loudest alarm. Guns also allow feelings of community with other gun owners. I am unaware that there are large communities of padlock owners sharing their insights and experiences. Guns confer feelings of power, whereas leaving a light on simply consumes power. But all of this shows that precautions that make us feel good in various ways—whether by inherent gratification or fun, by giving access to new friends, and in terms of enhanced status and taste, just to name a few—any of these make it more likely that someone will prefer the precaution.

Tying everything together, the ideas of certainty, swiftness, and severity appear to interact in complex ways and depend on not just immediate conditions but also the mind of the observer. Many people, for instance, appear to buy firearms not because of a rational assessment of their own risk but because possession makes them feel good. Whatever gratification others might feel is extinguished by the fact that firearms are expensive, difficult to maintain or store safely, and dangerous. If guns are gratifying, they are not always easy to own or acquire. Yet, my work shows that experience with prior victimization makes one more likely to choose a defensive weapon over doing something else (Schreck, Berg, Fisher, & Wilcox, 2018). So, people who perceive a high enough certainty of their own victimization ought to have an easier time overlooking danger and inconvenience. There is so much here that research can untangle.

The Role of Self-Control

Probably the defining paper of my career was the one I wrote linking low self-control to criminal victimization (Schreck C. J., 1999), so it is natural that I would speak of it here. Self-control is a good (or at least thorough) decision process, and applies not only to choices to commit crime but also decisions that might shape one’s vulnerability to crime. Self-control is the habitual tendency to take into account foreseeable long-term risk when making decisions. The risk of victimization certainly fits here. Even with no burglar or thief in sight, we can all foresee that if we leave a door unlocked then our house or car is vulnerable. Many people probably feel anxious at just the thought of it. For those with low self-control, however, it is easier to see going back as too much hassle and that it is much simpler to do nothing and hope for the best. Those of us who have ever forgotten to lock up the car, and realized it after getting to the office door—and then hesitated about going back—will understand how the temptation of doing nothing rather than checking to be sure. In this way, low self-control produces choices that give offenders larger windows of opportunity to act safely.

Having low self-control does not make victimization become inevitable, nor does it mean that victimization is deserved. Rather, it means that there are certain people who need extra help. In other words, self-control theory allows us to anticipate and understand noncompliance with precautions, which policy can then address. For one thing, the idea of self-control means that any policies that require individuals to take precautions can expect that those with low self-control will have clear preferences and tendencies. We should expect that the effects of certainty, swiftness, and severity (which affect everyone) would be exaggerated. Someone with low self-control probably would be less likely to react to ambiguous environmental cues or warnings that victimization might be imminent, at least if they are occupied with something more immediately enjoyable (like drinking). Everything else being equal, any precautionary behavior requiring noticeable effort should appear harder, more tedious, and more pointless. This suggests that they will defer any annoying defensive precautions as far into the future as possible; however, putting off the choice so late means they overestimate their ability to manage the offender and thus discount the offender’s opportunism. We can also predict someone with low self-control will prefer gratifying defensive choices irrespective of whether these are appropriate or dangerous. Recall how guns have all kinds of gratifying benefits not usually associated with other protective actions. A few years ago, I collaborated on a paper examining the mechanisms behind the decision to procure a defensive weapon versus something else, like getting better locks or an alarm (Schreck, Berg, Fisher, & Wilcox, 2018). We found that people who obtained weapons were somewhat less likely to try any other precautionary behavior, even something as effortless as leaving on the lights. One of the things that struck me in our results was how variables known to predict crime—being younger, male, less educated, prior criminality (of course), social isolation, beliefs in an honor code—all made a person more likely to prefer a weapon. After all, if someone is incapable of taking long-term risk seriously, what could possibly go wrong with leaving an unsecured and loaded gun lying around? What this means is that policy hoping to influence target decision-making faces particular challenges, at least among those who tend to have high risk of victimization due to low self-control.


This theory is based on an assessment of the evidence we have so far. There are very few data sets measuring precautionary behavior. There are even fewer that include a detailed range of measures designed to explain precautionary behavior, so we can better understand when people resort to it. My hope is that this address will inspire us to see the value of collecting this data. For those of my colleagues looking to stake a claim in an original area of research, understand that the theory I have described by no means represents the final version. It is an initial best guess, and something to be built and expanded upon (within limits; see Schreck & Berg, 2021). Nevertheless, I can give a sampling of research opportunities this theory proposes:

  1. Qualitative researchers should consider interviewing offenders and their known victims and intended victims, with the goal of exploring an all-around perspective on why offenders quit their plans or simply fail. Why did targets choose some other goal over their own safety? How far are offenders willing to go to manipulate the target into being vulnerable? At what point do targets recognize their danger and act? There is so much that would be fascinating to learn here.

  2. This theory suggests that people, irrespective of their identification with any demographic group, will try to protect themselves and their belongings in whatever way they think best, if they perceive victimization as a possibility. That is, the general principles that shape precautionary action—certainty, swiftness, severity, and self-control—should be applicable to anyone. The theory, however, does not suggest that all groups have easy access to the same options for protection, or necessarily take the same view of the efficacy of the choices available. Groups may differ considerably in their comfort with some forms of protection, favoring some while disliking others intensely—for instance, cultural differences in the acceptability of defensive weapons (Schreck et al., 2018; Warner & Ratcliff, 2021). Similarly, children do not have the independence to move to another neighborhood in order to protect themselves; adults typically make that decision. This suggests an unusual path forward for life course research. Researchers interested in documenting the effects of disadvantage, sexism, and racism might have much work to do here also. For instance, Jamie Fader’s (2021) in-depth interviews exploring the avoidance behavior of adult men in Philadelphia neighborhoods disadvantaged by mass incarceration revealed that avoidance carries a steep price: a catastrophic reduction in involvement in their own communities. I do not advocate that anyone stop taking precautions; I do say that managing risk should not disadvantage any group and that this theory can give structure to the issue and serve as a means to evaluate prospective solutions.

  3. This model suggests that research on offender decision-making should explore the connection between target protective behavior and how these magnify perceptions of formal and informal deterrence. The target’s actions may be what make the deterrent value of the police and criminal laws effective in the first place.

  4. If the target’s precautionary behavior is more effective at deterring offenders than coercive policing or sentencing, as the theory suggests, policymakers have less reason to endorse them and all of the consequences they often entail, such as: the increasing of the tax burden on the public, the weakening of civil liberties, and fomenting of social unrest. The police and criminal justice system could then limit its focus to administering justice and managing offenders, with no pretense that refinements in policing and the criminal law would have a significant direct effect on offender decision-making. The police instead would support the effectiveness of target precautionary actions through the maintenance of a presence within the community, the provision of accurate and accessible information about local crime risk, and prompt responsiveness to criminal complaints. An emphasis on precautionary behavior does not diminish community outreach and intervention. These efforts would be expanded to include a focus on supporting potential targets—identifying effective precautions for local crime problems, and developing strategies to encourage their consistent and appropriate use.

  5. Treatments or interventions designed to help build self-control should simultaneously make people safer from crime, at least to the degree that crime and vulnerability arise from the same flawed decision-making process characteristic of those with low self-control. Evaluators examining treatment programs that improve decision-making competencies (or build self-control) might thus consider including measures of victimization. If these programs are effective, as evidence indicates (Friese, Frankenbach, Job, & Loschelder, 2017; Pandey, et al., 2018), they should result in reductions in victimization experiences as well as offending (Berg & Schreck, Forthcoming).


I began with the goal of fostering an appreciation of what survivorship bias is, and what it means for the work we do. Our tendency to fixate on the criminal’s successes and ignore the failures ascribes to offenders borderline superhuman qualities, and is readily observable in the research we do and the policies we advocate—we overwhelmingly focus on the offender and urge the expansion of the power of the state. So little of what we produce improves the ability of targets to defend themselves or offers much insight about the difficulties that people face doing so. Luckily, the public has more sense than we criminologists give them credit for, as is evident from how precautionary behavior of some sort (like locking doors) is widespread. Even criminologists do it. This does not mean that people are necessarily consistent with this or that they have good or reasonable options, or that what they do is appropriate to the situation. This is where criminology can help. The model I propose here represents what I hope is a starting point for developing the human side of precautionary behavior.


Berg, M. T., & Schreck, C. J. (Forthcoming). The meaning of the victim-offender overlap for criminological theory and crime prevention policy. Annual Review of Criminology.

Braga, A. (2015). Crime and policing revisited. In New Perspectives on Policing. Washington, DC: National Institute of Justice.

Bursik, R. (2009). The Dead Sea Scrolls and criminological knowledge: 2008 presidential address to the American Society of Criminology. Criminology, 47, 5-16.

Casselman, B. (2016, June). The legend of Abraham Wald. Retrieved from

Cusson, M. (1993). Situational deterrence: Fear during the criminal event. Crime Prevention Studies, 1, 55-68.

Fader, J. J. (2021). "I don't have time for drama": Managing risk and uncertainty through network avoidance. Criminology. Retrieved from

Ferraro, K. (1995). Fear of Crime: Interpreting Victimization Risk. Albany, NY: State University of New York Press.

Friese, M., Frankenbach, J., Job, V., & Loschelder, D. D. (2017). Does self-control training improve self-control? A meta-analysis. Perspectives on Psychological Science, 12, 1077-1099.

Gottfredson, M. R. (2011). Sanctions, situations, and agency in control theories of crime. European Journal of Criminology, 8, 128-143.

Gottfredson, M. R., & Hirschi, T. (1990). A General Theory of Crime. Palo Alto, CA: Stanford University.

Gottfredson, M., & Hirschi, T. (2003). Self-control and opportunity. In C. Britt, & M. Gottfredson, Advances in Criminological Theory (pp. 5-19). New Brunswick, NJ: Transaction.

Gottfredson, M., & Hirschi, T. (2019). Modern Control Theory and the Limits of Criminal Justice. New York: Oxford.

Hirschi, T. (1986). On the compatibility of rational choice and social control theories of crime. In D. B. Cornish, & R. V. Clarke, The Reasoning Criminal: Rational Choice Perspectives on Offending (pp. 105-118). New Brunswick, NJ: Transaction.

Kopor, C. S. (1995). Just enough police presence: Reducing crime and disorderly behavior by optimizing patrol time in crime hot spots. Justice Quarterly, 12, 649-672.

Melde, C., Berg, M. T., & Esbensen, F. A. (2016). Fear, social interactions, and violence mitigation. Justice Quarterly, 33, 481-509.

Pandey, A., Hale, D., Das, S., Goddings, A. L., Blakemore, S. J., & Viner, R. M. (2018). Effectiveness of universal self-regulation-based interventions in children and adolescents: A systematic review and meta-analysis. JAMA Pediatrics, 172, 566-575.

Perloff, L. S. (1983). Perceptions of vulnerability to victimization. Journal of Social Issues, 39, 41-61.

Rountree, P., & Land, K. C. (1996). Burglary victimization, perceptions of crime risk, and routine activities: A multilevel analysis across Seattle neighborhoods and census tracts. Journal of Research in Crime and Delinquency, 33, 147-180.

Schreck, C. J. (1999). Criminal victimization and low self-control: An extension and test of a general theory of crime. Justice Quarterly, 16, 633-654.

Schreck, C. J., & Berg, M. T. (2021). What ideas of victimization and vulnerability mean for criminological theory: A logical appraisal. In J. Turanovic, & T. C. Pratt, Revitalizing Victimization Theory (pp. 15-55). New York: Routledge. Retrieved from

Schreck, C. J., & Fisher, B. (2004). Specifying the influence of family and peers on violent victimization: Extending routine activity and lifestyles theories. Journal of Interpersonal Violence, 19, 1021-1041.

Schreck, C. J., Berg, M. T., & Rogers, E. M. (2021). Predicting the probability of violence in actor-target dyads: Self-control and interpersonal provocations as mutual properties. British Journal of Criminology. Retrieved from

Schreck, C. J., Berg, M. T., Fisher, B. S., & Wilcox, P. (2018). That door you just kicked in was locked for your protection, not mine: Developing and testing competing theoretical models of crime prevention behavior. Journal of Research in Crime and Delinquency, 55, 316-345.

Schreck, C. J., Fisher, B., & Miller, J. (2004). The social context of violent victimization: A study of the delinquent peer effect. Justice Quarterly, 21, 23-47.

Schreck, C. J., Wright, R. A., & Miller, J. M. (2002). A study of individual and situational antecedents of violent victimization. Justice Quarterly, 19, 159-180.

Sherman, L. W. (1990). Police crackdowns: Initial and residual deterrence. In M. Tonry, Crime and Justice: A Review of Research (pp. 1-48). Chicago: University of Chicago.

Skogan, W. G., & Maxfield, M. G. (1981). Coping with Crime: Individual and Neighborhood Reactions. Thousand Oaks, CA: Sage.Sutherland, E. H. (1956). In A. K. Cohen, A. R. Lindesmith, & K. Schuessler, The Sutherland Papers. Bloomington, IN: Indiana University.

Tseloni, A., & Zarafonitou, C. (2008). Fear of crime and victimization. European Journal of Criminology, 5, 387-409.

Warner, T. D., & Ratcliff, S. (2021). What guns mean: Who sees guns as important, essential, and empowering (and why)? Sociological Inquiry, 91, 313-346.

Wikström, P. (2006). Individuals, settings, and acts of crime: Situational mechanisms and the explanation of crime. In P. Wikström, & R. Sampson, The Explanation of Crime: Context, Mechanism, and Development (pp. 61-107). Cambridge: Cambridge University Press.

Wikström, P. O., Oberwittler, D., Treiber, K., & Hardie, B. (2012). Breaking Rules: The Social and Situational Dynamics of Young People's Urban Crime. London: Oxford University.

Wilcox, P., Jordan, C. E., & Pritchard, A. J. (2007). A multidimensional examination of campus safety: Victimization, perceptions of danger, worry about crime, and precautionary behavior among college women in the post-Clery era. Crime & Delinquency, 53, 219-254.

Wright, R. T., & Decker, S. H. (1996). Burglars on the Job: Streetlife and Residential Break-Ins. Boston: Northeastern University.

Author Bio

Christopher J. Schreck is professor and chair in the Department of Criminal Justice at the Rochester Institute of Technology. His most recent work on victimization and precautionary behavior appears in Advances in Criminological Theory (edited by Jillian Turanovic and Travis Pratt), the British Journal of Criminology, and the Annual Review of Criminology.

1 of 2
A Supplement to this Pub

No comments here