A self-control/life-course theory of criminal behavior

Proposition 1: Self-control predicts problematic behavior at all relevant points of the life course

This first proposition is neither new nor controversial, but it is a good place to start. The phrase ‘relevant points of the life course’ is used because the present theory is not intended to explain the behavior – criminal or otherwise – of infants. Studies have, however, consistently shown that low self-control is linked to behavioral problems in young children (for example, physical aggression; see Meece and Robinson, 2014), to delinquency during the adolescent and teen years (Hay, 2001; Perrone et al., 2004), and to criminal behavior in adulthood (Pratt and Cullen, 2000; Reisig et al., 2012; Wright et al., 1999). Recent work has even shown that low self-control predicts offending among the elderly (Wolfe, 2014). Thus, just as little kids who lack self-control are more likely to scratch and bite their peers, they are also more likely to shoplift and steal pills when they reach retirement age. Put simply, neither crime nor low self-control should be viewed as strictly youthful problems (see also Reisig and Holtfreter, 2013). Instead, at all points in the life course, within any age cohort there will be variation in levels of self-control and such variation will significantly explain individuals’ patterns of misbehavior. ¹

Proposition 2: Self-control varies within individuals at all relevant points of the life course

The traditional view of self-control in criminology is that it is an individual personality trait that is established rather early on in life – primarily through effective parenting – and is highly stable over the life course (Gottfredson and Hirschi, 1990). ² And in terms of empirical support, the link between effective parenting and the development of self-control in children has been well established (Cecil et al., 2012; Hay, 2001; Pratt et al., 2004), albeit with some important qualifications (Burt et al., 2006; Wright and Beaver, 2005; Wright et al., 2008). Faring far less well, however, is the notion that self-control is stable within individuals – a proposition that has been called into question by several emerging lines of research.

For example, Hirschi’s (2004) own revision of self-control theory argues that it is the social bonds one imports into a given situation that will serve as the source of self-control. Of course, those bonds can change – for example, friendships with peers can be lost or gained over time – which would indicate that one’s level of self-control can change as well. In addition, studies that have disaggregated the desire to exercise self-control versus the ability to exercise self-control (Piquero et al., 2010; Tittle et al., 2004) also reveal that one’s preferences in the moment will influence whether their self-control will actually kick in when facing a given situation. ³ Perhaps most telling is the recent research on the ‘depletion’ of self-control under stressful conditions (Hagger et al., 2010; Uziel and Baumeister, 2012; see also Baumeister, 2002; Meldrum et al., 2015; Muravin et al., 2006). This line of work conceives of self-control as a limited resource – like a muscle – that can get tired out if used too often or too intensely. As such, some people start out ‘stronger’ in self-control than others (that is, individuals vary in terms of their levels of pre-existing self-control), some people are in ‘better shape’ than others (that is, individuals vary in terms of their resistance or susceptibility to self-control depletion), and, once depleted, it may take considerable time for one’s self-control reserves to be replenished.

All of these lines of research collectively hint at – to a greater or lesser degree – three primary components related to the present proposition. First, self-control can change within individuals at any given time and is subject to depletion at all points in the life course. Second, individuals vary with respect to how susceptible they are to self-control depletion throughout the life course. Of particular importance are the criminogenic social contexts and stressful life events that can drain one’s level of self-control at any given time (Pratt, 2009). And third, young people are more vulnerable to self-control depletion than older people are. Recent neuropsychological research on changes in executive functioning over time (for example, the ability to think of and anticipate the consequences of one’s actions) indicates that as we age we are less likely to lose our self-control in situations where we probably would have when we were younger (Kray et al., 2004). So, to put it simply, it is the changing nature of self-control during the aging process that may serve as the ‘missing link’ in the life-course theoretical tradition – a point that is elaborated upon in the next proposition.

Proposition 3: Following the peak crime years, self-control increases within individuals over time

Few empirical regularities in criminology have been as well documented as the ‘age–crime curve’ (Loeber and Farrington, 2014). That criminal behavior generally increases during adolescence, peaks at around age 17 or so, and declines as individuals age into adulthood has taken on the flavor of a criminological fact (Farrington, 1986; Piquero et al., 2003). And while research has revealed certain distinct trajectories that depart from the curve (Nagin, 2005), most individuals fit the well-known pattern (Farrington et al., 2013). But only slightly less well known is this: our inability to fully explain it. Indeed, Hirschi and Gottfredson (1983) argued long ago that the age–crime curve is essentially invariant and cannot be explained by any known psychological or sociological variables – a challenge presented to the field that was labeled three decades later as the ‘inexplicability’ thesis (Sweeten et al., 2013: 2).

Criminologists responded to this challenge in two ways. The first approach entailed a virtual explosion in longitudinal research. This development has enhanced the field considerably by producing new knowledge about how people’s lives change over time, as well as new theoretical perspectives intended to explain the causes and consequences of those very changes (Erosheva et al., 2014; see also Warr, 1998). And although this new criminological paradigm has come close to explaining the age–crime curve, it still has not fully delivered (Sweeten et al., 2013). Accordingly, the other approach to addressing the inexplicability thesis has been to just give up altogether (Gottfredson and Hirschi, 1990). For those favoring this tactic, the age–crime curve is simply chalked up to ‘the inexorable aging of the organism’ (Gottfredson and Hirschi, 1990: 141), or what has been dubbed the process of ‘maturational reform’ (Glueck and Glueck, 1968; Matza, 1964) – something that recent life-course work has, to a certain extent, agreed with (Sweeten et al., 2013).

The problem is that neither of these approaches takes into account the dynamic nature of self-control and how it changes within individuals over time. The denial of self-control’s dynamic properties is particularly problematic with respect to the maturational reform argument, where the causal mechanisms underlying what maturation is and why it may affect criminal behavior are unspecified theoretically and therefore go unmeasured empirically (see, for example, the discussion by Rocque, 2014). What we are left with, then, is the proposition that maturational reform causes crime to decrease with age, even though we are not really sure what it is and, even if we were sure, we have no idea whether it could be measured reliably.

Yet given what we know about within-individual changes in self-control, it is reasonable to conclude that social and biological factors (both of which would conceivably be implicated in the maturational reform explanation) contribute to changes in self-control over the life course (see, for example, the discussion by Rocque et al., 2015). Such changes, in turn, should mirror the age–crime curve. This is not to say that the exact nature and timing of these changes in self-control will be uniform for everyone. Within any given age cohort, for example, levels of self-control will still be influenced by variations in the social contexts and life experiences that either facilitate or challenge the development of self-control. ⁴ Even so, the broad pattern is that self-control – at least the form of it that is linked to criminal behavior – is lower during the crime-prone years, increases thereafter, and shadows the rate of offending over the life course. ⁵ This is a simple proposition that is backed by evidence from multiple sources (see, for example, Burt et al., 2014; Forrest and Hay, 2011; Roberts et al., 2008; Smith et al., 2014). It is time we imported this notion into both the self-control and life-course traditions. And viewed in this way, it is possible that changes in self-control (along with changes in its variability within individuals as people get older) may provide a better explanation of the age–crime curve than has thus far been offered up by criminologists.

Proposition 4: Self-control influences selection into negative life events

To scholars working within the self-control tradition, the idea that those with low self-control will be more likely to self-select into a wide array of negative life events is nothing new. Gottfredson and Hirschi (1990) explicitly noted that those who have a tough time controlling their impulses are likely to end up with a lengthy roster of life’s setbacks and tragedies. But the life-course perspective – which tends to focus rather heavily on the consequences of negative life events – tends to devote far less attention to their sources. Some have even accused certain versions of the life-course perspective of assuming that negative life events enter into individuals’ lives largely at random. Paternoster and Bushway (2009: 1148), for example, argued that Laub and Sampson’s (2003) hugely influential life-course theory relies so heavily on ‘structural events’ that any given person ‘is one who responds or simply reacts to events that he finds himself in, but having had no or little part in creating.’

And yet there is considerable evidence demonstrating that the occurrence of negative life events is often far from random and that self-control is an important component in explaining why sometimes bad things happen to people. Self-control has been linked, for example, to being violently victimized (Turanovic and Pratt, 2014; Turanovic et al., 2014), largely because those who lack self-control are more likely to be engaging in violent behavior themselves. Those with low self-control are also more likely to be targeted and ripped off by scam artists because, even if they think they might get burned by it, the temptation of easy money is just too strong to pass up (Holtfreter et al., 2008; Pratt et al., 2010). Those with low self-control are also more likely to get arrested and incarcerated (Longshore and Turner, 1998), to get divorced, and to lose their job (Evans et al., 1997). All of these negative life events have implications for offending as the life course proceeds (Gibson and Krohn, 2013). It is therefore necessary to recognize that individuals are not merely leaves on the wind, getting tossed around by forces beyond their control, but rather that people can inflict considerable damage upon themselves as a result of their own impulsive, short-sighted, selfish behavior.

Proposition 5: Self-control influences coping strategies following negative life events

Although it is clear that self-control influences individuals’ selection into negative life events, it can also shape the consequences of those events. For example, being violently victimized, losing a job, getting divorced, or losing a relationship are all stressful life events that can cause considerable emotional and psychological distress (Agnew, 1992). As a result of feeling badly, individuals will seek out ways to cope in order to feel better (Agnew, 2002).

These coping strategies can come in different forms. They can be prosocial, where people could seek out counseling, talk with friends or loved ones, or engage in some spirited physical exercise. These are all healthy things that people can do to help them deal with life’s adversities. But coping strategies can be maladaptive as well, such as engaging in alcohol/substance abuse and criminal offending (see Baron, 2004; Hay and Evans, 2006). And whether somebody chooses one strategy over another is heavily influenced by their level of self-control. Because those with low self-control tend to place a premium on immediate gratification, they will find maladaptive coping strategies to be much more attractive because they are perceived to have benefits that will come right away (Turanovic and Pratt, 2013). Counseling takes time and jogging takes effort, but punching someone you are mad at or getting drunk to try to forget about an ex-girlfriend can make people feel better immediately. Of course, adopting these maladaptive coping strategies can make a bad situation worse – they rarely represent effective long-term solutions to a problem. They can also carry enduring adverse consequences for one’s physical, emotional, and psychological health (Macmillan, 2001). Thus, understanding how the choice of coping strategy is influenced by self-control is critical for understanding the criminogenic consequences of negative life events.

Proposition 6: Self-control is a key factor linking neuropsychological deficits to early onset offending

Moffitt’s (1993) dual taxonomy has become one of the more popular ideas within life-course criminology. According to this perspective, the most common trajectory of offending over time is adolescence-limited offending, where through a largely ‘normal’ developmental process, most individuals’ offending rates mirror the traditional age–crime curve. A small group of individuals, however, fall into the category of life-course persistent offenders – those who offend at a high rate throughout their lives. These two offending trajectories are considered to be distinct, with one group characterized by change and the other by stability in criminal behavior over time.

The key factor that distinguishes life-course persistent offenders from everybody else, according to Moffitt (1993), is the presence of neuropsychological deficits. These deficits can emerge in children as a result of prenatal problems associated with exposure to toxins (Moffitt, 1993). This can happen either involuntarily, as a consequence of living in an area where exposure to pollution is likely, or it could come about voluntarily when pregnant moms to be smoke, abuse alcohol, or take other drugs and end up doing physical damage to the unborn child (see McGloin et al., 2006). Such deficits are assumed to create problems with executive functioning – that is, one’s ability to consider the long-term consequences of their actions. And an important point according to Moffitt (1993) is that those who have such deficits will begin their criminal careers at an early age. As such, several studies have confirmed empirically the link between neuropsychological deficits and the early onset of offending (see, for example, de Boer et al., 2012; Raine et al., 1997; Tibbetts and Piquero, 1999).

And yet the key causal mechanism linking neuropsychological deficits to the early onset of offending – problems with executive functioning – looks a lot like low self-control. For example, executive functioning entails one’s ability to control their impulses and to predict the consequences of their actions – properties that also characterize self-control. This similarity has not been lost on criminologists and self-control has even been described as executive functioning (Beaver et al., 2007; see also Wikström and Treiber, 2007), and recent research has demonstrated empirically that neuropsychological deficits are an important precursor to low self-control among adolescents (Jackson and Beaver, 2013). It is now time to explicitly incorporate self-control into the theoretical model linking neuropsychological deficits to life-course persistent offending. This will require the self-control tradition to be more open to the possibility of biosocial sources of self-control, and for the life-course perspective to be more willing to consider low self-control as an important part of the causal process that explains offending over time.

Proposition 7: Self-control influences late adolescent/teen sensitivity to the maturity gap

Although the life-course persistent offending group within Moffitt’s (1993) framework has garnered the lion’s share of criminologists’ attention, there is also something important to be learned when looking at the reasons underlying adolescence-limited offending as well. For Moffitt (1993), a certain amount of offending during the mid to late teen years is normal and is generally assumed to be a response to the ‘maturity gap.’ Put simply, younger kids want to do what the older kids are doing – like drinking, smoking, having sex, and breaking the law – as a way of asserting their own maturity. ⁶

But not all youths are equally influenced by the maturity gap (Galambos et al., 2003; Hay and Meldrum, 2015). Some teenagers try to make themselves come off as more mature by getting jobs, or by surrounding themselves with the trappings of adult sophistication, such as carrying classic novels around with them or dressing like a grown up. Such kids are likely to possess higher levels of self-control than many of their peers. Alternatively, it is likely that those with lower levels of self-control will find the deviant activities of those who are slightly older than them to be most appealing; destroying public property, for example, fulfills the need for immediate gratification in a way that doing math homework does not. What this means is that there should still be considerable heterogeneity in the offending trajectories of those within the adolescence-limited offender category (see, for example, Nagin and Land, 1993), and variations in levels of self-control should go a long way toward explaining that heterogeneity. In particular, those with more self-control will still follow the traditional pattern of the age–crime curve, but they will likely begin offending later, they will offend less often during their ‘peak years,’ and they will begin their decline in offending more quickly.

Proposition 8: Self-control influences individuals’ sensitivity to informal and formal social control at all points in the life course

Although it comes in many different forms, the control tradition in criminology begins with a rational decision maker (Cornish and Clarke, 2014). In this view, would-be offenders weigh the relative costs and benefits of the behavioral alternatives in front of them (even if they do so quickly and imperfectly) and make a choice that maximizes pleasure and minimizes pain (Paternoster, 1987). The problem, however, is that the benefits of criminal behavior are likely to be immediate (for example, money, excitement, peer reinforcement). The costs (for example, school suspension and parental sanctions for kids; arrest, job loss, and relationship problems for adults), should they ever be incurred, are off in some distant and uncertain future. Accordingly, those with low self-control are less likely to even consider the potential costs of their criminal behavior because thinking into the future is not their strong suit. Such individuals are therefore much less sensitive to various forms of social control.

And along those lines, there is a rather large body of criminological literature linking self-control to both informal and formal social control mechanisms. Specifically, research indicates that those with low self-control are less likely to feel shame and remorse should their criminal behavior disappoint loved ones (Antonaccio and Tittle, 2008; Tibbetts and Myers, 1999), they are less likely to think they are going to get caught if they break the law (Pratt et al., 2006), and if they do get caught they are more likely to think that they were just supremely unlucky that one time and that they will therefore not get caught in the future if they do it again (Pogarsky and Piquero, 2003). Recent research indicates that self-control influences other potential sources of social control as well, including perceptions of procedural justice, police legitimacy, and legal cynicism (Reisig et al., 2011; Wolfe, 2011) – factors that, in turn, have been linked to offending (Reisig et al., 2014; Tyler, 2006). In short, self-control plays a critical role in determining whether social control efforts – regardless of whatever formal or informal package they come in – will have their intended effects on criminal behavior.

Proposition 9: Self-control influences selection into social ties and turning points over the life course

The importance of social ties to criminological theory and research cannot be overstated. The bonds that people have to family, friends, and prosocial institutions are among the most studied in the criminological literature (Britt and Gottfredson, 2003; Lilly et al., 2011; Turanovic and Pratt, 2015). Within the life-course tradition, these social ties are assumed to keep our behavior in check by providing sources of social control (Sampson and Laub, 1993), by structuring who we hang out with (Warr, 1998), and by placing constraints on our daily activities as we move through different stages in the life course (Laub and Sampson, 2003).

And yet similar to proposition #4 above, criminologists have a long history of paying way more attention to the consequences of social ties than they do to where they come from. And the self-control perspective has something to say about that. For example, research indicates that self-control influences the bonds that young people form to school (Catalano et al., 2004) and that self-control is a strong predictor of selection into peer groups – both prosocial and deviant (McGloin and Shermer, 2009; Young, 2011). Among adults, self-control influences selection into jobs and marriage (Boutwell and Beaver, 2010; Evans et al., 1997) – factors that are assumed, in turn, to keep people from offending and to cause offenders to desist from a life of crime (Laub and Sampson, 2003).

It is also reasonable to argue that self-control will influence the selection into the kinds of ‘cognitive transformations’ that foster desistance from offending (see, for example, Giordano et al., 2002). Recent research suggests that such cognitive transformations might actually happen before people enter into the kinds of social ties (for example, marriage and work) that have previously been assumed to cause desistance from crime (Lyngstad and Skardhamar, 2013; Skardhamar and Savolainen, 2014). It would make sense that a certain amount of self-control would need to be invoked to prompt such transformations. After all, cleaning up your act takes effort, discipline, and an eye toward what one hopes will be a better future – all of which will need a good dose of self-control to make happen. Either way, the broad point here is that self-control plays an important part in how people select into and form social ties as they move through their lives.

Proposition 10: Self-control influences the quality of social ties over the life course

It is important to note that not only does self-control influence selection into social ties, but self-control also plays a large role in determining the quality of those ties. It is no secret that people vary considerably in terms of the quality of relationships they have with their spouses and with their occupations (Towler and Stuhlmacher, 2013), and evidence suggests that such variation in quality explains some of the variation in offending (Laub et al., 1998). Put simply, the quality of social ties matters. And what is more, the role that self-control plays in shaping the quality of social ties takes on added importance with age. To be sure, young people typically have social ties provided to them – usually by (and/or in the form of) their parents. As people get older, they become more responsible for cultivating, maintaining, and nurturing their own social ties. And yet those with low self-control will typically behave in ways that make maintaining those ties difficult if not impossible.

A prime example of this phenomenon can be found in Laub and Sampson’s (2003) discussion of the life-course persistent offenders in their follow-up study of the Glueck men. Laub and Sampson (2003) noted that one factor that seems to characterize this group of men is that the life-course persistent offenders were still heavy drinkers even up to age 70. The explanation given is that hard drinking has eroded the social ties that these men might have otherwise had in their lives. And since their affection for alcohol drove their wives away and made them unfit to hold a steady job, these men were not subject to these forms of social control and were therefore at liberty to keep offending even in late life.

Yet it is equally possible that those who continue to drink heavily in the later stages of the life course have major problems with self-control. If so, they have likely created a long and diverse list of questionable decisions over the course of their lives that have destroyed their social ties, which may have little or nothing to do with their drinking (Costello et al., 2014). This should not be a surprise, since people who lack self-control are often really unpleasant to be around. These are people who victimize others, who have a tough time recognizing when they are being insensitive and socially inappropriate, and who are often too self-centered to understand that quality relationships require the reciprocal investment of time, care, and energy. As a result, such people tend to make lousy friends, husbands, and wives, and prosocial others rarely want to have anything to do with them. What this all means is that it is plausible that the relationships surrounding the erosion of social ties in adulthood, drinking too much well into late life, and continued criminal behavior may not be causal, but rather all may be the consequence of the behaviors that people engage in as a result of their lack of self-control.