Money Now,Money Later: Linking Time Discounting and Criminal Convictions in the Cambridge Study in Delinquent Development

Abstract

Two prominent criminological theories offer time discounting, or the preference for an immediate reward over a later one, as a central part of understanding involvement in criminal activity. Yet, there exist only a few studies investigating this issue, and they are limited in a few respects. The current study extends prior work in this area by using multiple measures of time discounting collected at three different periods of the life course to examine the link to criminal offending into late middle adulthood in the Cambridge Study in Delinquent Development. Results show that greater time discounting is positively related to a higher number of criminal convictions by late middle adulthood, and this effect remains after controlling for early life-course individual and environmental risk in a multivariate framework. Study limitations and implications are also discussed.

Keywords

time discounting criminal convictions individual differences self-control risk factors

Introduction

Many readers are acutely aware of the fact that one key correlate of delinquency and criminal behavior is the extent to which offenders fail to consider the long-term consequences of their actions. Two criminological theories offer this perspective as the key cog in their frameworks: Gottfredson and Hirschi’s (1990) general theory of crime and Wilson and Herrnstein’s (1985) crime and human nature theory (for other works that discuss time discounting/delaying gratification, see Baumeister, Bratslavsky, Muraven, & Tice, 1998; Baumeister, Heatherton, & Tice, 1994; Baumeister, Schmeichel, & Vohs, 2007; Baumeister, Vohs, & Tice, 2007; Duckworth & Carlson, 2013; Duckworth, Gollwitzer, Kirby, & Oettingen, 2013; Duckworth, Tsukayama, & Kirby, 2013; Duckworth, Tsukayama, & May, 2010; Mischel, Ebbesen, & Zeiss, 1972; Mischel, Shoda, & Peake, 1988; Mischel, Shoda, & Rodriguez, 1989; Nagin & Paternoster, 1993). Although we do not offer an explicit test of these two theories per se, we rely on them for theoretical background.

For Gottfredson and Hirschi, the image of the offender is one characterized by low self-control. Persons having lower self-control are said to be impulsive, prone to anger, interested in simple tasks, have a preference for physical over mental activities, are insensitive to the feelings of others, and prefer risky activities. As a result, such persons are also more likely to have a present orientation and are very likely to discount the future and prefer whatever is easiest and imminent. More specifically, Gottfredson and Hirschi (1990) argue that a lack of self-control is not a characteristic that is particularly “conducive to hard work, delayed gratification, or persistence in a course of action” (p. 115). In this same vein, they also contend that “crimes provide few or meager long-term benefits. They are not equivalent to a job or career. On the contrary, crimes interfere with long-term commitments to jobs, marriages, family, or friends” (p. 89; emphasis in original). Therefore, “people who discount the future [such as those with low self-control and who are more likely to commit crime] do not exert themselves for uncertain future benefits” (Gottfredson & Hirschi, 1990, p. 114).¹

Also focusing on the relationship between impulsiveness and offending, Wilson and Herrnstein (1985) argue that persons differ in their underlying criminal tendencies, and that whether a person chooses to commit a crime in any situation depends on whether the perceived benefits of offending are considered to outweigh the perceived costs. In fact, Wilson and Herrnstein have also specifically argued that delinquents discount future consequences more so than non-delinquents do, and this time discounting also applies to rewards as well as how the effort/reward ratio is taken into consideration. This is a clear point of departure from Gottfredson and Hirschi’s (1990) view toward time discounting as Wilson and Herrnstein focus more narrowly on the role of time discounting specifically as it relates to offending, whereas Gottfredson and Hirschi consider time discounting and its relation to crime within the theoretical framework of the general construct of low self-control and its relation to a variety of analogous behaviors. The benefits of offending, including material gain, peer approval, and sexual gratification, tend to be contemporaneous with it. In contrast, many of the costs of offending, such as the risk of being caught and punished, are uncertain and often delayed. The key individual difference factor in the Wilson and Herrnstein theory is the extent to which people’s behavior is influenced by immediate as opposed to delayed consequences. They suggested that individuals vary in their ability to think about or plan for the future. As the authors (Wilson & Hernnstein, 1985) surmise, “. . . time discounting becomes extremely important in explaining criminal behavior” (p. 50) as the “strength of any reward [from crime] declines with time, [and] people differ in the rate at which they discount the future” (p. 61).

Given the centrality of the time-discounting concept in two major criminological theories, one would suspect that there would be a large literature on it. Unfortunately, and as recently noted by Akerlund, Golsteyn, Gronqvist, and Lindahl (2016) empirical evidence on this issue is “essentially nonexistent” (p. 1). With respect to criminal offending, there are two relevant studies worth highlighting.

Using data from the Pathways to Desistance Study, a large sample of serious adolescent offenders followed for 7 years from mid-adolescence to early adulthood, Piquero (2016) investigated the link between time discounting, measured as perceived age at death, and self-reported offending trajectories. To measure perceived age at death, at the baseline interview, respondents were asked, “How old do you think you will live to be?” This concept can be considered a type of time-discounting measure because, in theory, youth who perceive that they will die earlier are, in turn, likely to live and act in the here-and-now by focusing on immediate rewards in lieu of placing a premium on the costs of their actions, which may not be realized until well into the future. To measure self-reported offending, respondents were interviewed and provided responses to more than 20 criminal acts, including both property and violent offenses, over the course of the study. From those responses, offending trajectories were estimated, which revealed the long-term variability in offending patterns. Piquero hypothesized that adolescents who perceive an earlier age at death would be more likely to discount the future and offend more frequently. His results showed just that: An earlier anticipated death age was linked to the most chronic offending patterns.

In a second study, Akerlund et al. (2016) used data from the Stockholm Birth Cohort to examine the relationship between time discounting and criminal behavior. Specifically, they compared a survey-based measure of time-discount rates at age 13 with criminal behavior data above 18 years. The measure of time discounting was based on responses where the children were asked “to rate the extent to which they prefer: Swedish Krona (SEK) 900 [US dollar (USD) 140] today over SEK 9,000 (USD 1,400) in 5 y using a five point scale (in 2013 year’s price level)” (p. 1). Response options were as follows: (a) Certainly SEK 900 now, (b) probably SEK 900 now, (c) cannot choose, (d) probably SEK 9,000 in 5 years, and (e) certainly SEK 9,000 in 5 years. Their measures of crime originated from two sources: “(i) interventions for social authorities due to delinquent behavior of children up to age 18; [and] (ii) the universe of criminal convictions for all individuals between ages 15 and 31” (p. 1). Their results showed that “individuals with short time horizons have a significantly higher risk of criminal involvement later in life” (p. 1). Or in other words, persons who do not express a preference for here-and-now gratification and have a willingness to delay the reward are less likely to be convicted for a crime. Despite the initial and important contributions of Piquero’s (2016) and Akerlund et al.’s (2016) time-discounting studies, respectively, they both are limited as they only considered offending in late adolescence/early adulthood and only relied on one measure of time discounting obtained at a particular age and not multiple measures of time discounting obtained at different ages to gauge the sensitivity of the measurement of time discounting.

Current Study

In the current study, we seek to extend the small extant knowledge base on the relationship between time discounting and criminal offending over a larger period of the life course than has previously been the case. Specifically, we use data from the Cambridge Study in Delinquent Development (CSDD), a longitudinal study following more than 400 South London males into late middle adulthood, along with three different time-discounting measures, collected at ages 18, 32, and 48 to predict the total number of convictions to age 56. Thus, not only does our work attempt to replicate one of the first empirical links between time discounting (assessed via money) and criminal behavior (Akerlund et al., 2016), but also, it does so using multiple measures of time discounting collected at three different periods of the life course along with criminal behavior information collected into late middle adulthood.

Data and Method

Data for this study are drawn from the CSDD, one of the longest prospective longitudinal studies known to criminology, the results of which have appeared in well more than 200 published works (Farrington, 2003; Farrington, Coid, & West, 2009; Farrington, Piquero, & Jennings, 2013; Piquero, Farrington, & Blumstein, 2007). The participants include 411 boys who were recruited in 1961-1962 from South London. The sample was generated by targeting all boys aged 8 to 9 on the registers of six primary schools within a one-mile radius of the research office. The sample was working class and mostly White (97%; Farrington, 1995; West & Farrington, 1973). Interviews have been conducted when the boys were approximately aged 8, 10, 14, 16, 18, 21, 25, 32, and most recently, at age 48. The attrition rate was very low. Of those still alive, 94% were interviewed at age 32 and 93% at age 48.

Dependent Variable

The primary dependent variable is criminal offending, here measured by the total number of convictions accrued by the males by age 56 (M = 2.051, SD = 4.187, range = 0-29). Conviction records for the men in the CSDD were intensively gathered by the research team by examining microfiche records at Scotland Yard (prior to 1994) and the Police National Computer (PNC) after 1994 (see Farrington et al., 2013, for further details).

Explanatory Variables

Time discounting

Our primary independent variable of interest is a measure of time discounting. In the CSDD, three questions on time discounting within the context of money were asked at ages 18, 32, and 48. Specifically, the males were asked the following questions: At age 18, “I would rather have £10 now than £20 in a year’s time” (61.7% true, 38.3% false); at age 32, they were asked, “I would rather have £50 now than £100 in a year’s time” (61.4% true, 38.6% false); and at age 48, they were asked, “I would rather have £50 now than £100 in a year’s time” (56.1% true, 43.9% false). Responses to the three items were strongly related to one another.

For example, between ages 18 and 32, 70.4% of those who said “now” at 18 also said “now” at 32, compared with 47.5% of those who said “later” at 18 (odds ratio [OR] = 2.62, 95% confidence interval [CI] = [1.69, 4.06]). The strongest relationship was between ages 32 and 48: 72.1% of those who said “now” at 32 also said “now” at 48, compared with 29.5% of those who said “later” at 32 (OR = 6.18, CI = [3.84, 9.95]). The relationship between ages 18 and 48 was weaker, as expected with such a long period in between interviews, but still relatively strong and statistically significant: 63.9% of those who said “now” at 18 also said “now” at 48, compared with 43.1% of those who said “later” at 18 (OR = 2.34, CI = [1.50, 3.65]). This strong relationship in time discounting over a 30-year period is remarkable. Finally, when we summed the responses to the three time-discounting items, 113 males reported that they would prefer the money now rather than in a year’s time at all three assessments. The corresponding values were 0 (n = 48), 1 (n = 81), 2 (n = 98), and 3 (n = 113).

Control Variables

Early life-course individual and environmental risk factors

The analysis controls for two key risk factor indexes that have been found in previous work with the CSDD to be related to criminal convictions, namely, an individual risk factor index and an environmental risk factor index. Following Piquero et al. (2007; see also Farrington et al., 2013), we construct an individual risk factor index and an environmental risk factor index. The individual risk factor index is the sum of 12 items, measured when the males were aged 8 to 10 (and scored dichotomously where 1 indicates the presence of the risk factor, 0 = otherwise) and include (a) low school attainment, (b) daring disposition, (c) small stature/height, (d) low non-verbal IQ, (e) nervous/withdrawn personality, (f) high extraversion, (g) high neuroticism, (h) psychomotor impulsivity, (i) dishonesty, (j) unpopularity, (k) troublesomeness, and (l) lacking concentration/restlessness (M = 2.948, SD = 2.118, range = 0-10).

The environmental risk factor index is scored similarly, using 15 dichotomous items measured at ages 8 to 10 that are summed to create an overall risk factor index. The individual items comprising the index are as follows: (a) parental harsh attitude/discipline, (b) mother was a teenager at birth of her first child, (c) behavior problems of siblings, (d) parental criminal record, (e) delinquent older sibling(s), (f) large family size, (g) poor housing, (h) low family income, (i) parental disharmony, (j) neurotic/depressed father, (k) neurotic/depressed mother, (l) low socioeconomic status, (m) separated parents, (n) poor supervision, and (o) high delinquency rate school (M = 3.561, SD = 2.924, range = 0-13). Both of the indexes, which were positively and significantly correlated with one another (r = .452, p < .05), are calculated such that higher scores indicate more risk.²

Analysis Plan

Two analyses are presented. First, we examine the bivariate relationship between all three time-discounting measures and the total number of convictions. Then, we perform a negative binomial regression model predicting the total number of convictions accumulated by the sample by age 56 from each of the time-discounting measures while controlling for the individual and environmental risk factor indexes.³

Results

Figure 1 presents the bivariate relationship between the average number of convictions summed between ages 10 and 56 and each response option for each of the three time-discounting measures obtained at ages 18, 32, and 48. As can be seen, on average, males who reported desiring the money now, as opposed to a year later, were more likely to accumulate a higher number of convictions between ages 10 and 56, on average. The differences were statistically significant at ages 18 and 48 (p <.05), but not at age 32 (p > .05).

Figure 1.

Bivariate relationship between time discounting and criminal convictions.

Although a bivariate relationship has been observed between time discounting and criminal convictions, we next estimated two negative binomial regression models to examine whether this relationship held after controlling for individual and environmental risk factors. In the first, we estimated the relationship between the time-discounting summary measures (that summed the three individual responses). These results may be found in Model 1 of Table 1. As can be seen, the relationship was positive and significant (incidence rate ratio [IRR] = 1.283, p <.05), indicating that a one-unit change in time discounting corresponded to a 28% increase in the number of convictions. In Model 2, we once again examined the relationship between time discounting but now controlled for two important risk indexes that have been found to be significantly related to criminal offending in previous analyses with the Cambridge Study (Farrington et al., 2013; Piquero et al., 2007). As can be seen, although both individual and environmental risk factor indexes were significantly related to the total number of convictions (IRR = 1.226, IRR = 1.175, p <.05, respectively), the time-discounting summary measure remains significantly related to the number of convictions (IRR = 1.260, p < .05) with virtually the same coefficient as in the original model without the inclusion of the two risk factor indexes.⁴

Table 1.

Negative Binomial Regressions Predicting Total Number of Convictions, Ages 10-56.

	Model 1		Model 2
	IRR (SE)	95% CI	IRR (SE)	95% CI
Time discounting	1.283 (0.162)*	[1.001, 1.646]	1.260 (0.137)*	[1.018, 1.561]
Individual risk			1.226 (0.062)*	[1.110, 1.354]
Environmental risk			1.175 (0.047)	[1.086, 1.272]
Log-likelihood	−560.640		−532.034
χ² statistic	χ²(1) = 3.78, p < .05		χ²(3) = 60.99, p < .05

Note. IRR = incidence rate ratio; CI = confidence interval.

p < .05.

Supplemental Analyses

Finally, we undertook two additional sets of analyses. In the first, we estimated a negative binomial regression model where we summarized the individual and environmental risk indexes, thus creating a total risk index, and included it in a model predicting the total number of convictions along with the time-discounting summary measure. Once again, although the total risk factor index was positively and significantly related to the total number of convictions, the effect for the time-discounting measure remained positive and significant (IRR = 1.249, standard error [SE] = 0.134, p < .05). And, when we recoded the total risk factor index into a dichotomous measure equal to 1 for those males who were in the top 5% of the total risk factor index, the positive and significant relationship between the time-discounting summary measure and the total number of convictions remained unchanged from the results reported above (IRR = 1.233, SE = 0.145, p < .05). Importantly, the coefficient size and strength from these two supplemental models were quite similar to those reported upon in the main analyses.

In the second supplemental analysis, we estimated a logistic regression model where we used the time-discounting measures to predict conviction status (no/yes) by age 56 to see whether time discounting differentiated non-offenders from offenders. When we performed this analysis for all four time-discounting measures (ages 18, 32, 48, and their sum), we only found a significant relationship for the age-18 measure. This was also the case whether we controlled for the two risk factor indexes or not. This leads us to conclude that time discounting distinguishes best within the population of offenders than between non-offenders and offenders. This may point, then, to the possibility that there are differences in time preferences among those who offend only once and those who offend many more times.

Discussion

Herein, we sought to contribute to the small set of studies that have investigated the linkage between time discounting and criminal offending. In so doing, we extended prior research in several ways, but primarily by assessing the relationship between time discounting (measured at three different ages: 18, 32, and 48) and criminal convictions into late middle adulthood, thereby offering the longest investigation of this presumed linkage. Our analyses showed that individuals who had greater time discounting—preferring immediate rewards (less money now compared with more money in the future)—were at increased risk of accumulating more convictions up through age 56. Moreover, this relationship held when we controlled for a series of individual and environmental risk factors as well as in different model specifications.

The substantive findings, then, support the main tenets of two popular criminological theories: Gottfredson and Hirschi’s (1990) general theory of crime and Wilson and Hernnstein’s (1985) crime and human nature—both of which place a premium on the differences across persons regarding future orientation. Specifically, the males who engaged in greater time discounting were likely those persons who were demonstrating a lack of self-control and hence not characteristics “conducive to hard work, delayed gratification, or persistence in a course of action” (Gottfredson & Hirschi, 1990, p. 115), which in turn increased their risk of accumulating a greater number of convictions by age 56. Similarly, consistent with Wilson and Herrnstein’s assumptions, time-discounting preferences significantly differentiated the males in terms of their frequency of conviction by age 56 as well. The fact that we find that time discounting predicts the frequency of offending more consistently than it does for predicting prevalence/participation suggests that time discounting may be better able to distinguish within the sample of people who offend and that the two variables (time discounting and the frequency of convictions) are linearly and positively correlated.

Although our data and analyses are important extensions of the previous research literature on time discounting and criminal behavior, a few limitations should be noted. First, our measure of time discounting was based on a single item that asked respondents to decide between a particular amount of money now as compared with a higher amount of money in a year’s time. Although this type of time-discounting measure is the norm in this line of research, subsequent studies should consider developing alternative measures that may consider different types of immediate versus delayed rewards. For example, one interesting question would ask respondents about whether they would like to lose 10 pounds now or 20 pounds in a year’s time. Not only would this provide a counter approach to measuring time discounting in a non-financial manner, but it also could assess reward structures that are built differently from money. As well, such a measure may also open new opportunities for considering gender differences, given the societal pressures of thinness in the American culture—particular to females. Second, our measure of criminal offending was based on official records. And although we were able to link self-reported time-discounting measures to officially based measures of offending, future research should consider replicating our work with self-reported offending information. Third, it would also be useful to assess the extent to which time discounting relates to other non-criminal behaviors. For example, are persons who prioritize immediate rewards more likely to engage in analogous behaviors that Gottfredson and Hirschi believe come together in the same (low self-control) persons, such as excessive drinking, erratic employment records, promiscuous sexual behavior, poor interpersonal relationships, and so on. Finally, it is important to acknowledge that although the time-discounting effect is “real” and robust as it withstands controls for two strong risk factor indices, the magnitude of its effect is more modest.

Our results also offer some suggestions about how an individual’s time discounting could be addressed, and in turn, their criminal offending as well. The first focuses on the event itself. Deterrence researchers have long appreciated that the certainty of punishment is more effective at reducing the likelihood of crime than is the severity of punishment (Nagin, 1998; Paternoster, 2010). Persons with high discount rates will not be deterred by uncertain or delayed punishments (e.g., Akerlund et al., 2016; Nagin & Paternoster, 1993; Piquero, Paternoster, Pogarsky, & Loughran, 2011; Piquero & Tibbetts, 1996). Thus, to the extent that agents of the criminal justice system (and the system itself) can increase the certainty (and celerity) of punishment, this may go a long way toward helping to re-orient individual time-discounting preferences.

A second approach focuses on the time discounter himself or herself by developing a better understanding of how persons initially begin to think about immediate versus longer term rewards. Returning to Gottfredson and Hirschi, these theorists claim that an individual’s self-control is developed due, in large part, to effective parental socialization, which includes parental monitoring, parental recognition of deviance, and appropriate punishment of deviant behavior. Thus, to the extent that parents can effectively socialize their children to exercise more self-control, a personal characteristic that is developed in the first decade or so of a child’s life, they will focus less on immediate rewards and instead place greater emphasis on long-term rewards. Fortunately, there are many evidence-based programs that are effective at improving children’s self-control, and in turn, at reducing their involvement in delinquency and criminal activity (see Piquero et al., 2016; Piquero, Jennings, & Farrington, 2010). Thus, a two-pronged strategy focusing on time discounting would consider both the characteristics of the event and the characteristics of the person.

In this regard, the fact that we find that time discounting appears to have more of a consistent effect on predicting the frequency of offending relative to predicting the prevalence/participation of offending suggests that there may be some sort of linear effect for time discounting on offending once a certain threshold (two or more offenses, for example) is met. Alternatively, it may also indicate that the consideration of the risks/rewards ratio principle, espoused by Wilson and Herrnstein, does not fully come into play for the initial commission of an offense, but it does influence criminal decision making insofar as increasing the likelihood of repeating the criminal behavior. As such, policies and interventions aimed at preventing subsequent criminal behavior should make an effort to incorporate components to minimize the extent to which active offenders engage in time discounting.

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.

Notes

References

Akerlund

Golsteyn

B. H. H.

Gronqvist

Lindahl

(2016). Time discounting and criminal behavior. Proceedings of the National Academy of Sciences, 113, 6160-6165. doi:10.1073/pnas.1522445113

Baumeister

R. F.

Bratslavsky

Muraven

Tice

D. M.

(1998). Ego depletion: Is the active self a limited resource? Journal of Personality and Social Psychology, 74, 1252-1265.

Baumeister

R. F.

Heatherton

T. F.

Tice

D. M.

(1994). Losing control: How and why people fail at self-regulation. San Diego, CA: Academic Press.

Baumeister

R. F.

Schmeichel

B. J.

Vohs

K. D.

(2007). Self-regulation and the executive function: The self as controlling agent. In Kruglanski

A. W.

Higgins

E. T.

(Eds.), Social psychology: Handbook of basic principles (pp. 516-539). New York, NY: Guilford Press.

Baumeister

R. F.

Vohs

K. D.

Tice

D. M.

(2007). The strength model of self-control. Current Directions in Psychological Science, 16, 351-355.

Duckworth

A. L.

Carlson

S. M.

(2013). Self-regulation and school success. In Grouzet

F. M. E.

Muller

Sokol

B. W.

(Eds.), Self-regulation and autonomy: Social and developmental dimensions of human conduct (pp. 208-230). New York, NY: Cambridge University Press.

Duckworth

A. L.

Gollwitzer

Kirby

Oettingen

(2013). From fantasy to action: Mental Contrasting with Implementation Intentions (MCII) improves academic performance in children. Social Psychological & Personality Science, 4, 745-753.

Duckworth

A. L.

Tsukayama

Kirby

(2013). Is it really self-control? Examining the predictive power of the delay of gratification task. Personality and Social Psychology Bulletin, 39, 843-855.

Duckworth

A. L.

Tsukayama

May

. (2010). Establishing causality using longitudinal hierarchical linear modeling: An illustration predicting achievement from self-control. Social Psychological & Personality Science, 1, 311-317.

10.

Farrington

D. P.

(1995). The twelfth Jack Tizard memorial lecture: The development of offending and antisocial behaviour from childhood: Key findings from the Cambridge Study in Delinquent Development. Journal of Child Psychology and Psychiatry, 36, 929-964.

11.

Farrington

D. P.

(2003). Key results from the First 40 years of the Cambridge Study in Delinquent Development. In Thornberry

T. P.

Krohn

M. D.

(Eds.), Taking stock of delinquency: An overview of findings from contemporary longitudinal studies (pp. 137-183). New York, NY: Kluwer.

12.

Farrington

D. P.

Coid

J. W.

West

D. J.

(2009). The development of offending from age 8 to age 50: Recent results from the Cambridge Study in Delinquent Development. Monatsschrift fur Kriminologie und Strafrechtsreform / Journal of Criminology and Penal Reform, 92, 160-173.

13.

Farrington

D. P.

Piquero

A. R.

Jennings

W. G.

(2013). Offending from childhood to late middle age: Recent results from the Cambridge Study in Delinquent Development. New York, NY: Springer.

14.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Stanford, CA: Stanford University Press.

15.

Mischel

Ebbesen

E. B.

Zeiss

A. R.

(1972). Cognitive and attentional mechanisms in delay of gratification. Journal of Personality and Social Psychology, 21, 204-218.

16.

Mischel

Shoda

Peake

P. K.

(1988). The nature of adolescent competencies predicted by preschool delay of gratification. Journal of Personality and Social Psychology, 54, 687-696.

17.

Mischel

Shoda

Rodriguez

M. L.

(1989). Delay of gratification in children. Science, 244, 933-938.

18.

Nagin

D. S.

(1998). Criminal deterrence at the outset of the twenty-first century. In Tonry

(Ed.), Crime and justice: A review of research (Vol. 23, pp. 1-42). Chicago, IL: University of Chicago Press.

19.

Nagin

D. S.

Paternoster

(1993). Enduring individual differences and rational choice theories of crime. Law & Society Review, 27, 467-469.

20.

Paternoster

(2010). How much do we really know about criminal deterrence? Journal of Criminal Law and Criminology, 100, 765-824.

21.

Piquero

A. R.

(2016). “Take my license n’ all that jive, I can’t see . . . 35”: Little hope for the future encourages offending over time. Justice Quarterly, 33, 73-99.

22.

Piquero

A. R.

Farrington

D. P.

Blumstein

(2007). Key issues in criminal career research: New analyses of the Cambridge Study in Delinquent Development. Cambridge, UK: Cambridge University Press.

23.

Piquero

A. R.

Jennings

W. G.

Diamond

Farrington

D. P.

Tremblay

R. E.

Welsh

B. C.

Gonzalez

J. M. R.

(2016). A meta-analysis update on the effects of early family/parent training programs on antisocial behavior and delinquency. Journal of Experimental Criminology, 12, 229-248.

24.

Piquero

A. R.

Jennings

W. G.

Farrington

D. P.

(2010). On the malleability of self-control: Theoretical and policy implications regarding a general theory of crime. Justice Quarterly, 27, 803-834.

25.

Piquero

A. R.

Paternoster

Pogarsky

Loughran

(2011). Elaborating the individual difference component in deterrence theory. Annual Review of Law and Social Science, 7, 335-360.

26.

Piquero

Tibbetts

S. G.

(1996). Specifying the direct and indirect effects of low self-control and situational factors in offenders’ decision making: Toward a more complete model of rational offending. Justice Quarterly, 13, 481-510.

27.

West

D. J.

Farrington

D. P.

(1973). Who becomes delinquent? London, England: Heinemann.

28.

Wilson

J. Q.

Herrnstein

R. J.

(1985). Crime and human nature. New York, NY: Simon & Schuster.