Illegal Earnings,Opportunity Structures and the Timing of Life Transitions

Abstract

We draw from theories of choice and the life course perspective to assess if earning money from crime during adolescence is associated with the timing of entry into stable employment, cohabitation and marriage and consider the timing of two key structural factors—incarceration and bachelor’s degree completion—that may attenuate any observed relationships. We analyze a subsample of the National Longitudinal Survey of Youth 1997. Results of proportional hazard models with propensity score matching suggest that illegal earnings in adolescence is associated with hastening incarceration and delaying bachelor’s degree completion. Illegal earnings in adolescence is directly associated with cohabitation, incarceration and bachelor’s degree completion are related to the timing of stable employment and marriage. Findings are discussed.

Keywords

Illegal earnings life course transitions turning points

Individuals choose the paths they follow, yet choices are always constrained by life opportunities structured by social institutions and culture. (Elder et al., 2003)

Financial returns, one dimension of success in crime (McCarthy & Hagan, 2001; Nguyen, 2020), are a key element in theories of choice. Theoretically, people choose to commit a crime if the expected marginal benefit of allocating time to crime exceeds the marginal cost of doing so (Becker, 1968; Ehrlich, 1973). The returns to crime are particularly salient because they can shift the decision calculus toward crime; yet as Draca and Machin (2015) note, this decision “seems to be the most understudied element of crime determinants that arise from the basic economic model of crime” (p. 399). We examine if financial returns from crime are an early formidable experience with long-term consequences. Financial returns from crime can influence perceptions of and preferences for illegal and legal lines of action, as well as shape and constrain criminal and non-criminal opportunities and outcomes. Over time, adolescents who earn money illegally can become entrenched in a criminal milieu, which can disrupt the desistance process (Paternoster & Bushway, 2009; Pezzin, 1995; Shover & Thompson, 1992) and delay meaningful transitions that mark the movement from adolescence to emerging adulthood, such as entrance into the labor force and romantic partnerships (Massoglia & Uggen, 2010).

Interest in the intersection of structure and choice has grown over the last 30 years, as scholars have heeded the call to incorporate social structure and social relations in the study of decision making (Coleman, 1990; Granovetter, 1985). The lion’s share of this research is interested in how larger structural or social contexts influence individual perceptions, preferences and choices. Conversely, we focus on how decision making later impacts social structure. Inherent in our framework is the notion that early experiences reverberate across the life course (Ferraro et al., 2009). Research demonstrates that early returns to conventional human capital (e.g., education, skills) have pivotal implications for life transitions and social and economic well-being (Mortimer & Shanahan, 2007; O’Rand, 2006). Educational attainment, or the lack thereof, can trigger cumulative processes of stratification that set people on different pathways of relative advantage or disadvantage—early advantage, for example, increases access to beneficial opportunity structures.

We take a life course approach, which examines age-graded patterns embedded in social institutions and contexts (Elder, 1998). Traditionally, economists have not embraced the larger structure in which decisions are embedded and sociologists have been reluctant to incorporate a choice approach to decision making with a structural analysis (McCarthy, 2002). A life course framework parsimoniously accommodates both choice and structure and is a useful overarching perspective for examining micro processes within a larger sociological orientation. We use this approach to examine illegal earnings in adolescence, opportunity structures (i.e., incarceration and bachelor’s degree completion) and the timing of life transitions (i.e., stable employment, cohabitation and marriage) in self-report data from the National Longitudinal Survey of Youth 1997 (NLSY97).

Constrained Choices

In criminology, the study of choice has advanced past the neoclassical model of rational decision making. In the neoclassical model, individuals assess the options available to them and select the option they believe will provide the greatest expected utility.¹ Studies demonstrate some support for predictions that the costs of crime (e.g., punishment) produce a deterrent effect (Apel, 2022; D. S. Nagin, 2013) while the returns to crime (e.g., income) increase the probability of future offending (Nguyen et al., 2021). Building on these findings, criminologists have worked toward specifying elements associated with decision making.²

Decision making models assume that individual perceptions and preferences influence utility functions and the options that individuals see as available to them, which in turn influence decision making. Research suggests that external structural factors—such as neighborhood disadvantage (Thomas et al., 2022), social structure (e.g., Matsueda et al., 2020) and peer dynamics (Barnum & Pogarsky, 2022; Hoeben & Thomas, 2019; McGloin & Thomas, 2016)—affect the formation of perceptions and/or preferences, which supports the notion that opportunity structures may influence behavior.

Our approach to studying the intersection of structure and choice examines how choices at one point in time yield structural constraints or opportunities in later life and draws on the notions of state dependence (D. S. Nagin & Paternoster, 1991) and cumulative disadvantage (Sampson & Laub, 1993, 1997). State dependence is the idea that continued deviant behavior is explained, in part, by the opportunities and rewards for illegal activities that stem from prior offending and the erosion of conventional opportunities (D. Nagin & Paternoster, 2000). Sampson and Laub (1997) emphasize that deviant behavior incrementally weakens adult bonds to prosocial institutions (e.g., labor force, family) and can contribute to an accumulation of structural disadvantages that limit opportunities. However, they explicitly downplay the role of individual-level processes. Indeed, they argue that structural effects of deviant behavior and stigma emerge through “social allocation mechanisms that have nothing to do with a redefinition of the self or other social-psychological processes that operate within the individual” (p. 9).³ We explore a different possibility—early life choices shape perceptions, preferences and opportunity structures, and thus shape subsequent choices over the life course. Specifically, we examine illegal earnings in adolescence, opportunity structures and the timing of stable employment and romantic relationship transitions in adulthood; thus, we theorize about how individual-level processes impact later life outcomes.

Stable Employment

Scholars who take a choice approach to understanding the overlapping nature of illegal and legal work note that the relationship between income-generating crime and legal employment involves trade-offs between crime returns, punishment costs, legal work opportunity costs and tastes for both types of work (Fagan & Freeman, 1999). Fagan and Freeman (1999) highlight the long-term consequences of financial criminal success in early life: “Whatever the origins of the decision to engage in illegal work, the onset of an illegal career is often viewed as narrowing later economic and social options for engaging in legal work, foretelling a lengthy career of illegal pursuits outside the social world of legal enterprise” (p. 228).

We build on this work to hypothesize about how illegal earnings in adolescence may influence the timing of stable employment in adulthood. Theoretically, individuals’ perceptions about the costs and benefits of crime change in response to their criminal activity (Anwar & Loughran, 2011; Lochner, 2007). Greater returns to crime can shift expectations of monetary rewards, as individuals update their expectations based on previous experiences with income-producing crime (Anwar & Loughran, 2011). Illegal earnings can also lead some people to reduce the marginal disutility of the potential costs of justice system involvement. For example, the formal (e.g., incarceration) and social (e.g., stigma) costs of punishment might not be enough to discount illegal earnings for individuals whose legal alternatives are unappealing or unavailable (Fagan & Freeman, 1999; LeBel et al., 2008). Higher criminal returns can also increase a person’s disutility of a stable legal job, including potential costs such as a fixed work schedule, less freedom and mobility during the day or night and answering to a supervisor. Put simply, early financial success in crime can alter how individuals value the perceived outcomes of income-generating crime and legal employment. Recent research finds that adolescent delinquency is associated with a slower transition into full time employment (Kang, 2019). We expect that illegal earnings during adolescence will also reduce or delay entry into stable legal employment.

Illegal earnings in adolescence can be indirectly related to the timing of stable employment through opportunity structures. The choice to engage in income-generating crime can increase the likelihood of continued involvement in offending, thereby raising the probability of justice system contact in adulthood. Arrests, charges and incarceration can reduce opportunities for work and wage premiums in the legal labor market. A growing body of evidence documents pervasive labor market barriers faced by applicants with a criminal record, particularly those with a history of incarceration (e.g., Pager, 2003; Sugie et al., 2020; Uggen et al., 2014; Western & Pettit, 2000). Additionally, time spent in correctional facilities can provide opportunities to accumulate criminal capital—including criminal social (e.g., potential co-offenders and mentors) and criminal human (e.g., skills and knowledge) capital—that translates into greater illegal earnings post-release (Nguyen et al., 2017). In other words, incarceration, a potential consequence of the choice to earn money illegally, can shift access to both legal and illegal opportunity structures. Lack of good opportunities in the legal labor market and low subjective expectations of the returns to a legal job can constrain a person’s choice and influence the decision to further engage in criminal behavior, thereby hastening the chance of incarceration and delaying entry into stable employment.

Early financial success in crime can also reduce the utility of spending time and resources in growing conventional human capital (Cloward & Ohlin, 1960; Machin & Meghir, 2004), including the choice to invest in post-secondary education (Lochner, 2004). Illegal earnings in adolescence can increase opportunities to grow criminal social and criminal human capital, assets that further opportunities for illegal income (Bouchard & Nguyen, 2010; Loughran et al., 2013; Morselli et al., 2006; Tremblay, 1993). A central feature of time allocation models is the notion that criminal capital is a substitute for human capital and can improve prospects in the crime market (see, e.g., Lochner, 2004; Mocan et al., 2005). Voluntary investment in post-secondary education may be particularly unappealing and thus delayed for individuals already earning money from crime because education is financially costly and reduces time for leisure and illegal income-generating activities.

Bachelor’s degrees have, however, become increasingly prevalent and increasingly important in shaping labor market opportunities (Baker, 2014). In 2020 about 37.9% of the U.S. population aged 25 and older were college graduates, which is a significant increase from 7.7% in 1960 (Statista Research Department, 2023). Moreover, education and labor market outcomes are strongly linked: individuals with a bachelor’s degree are more likely to be employed and have higher earnings than their counterparts (U.S. Bureau of Labor Statistics, 2022). We expect illegal earnings in adolescence to delay bachelor’s degree completion, which we expect to hasten time to stable employment. As such, any relationship between illegal earnings in adolescence and the timing of stable employment may be offset among individuals who complete a bachelor’s degree.

Romantic Relationships

Illegal success may also affect the subjective expectations, marginal utilities and opportunities associated with romantic relationship trajectories. Several scholars use a choice approach to understand romantic relationships and activities (e.g., sex) across a variety of contexts (Becker, 1981; Oppenheimer, 1988) including adolescence (Loewenstein & Furstenberg, 1991) and emerging adulthood (Oppenheimer, 2003). In these models, the choice to pursue romantic relationships is influenced by several factors, including a preference for the potential returns relationships provide, an acceptance of the likely costs that accompany relationships, access to relationship opportunities and a set of personal resources (e.g., economic, sexual and other types of capital) that attract potential partners. Economic resources are a key factor in cost-benefit analyses during mate selection. As such, illegal earnings in adolescence may increase opportunities for romantic partnerships.

Money increases opportunities to buy goods and services that can enhance one’s attractiveness to potential mating partners. Visible expenditures and conspicuous consumption convey high status and may help individuals differentiate themselves in the competition for partners in cohabitation and marriage markets (Eastwick et al., 2014; Huang et al., 2011). Individuals who are financially successful may update their subjective expectations of attracting a partner because they believe that their economic resources make them more attractive partners, which may be especially likely in late adolescence and emerging adulthood where the visual displays of economic resources are key to attracting an intimate partner. Anderson (2000) notes, for example, that young people spend much of their time in “staging areas” where people hangout. In these settings, young people are acutely aware of the conspicuous consumption of their peers and have subjective expectations of attracting partners.

Money may also lead to a change in the expected marginal utility for romantic relationships. Utility functions can be state dependent and impacted by experiences and perceptions (see Heckman, 1981). Individuals who believe that they have low odds of attracting a partner may devalue the expected marginal utility of a relationship to cope with their perceived low odds and elevate the expected marginal utility once they obtain resources. The competitive advantage of illegal income is likely greater in late adolescence and emerging adulthood when competitors who depend on legal income have fewer opportunities to generate it. Research finds that both income and a higher relative income (i.e., higher predicted income based on background characteristics) are positively associated with cohabitation (Clarkberg, 1999). Other studies document the extent to which low income is particularly disadvantaging for transitioning to cohabitation (Oppenheimer, 2003) and show that cohabitation and long-term potential income are essentially unrelated (Xie et al., 2003), underscoring the value of current earnings, particularly earlier in life (Smock, 2004).

Although marriage is the most common romantic relationship studied in criminology, cohabitation is now more common than marriage in many countries, including the U.S. Americans aged 18 to 29 are almost twice as likely to have cohabited (44%) as to have married (23%), and more adults aged 18 to 44 have lived with an unmarried partner (59%) than have ever been married (50%) (Horowitz et al., 2019). There is also some evidence that adolescent delinquency increases the likelihood of cohabitation (Kang, 2019; Lonardo et al., 2010), while decreasing the likelihood of marriage (Kang, 2019). Thus, we expect that illegal earnings in adolescence will hasten the transition to cohabitation but will have a weaker effect on the transition to marriage given the latter’s declining popularity.

Illegal earnings in adolescence may be indirectly related to the timing of cohabitation and marriage through opportunity structures. Incarceration may hamper the relationship between financial success in adolescence and transitions into romantic relationships. Though numerous studies investigate incarceration and relationship dissolution (Massoglia et al., 2011; Siennick et al., 2014; Turney, 2015), fewer studies investigate incarceration and transitions into cohabitation or marriage. Using the NLSY97, Apel (2016) estimates hazard models and finds that incarceration is a long-term impediment to the transition to marriage but not cohabitation. Also using the NLSY97, Bacak and Kennedy (2015) estimate marginal structural models with inverse probability weighting and find that incarceration decreases the chance of marriage. Huebner (2007) also finds that incarceration has a negative effect on marriage for men using data from the NLSY79. We expect incarceration to delay the timing of marriage but may have no impact on the timing of cohabitation. As such, any relationship between illegal earnings in adolescence and the timing of marriage may be offset by incarceration.

Bachelor’s degree completion and cohabitation rates in the United States have risen substantially over time, while marriage rates have declined and stagnated. Across education levels, most individuals have cohabited or been married at some point by the age of 44. For individuals with a bachelor’s degree, marriage has remained a more common life transition, and the rise in cohabitation has offset the decline in marriage (Horowitz et al., 2019). The positive relationship between education and marriage in the U.S. supports the idea of the economic and cultural attractiveness of having a college degree in the marriage market (McClendon et al., 2014). It is thus possible that holding a bachelor’s degree makes one more competitive in romantic partnership markets. We expect bachelor’s degree completion to hasten time to cohabitation and marriage. Thus, any relationship between illegal earnings in adolescence and the timing of cohabitation and marriage may be offset by bachelor’s degree completion.

Data and Methods

We analyze data from the National Longitudinal Survey of Youth 1997 (NLSY97), a longitudinal sample of 8,984 individuals born between January 1, 1980 and December 31, 1984 living in the United States during the initial survey round collected in 1997 to 1998. The NLSY97 is composed of two probability-based household samples: (a) a nationally representative cross-sectional sample (n = 6,748) and (b) an oversample of Black and Hispanic youth (n = 2,236). Interviews were conducted annually from 1997 to 2011 and biannually since. As such, respondents are followed from age 12 to 18 through age 34 to 40, with data collection ongoing.

We impose sample restrictions in line with our goal of examining whether illegal earnings in adolescence are associated with the timing of life transitions. First, we select respondents with an interview at age 12 and/or 13 (n = 2,837). By restricting our sample to the youngest respondents, we are able to examine illegal earnings over a greater period of adolescence. Second, we select respondents who report any valid information on our treatment of interest—illegal earnings at ages 14 to 17—which resulted in a loss of 288 respondents. This nets a maximum working sample of 2,549 respondents, 284 of whom reported first-time illegal earnings at ages 14 to 17.⁴

The NLSY97 is uniquely suited for our study because it captures information on illegal earnings in adolescence as well as key life transitions including employment, cohabitation, and marriage for over 20 years. We structure data on stable employment and romantic relationship formation by age in months (detailed below), which allows for fine-grained analyses of variation in the timing of these life transitions.

Key Measures

Illegal earnings: Respondents are asked about money earned from stealing something worth more than $50, other property crimes and drug selling specific to a calendar year. We use birth year to determine age at the start of the calendar year in which illegal earnings are reported. Empirical research documents considerable variability in the financial returns to crime. For some, illegal earnings are trivial and sporadic but for others crime is profitable, often exceeding what individuals could earn in the legitimate market (Freeman, 1996; Loughran et al., 2013; McCarthy & Hagan, 2001). We construct three binary indicators of illegal earnings between 14 and 17 years old: any illegal earnings (n = 281), $100+ at any age 14 to 17 (n = 243) and $500+ at any age 14 to 17 (n = 159).⁵ These cutoffs approximate the terciles of the distribution of illegal earnings among the analytic sample.

Stable employment: We conceptualize first stable employment as working for the same employer for 35+ hours a week on average for 12 months or more (Bushway, 1998; Holzer & Lalonde, 1999). We calculate age in months at first stable employment by using respondent birth month and year and the month and year of first stable employment (i.e., the month and year a respondent first meets the stable employment criteria).

Romantic relationship formation: We use two indicators of romantic relationship formation: age in months at first cohabitation and first marriage. We calculate age in months using respondent birth month and year and the month and year of first cohabitation and marriage.

Incarceration: We measure incarceration by age in months at first adult incarceration. We calculate age in months using respondent birth month and year and the month and year of first adult incarceration. Among respondents who reported incarceration in adulthood, the median months incarcerated is 12.5 months.

Bachelor’s degree: We measure post-secondary education by age in months at the completion of a bachelor’s degree. We calculate age in months using respondent birth month and year and the month and year of bachelor’s degree completion.

Analytic Plan

Our analytic plan unfolds in several steps. First, we compare the prevalence and age at first stable employment, cohabitation and marriage for respondents who report earning money illegally at age 14 to 17 versus those who do not. However, individuals who earn money illegally likely differ systematically from individuals who do not in observable and non-observable ways (Loughran et al., 2013), which makes unadjusted comparisons difficult. Our second step is to compare age 12 to 13 covariates between adolescent illegal earners and non-illegal earners to assess how they differ.

Third, we adjust for selection into illegal earnings during adolescence—any, $100+, and $500+—by estimating propensity scores, or balancing scores, which we use to match treated (those with illegal earnings at age 14–17) with control (no illegal earnings) respondents with the closest propensity score. We balance on 37 covariates measured at age 12 to 13 to prevent post-treatment bias, reduce the impact of confounders and account for selection into illegal earnings at age 14 to 17 (Rosenbaum, 2002; Rosenbaum & Rubin, 1983). As such, treated and control respondents have similar distributions of these covariates at age 12 to 13. The covariates tap into theoretically informed domains including demographics; household and contextual characteristics; parenting; schooling; health; attitudes; expectations; time use; peer behavior; substance use; and delinquency. We take an iterative approach for model selection to maximize covariate balance (1:1, nearest neighbor).

Fourth, we estimate Cox proportional hazard models with propensity score matching to examine if earning money illegally in adolescence is related to the timing of opportunity structures (i.e., incarceration and bachelor’s degree completion) and life transitions (i.e., stable employment, cohabitation and marriage). We also examine if opportunity structures attenuate any observed relationships between illegal earnings and the timing of life transitions. Event history models are ideal for examining timing to transitions because they capture an occurrence of an event over time (Allison, 2004; Cleves et al., 2016). Event history models are superior to traditional regression methods because they account for censoring of cases (i.e., cases that do not experience a transition in the follow-up period) and help with estimating the precision of the duration of the follow-up period in which cases remain transition free.⁶ By directly comparing the timing of transitions between treated and control respondents in the matched sample, we estimate the average treatment on the treated effect.

Finally, we conduct postestimation sensitivity analyses to assess the robustness of our results to potential unmeasured confounders (VanderWeele & Ding, 2017). We calculate the “E-value”—the minimum strength of association that an unmeasured confounder would need to have with both a treatment and an outcome to fully explain a treatment–outcome association, conditional on the measured covariates (Linden et al., 2020; VanderWeele & Ding, 2017). E-values are measured on the risk-ratio scale and have a lowest possible value of 1, which indicates that no unmeasured confounder is needed to explain away an observed association. The higher an E-value, the stronger the confounder associations must be to explain away an effect.

Results

Unadjusted Comparisons

We begin by comparing the prevalence of life transitions for respondents who did and did not report any illegal earnings at age 14 to 17. As shown in Panel A of Table 1, the prevalence of entering into stable work (85% vs. 84%) and marriage (47% vs. 50%) is similar for illegal earners and non-illegal earners. In contrast, cohabitation (77% vs. 63%, p < .001) is more prevalent among illegal earners than non-illegal earners. Conditional on experiencing an adult life transition, the timing of life transitions among the unadjusted sample shows patterns consistent with those above for prevalence. As shown in Panel B of Table 1, illegal earners and non-illegal earners report their first stable job at approximately 23 years old and first marriage at about 26 years old. Yet, age at first cohabitation occurs earlier for illegal earners, at about 23 years old compared to approximately 24 years old for non-illegal earners (p < .001).

Table 1.

Comparison of Life Transitions With Unadjusted Groups.

Life Transitions	Treated	Control	p
Panel A
Any stable work	0.846	0.844	.949
Any cohabitation	0.767	0.633	.000
Any marriage	0.467	0.502	.185
Panel B
Age stable work	23.092	22.928	.464
Age cohabitation	22.685	23.565	.001
Age marriage	25.657	25.671	.966

We next compare covariates at age 12 to 13 for adolescents who report earning illegal money between 14 and 17 years old (n = 284) and those who do not (n = 2,549). We compare the differences in covariates between illegal and non-illegal earners in terms of an ordinary difference in means test (t-statistic) and its associated p-value for a test of no association, where a p-value <.10 suggests an imbalance (i.e., illegal earners and non-illegal earners are not comparable on the covariate). We initially test 65 covariates prior to treatment; 37 of the 65 covariates are out of balance.

Table 2 displays the means of the 37 covariates conditional on illegal earnings. As expected, substance use—smoking (42% vs. 21%, p < .001), drinking alcohol (37% vs. 21%, p < .001), and smoking marijuana (15% vs. 5%, p < .001)—is more prevalent among illegal earners. Illegal earners also score significantly higher on delinquency variety (1.77 vs. 0.79, p < .001) and delinquent peers (1.77 vs. 1.63, p < .001) measures. In line with prior research, these findings suggest that there are pre-existing differences between individuals who earn money illegally and those who do not (e.g., Loughran et al., 2013; Nguyen et al., 2021). These differences are arguably associated with both illegal earnings and life transitions, necessitating an approach that addresses selection into illegal earnings.

Table 2.

Covariates at Age 12 or 13 and Initial Balance.

Covariate	Treated	Control	t	p > t
School unattachment scale	1.952	1.871	3.230	.001
A lot of cheating on tests and assignments	2.410	2.502	−1.890	.058
No. days/week does homework	3.713	3.898	−3.150	.002
No. weekdays watching tv	4.590	4.475	2.100	.036
Parental monitoring by mother	9.898	10.679	−4.740	.000
Family routines	14.592	15.734	−3.890	.000
Autonomy, control, and limit setting	3.273	3.433	−2.030	.043
Limit-breaking (youth report)	0.527	0.424	4.030	.000
Limit-breaking (parent report)	0.414	0.300	5.000	.000
Family/home risk index	326.910	258.200	5.990	.000
Physical environment risk index	153.940	135.680	2.600	.009
Enriching environment index	1.646	1.766	−2.940	.003
Gender	1.314	1.512	−7.620	.000
Non-White	0.585	0.526	2.250	.025
Optimism	2.774	2.863	−2.270	.023
Ever smoked	0.423	0.210	9.600	.000
Ever drank	0.366	0.206	7.310	.000
Ever smoked marijuana	0.146	0.054	6.990	.000
Delinquency variety score	1.771	0.792	13.960	.000
Ever arrested	0.076	0.030	4.560	.000
See someone get shot/shot at <12	0.181	0.094	5.350	.000
Ever live through hard times (parent interview)	0.083	0.056	2.280	.023
Any gangs in neighborhood or school	0.529	0.395	5.230	.000
Any brothers, sisters, friends in gang	0.235	0.165	3.500	.000
Had something stolen from school	0.329	0.239	3.960	.000
Threatened to be hurt at school	0.907	0.530	5.590	.000
In a fight at school	0.681	0.359	6.490	.000
Late for school without an excuse	1.114	0.728	4.910	.000
Days absent from school in Fall term	3.604	3.161	2.800	.005
% Chance released without charge if stole car	29.034	31.967	−1.680	.094
% Chance pay fine and released if stole car	49.974	53.909	−2.130	.033
Delinquent peers score	1.770	1.628	3.580	.000
Peers go to church regularly	2.768	2.910	−2.350	.019
Peers do volunteer work	2.169	2.306	−2.240	.025
Ever been on a date	0.378	0.178	9.500	.000
Live with both biological parents	2.732	2.401	3.370	.001
Household size	4.500	4.638	−1.770	.077

Propensity Scores: Balancing Illegal and Non-Illegal Earners

We estimate a propensity score for each respondent to account for pre-existing differences between our comparison groups and use this estimate as a method for creating balance on key covariates that may confound treatment effect estimates.⁷ Our goal is to examine the effect of illegal earnings in adolescence for youth who have similar characteristics (e.g., early deviance, deviant peers, family and school attachment). We employ a binary logit model and regress illegal earnings at age 14 to 17 (vs. none) on the 37 covariates to estimate the predicted probability of illegal earnings (i.e., the propensity score). We match our illegal earnings group with a comparable control group using several approaches, including caliper matching and nearest neighbor. Our final matching is 1:1 nearest neighbor matching without replacement. Overall, results are comparable across matching approaches. Only three cases are off common support, resulting in 281 “treated” respondents and 281 “control” respondents.⁸ Table 3 shows that after matching, the treated and control group do not significantly differ on any of the 37 covariates. To be sure, illegal earners are effectively matched with a “control” group of non-illegal earners who have similar risk factors at age 12 to 13. During the treatment period (ages 14–17), some members of the control group participated in illegal activities but did not report illegal earnings.

Table 3.

Covariates at Age 12 or 13 After Matching.

Covariate	Treated	Control	t	p>t
School attachment scale	1.951	1.951	−0.010	.991
A lot of cheating on tests and assignments	2.414	2.363	0.770	.440
No. days/week does homework	3.718	3.737	−0.240	.809
No. weekdays watching tv	4.589	4.567	0.360	.722
Parental monitoring by mother	9.897	9.849	0.220	.828
Family routines	14.591	14.676	−0.210	.830
Autonomy, control, and limit setting	3.276	3.254	0.210	.831
Limit-breaking (youth report)	0.526	0.540	−0.430	.668
Limit-breaking (parent report)	0.413	0.410	0.090	.927
Family/home risk index	326.030	331.120	−0.300	0.766
Physical environment risk index	153.120	154.800	−0.180	.858
Enriching environment index	1.647	1.627	0.390	.699
Gender	1.314	1.291	0.750	.455
Non-White	0.586	0.593	−0.210	.834
Optimism	2.774	2.794	−0.390	.695
Ever smoked	0.422	0.404	0.520	.601
Ever drank	0.365	0.362	0.070	.943
Ever smoked marijuana	0.144	0.142	0.100	.917
Delinquency variety score	1.754	1.681	0.600	.551
Ever arrested	0.073	0.073	0.000	.996
See someone get shot/shot at <12	0.179	0.182	−0.130	.896
Ever live through hard times (parent interview)	0.083	0.094	−0.570	.570
Any gangs in neighborhood or school	0.528	0.520	0.230	.817
Any brothers, sisters, friends in gang	0.233	0.242	−0.310	.757
Had something stolen from school	0.330	0.289	1.280	.199
Threatened to be hurt at school	0.898	0.917	−0.180	.859
In a fight at school	0.676	0.612	0.750	.452
Late for school without an excuse	1.107	1.017	0.770	.440
Days absent from school in Fall term	3.589	3.422	0.800	.426
% Chance released without charge if stole car	28.892	27.935	0.430	.669
% Chance pay fine and released if stole car	49.860	49.247	0.250	.804
Delinquent peers score	1.768	1.724	0.790	.430
Peers go to church regularly	2.770	2.828	−0.750	.456
Peers do volunteer work	2.169	2.262	−1.190	.236
Ever been on a date	0.376	0.383	−0.200	.843
Live with both biological parents	2.727	2.634	0.710	.476
Household size	4.504	4.501	0.020	.981

Illegal Earnings and Opportunity Structures

Tables 4 and 5 display the hazard ratios and confidence intervals for our three illegal earnings measures and time to incarceration and bachelor’s degree completion, respectively. As shown in Table 4, illegal earnings in adolescence significantly hastens time to incarceration in adulthood, and this rate increases as illegal earnings increase; reporting any illegal earnings in adolescence increases the expected hazard of incarceration by approximately 1.9 times (p < .01), $100+ increases the expected hazard of incarceration by 3.5 times (p < .001) and $500+ increases the expected hazard of incarceration by four times (p < .001). Sensitivity analyses suggest that the observed effects are reasonably robust with E-values ranging from 2.48 to 4.56 with confidence intervals that range from 1.67 to 3.08. Evidence that illegal earnings in adolescence hastens time to incarceration is reasonably strong, especially as the amount of illegal earnings increases.

Table 4.

Cox Hazard Models—Time to Incarceration.

Illegal Earnings Cut-offs	(1)	(2)	(3)	Sensitivity
Any illegal earnings	1.897**			2.485
N = 562	[1.286, 2.798]			[1.667]
$100+ illegal earnings		3.518***		4.141
N = 486		[2.219, 5.576]		[2.855]
$500+ illegal earnings			4.005***	4.564
N = 318			[2.427, 6.609]	[3.081]

Note. Exponentiated coefficients; 95% confidence intervals in brackets.

p < .10. *p < .05. **p < .01. ***p < .001.

Table 5.

Cox Hazard Models—Time to Degree Completion.

Illegal Earnings Cut-offs	(1)	(2)	(3)	Sensitivity
N = 562
Any illegal earnings	0.724 [0.481, 1.090]			2.105 [1.000]
N = 486
$100+ illegal earnings		0.389*** [0.246, 0.615]		3.231 [2.147]
N = 318
$500+ illegal earnings			0.446** [0.242, 0.824]	2.881 [1.549]

Note. Exponentiated coefficients; 95% confidence intervals in brackets.

p < .10. *p < .05. **p < .01. ***p < .001.

Conversely, greater amounts of illegal earnings in adolescence are associated with a delay in bachelor’s degree completion. As shown in Table 5, earning $100+ and $500+ are related to a 61% (p < .001) and a 55% (p < .01) reduction in the expected hazard of bachelor’s degree completion, respectively. The point estimate for any illegal earnings is also negative (HR = 0.72) but is not statistically significant. Sensitivity analyses suggest that the observed effects are reasonably robust, with E-values ranging from 2.88 to 3.23 with confidence intervals that range from 1.55 to 2.15. Overall, findings suggest that illegal earnings in adolescence is related to later life opportunity structures through hastening incarceration and delaying bachelor’s degree completion.

Illegal Earnings and Life Transitions

Tables 6 to 8 display the results of Cox hazard models with propensity score matching predicting time to stable employment in adulthood, cohabitation, and marriage, respectively. We first estimate the bivariate relationship between illegal earnings in adolescence and the timing of life transitions (models 1) then estimate the bivariate relationship between incarceration (models 2) and bachelor’s degree completion (models 4) and the timing of life transitions. Finally, we examine if incarceration (models 3) and bachelor’s degree completion (models 5) attenuate the relationship between illegal earnings and the timing of life transitions.

Table 6.

Cox Hazard Models—Time to Stable Employment.

Illegal Earnings Cut-offs	(1)	(2)	(3)	(4)	(5)	Sensitivity
N = 562
Any illegal earnings	1.007 [0.841, 1.205]		1.015 [0.848, 1.216]		1.016 [0.849, 1.216]
Incarceration		0.861 [0.631, 1.174]	0.859 [0.629, 1.172]
Degree completion				2.595*** [1.831, 3.678]	2.597*** [1.832, 3.682]	3.258 [2.406]
N = 486
$100+ illegal earnings	0.947 [0.782, 1.147]		0.968 [0.796, 1.176]		1.000 [0.824, 1.214]
Incarceration		0.828 [0.599, 1.143]	0.836 [0.602, 1.162]
Degree completion				2.536*** [1.803, 3.567]	2.536*** [1.797, 3.579]	3.195 [2.363]
N = 318
$500+ illegal earnings	1.065 [0.838, 1.352]		1.105 [0.866, 1.410]		1.061 [0.835, 1.347]
Incarceration		0.798 [0.557, 1.142]	0.774 [0.536, 1.116]
Degree completion				3.408*** [2.091, 5.554]	3.405*** [2.089, 5.550]	4.039 [2.709]

Note. Exponentiated coefficients; 95% confidence intervals in brackets.

p < .10. *p < .05. **p < .01. ***p < .001.

Table 7.

Cox Hazard Models—Time to Cohabitation.

Illegal Earnings Cut-offs	(1)	(2)	(3)	Sensitivity	(4)	(5)	Sensitivity
N = 562
Any illegal earnings	1.409*** [1.157, 1.716]		1.389** [1.138, 1.694]	1.822 [1.414]		1.413*** [1.160, 1.721]	1.856 [1.455]
Incarceration		1.350⁺ [0.954, 1.910]	1.252 [0.882, 1.777]
Degree completion					1.095 [0.753, 1.592]	1.124 [0.773, 1.635]
N = 486
$100+ illegal earnings	1.296* [1.055, 1.591]		1.254* [1.017, 1.547]	1.615 [1.121]		1.317** [1.072, 1.619]	1.714 [1.277]
Incarceration		1.437* [1.029, 2.007]	1.329 [0.945, 1.869]
Degree completion					1.770** [1.236, 2.535]	1.821** [1.270, 2.609]	2.393 [1.641]
N = 318
$500+ illegal earnings	1.238⁺ [0.964, 1.590]		1.194 [0.923, 1.545]			1.252⁺ [0.974, 1.609]
Incarceration		1.370 [0.938, 2.000]	1.284 [0.869, 1.895]
Degree completion					1.356 [0.836, 2.200]	1.398 [0.861, 2.272]

Note. Exponentiated coefficients; 95% confidence intervals in brackets.

p < .10. *p < .05. **p < .01. ***p < .001.

Table 8.

Cox Hazard Models—Time to Marriage.

Illegal Earnings Cut-offs	(1)	(2)	(3)	(4)	(5)	Sensitivity
N = 562
Any illegal earnings	0.975 [0.766, 1.242]		1.015 [0.796, 1.294]		0.988 [0.776, 1.258]
Incarceration		0.582* [0.381, 0.889]	0.580* [0.379, 0.888]			2.273 [1.391]
Degree completion				2.377*** [1.700, 3.324]	2.376*** [1.699, 3.323]	3.026 [2.241]
N = 486
$100+ illegal earnings	0.758* [0.589, 0.975]		0.808 [0.626, 1.044]		0.813 [0.630, 1.048]
Incarceration		0.577** [0.380, 0.876]	0.615* [0.402, 0.941]			2.147 [1.255]
Degree completion				2.233*** [1.583, 3.152]	2.134*** [1.506, 3.025]	2.760 [1.987]
N = 318
$500+ illegal earnings	0.956 [0.698, 1.310]		1.074 [0.777, 1.484]		0.991 [0.722, 1.360]
Incarceration		0.543* [0.330, 0.892]	0.529* [0.317, 0.881]			2.477 [1.408]
Degree completion				2.130** [1.325, 3.425]	2.128** [1.321, 3.428]	2.753 [1.720]

Note. Exponentiated coefficients; 95% confidence intervals in brackets.

p < 0.10. *p < .05. **p < .01. ***p < .001.

As shown in Table 6, none of the illegal earnings measures are related to the timing of stable employment (HR = 0.95–1.07, n.s., models 1). At the bivariate level, incarceration delays time to stable employment but fails to meet significance (HR = 0.80–0.86, n.s., models 2) and bachelor’s degree completion hastens time to stable employment (HR = 2.54–3.41, p < .001, models 4), with the strongest point estimate for $500+ (HR = 3.41, p < .001, model 4). Findings for illegal earnings, incarceration and bachelor’s degree completion are substantively similar when illegal earnings and incarceration (models 3) and illegal earnings and bachelor’s degree completion (models 5) are included simultaneously. Sensitivity analyses suggest that the observed effect of bachelor’s degree completion on time to stable employment is reasonably robust (E = 3.26–4.04). Overall, there does not appear to be a relationship between illegal earnings in adolescence and the timing of stable employment, yet incarceration delays time to stable employment (although not significant) while bachelor’s degree completion hastens time to stable employment.

As shown in Table 7, illegal earnings hasten time to cohabitation at the bivariate level (HR = 1.24–1.41, p < .10, models 1). Incarceration (HR = 1.35–1.44, models 2) and bachelor’s degree completion (HR = 1.10–1.77, models 4) also trend positively but significance varies by amount of illegal earnings; incarceration is significant for any and $100+, while bachelor’s degree completion is only significant for $100+. When illegal earnings and incarceration are simultaneously included in model 3, there is a slight attenuation of illegal earnings (HR = 1.39, p < .001) and the point estimate for incarceration remains positive but is reduced and no longer significant (HR = 1.25, n.s.). There is a similar trend for the $100+ and $500+ models. When illegal earnings and bachelor’s degree completion are simultaneously included in model 5, the point estimates slightly increase for both illegal earnings (HR = 1.41, p < .01) and bachelor’s degree completion (HR = 1.12, n.s.). Again, the trend is similar for the $100+ and $500+ models. These findings suggest that there may be some mutual mediation between illegal earnings and incarceration and some mutual suppression between illegal earnings and bachelor’s degree completion on the timing of cohabitation. However, sensitivity analyses suggest that the observed effects between illegal earnings and time to cohabitation are potentially sensitive to confounders (E = 1.62–1.86) and are less robust than the relationship between bachelor’s degree completion and time to cohabitation (E = 1.71 vs. E = 2.39) in the $100+ model. Overall, there is some suggestive evidence that illegal earnings, incarceration and bachelor’s degree completion all hasten time to cohabitation. However, although the most consistent, the finding that illegal earnings hasten time to cohabitation may be sensitive to confounding. Moreover, the relationship between illegal earnings and the timing of cohabitation does not appear to operate through our indicators of opportunity structures.

As shown in Table 8, illegal earnings in adolescence is inconsistently related to the timing of marriage in adulthood: any illegal earnings (HR = 0.98, n.s.) and $500+ (HR = 0.96, n.s.) are not related to the timing of marriage, but $100+ delays marriage (HR = 0.76, p < .05). At the bivariate level, incarceration delays marriage (HR = 0.54–0.58, p < .05, models 2) and bachelor’s degree completion hastens marriage (HR = 2.13–2.38, p < .01, models 4). Across all models 3 and 5, both indicators of opportunity structures retain significance when illegal earnings is simultaneously included. However, the relationship between $100+ illegal earnings and time to marriage is attenuated by the inclusion of incarceration (HR = 0.81, n.s., model 3) and bachelor’s degree completion (HR = 0.81, n.s., model 5), though the point estimates remain negative. Sensitivity analyses suggest that the relationships between incarceration (E = 2.15–2.48) and bachelor’s degree completion (E = 2.75–3.03) and time to marriage are moderately robust. Overall, opportunity structures appear to impact the timing of marriage in adulthood in the expected directions but results are mixed for the relationship between illegal earnings and the timing of marriage.⁹

To summarize, our findings show that compared to similarly situated adolescents, those who make money illegally experience incarceration earlier and bachelor’s degree completion later in adulthood. We also find that illegal earnings in adolescence accelerates the timing of cohabitation (although this finding may be sensitive to unmeasured confounders) but does not appear to affect the timing of stable employment or marriage. Yet, bachelor's degree completion hastens stable employment and marriage while incarceration delays the timing of marriage. Overall, our findings suggest that making money illegally early in life impacts later opportunities which, in turn, opens or constrains choices and the timing of some subsequent life transitions.

Discussion

Research has linked social and institutional structure to a diverse array of individual choices including health (Bird & Rieker, 2008; Vuolo et al., 2016), voting (Oskarson, 2005) and employment (McRae, 2003) decisions. As Bird and Rieker (2008) note, people’s choices, from where to live to what careers to pursue, are informed by the opportunities available to them. We take a life course approach, whereby we focus on early individual choices and how these choices shape or constrain later life opportunities. Specifically, we examine the link between early financial returns to crime, opportunity structures (i.e., incarceration and bachelor’s degree completion) and life transitions (i.e., stable employment, cohabitation and marriage). We hypothesize that early success in crime can impact perceptions, preferences and opportunity structures, which may alter the timing of life transitions.

Several interesting findings emerge. First, we find that early success in crime may shape opportunity structures that impact later life outcomes. We theorize that returns to crime can encourage continued offending and reduce the utility of spending time and resources in growing conventional human and social capital (Cloward & Ohlin, 1960; Machin & Meghir, 2004), including educational investment (Lochner, 2004). As expected, we find that illegal earnings in adolescence hastens time to incarceration and delays bachelor’s degree completion. These findings are consistent with notions of state dependence and cumulative disadvantage and align with prior research. Indeed, other analyses find that expectations of greater illegal earnings delay the desistance process (Pezzin, 1995; Shover & Thompson, 1992), with incarceration being one potential consequence. Likewise, research documents that delinquent adolescents put forth less effort in school and have lower grades and test scores, lower educational expectations and lower odds of matriculating into and graduating from college (Siennick & Staff, 2008).

Second, we find that illegal earnings in adolescence is related to the timing of one life transition: cohabitation. As expected, illegal earnings in adolescence accelerates the timing of cohabitation,¹⁰ which suggests that illegal money in adolescence may attract potential mates and that many of these mates become cohabitation partners. Contrary to our expectations, illegal earnings in adolescence is not related to the timing of stable employment or marriage. Some prior research finds that adolescent illegal activity is associated with delays in the timing of fulltime employment and marriage (Kang, 2019) while others find little evidence of “spillover” effects of adolescent delinquency on areas of adult life (e.g., work, family, friendships, and mental health; Hagan, 1991; Jessor et al., 1994). It is possible that, for most youth, illegal earnings in adolescence are not especially detrimental to these later life transitions.

Third, we find some evidence that opportunity structures are related to the timing of life transitions. As expected, we find that incarceration delays stable employment (though not significantly) and marriage. These findings are consistent with cumulative disadvantage, or that incarceration can trigger cumulative processes of stratification by limiting opportunities for stable employment and marriage. The strongest and most consistent findings across our analyses are that bachelor’s degree completion hastens the timing of stable employment and marriage. These findings also indicate cumulative processes of stratification, but for advantage rather than disadvantage. Our analytic approach should be considered when substantively interpreting the impact of bachelor’s degree completion on the timing of life transitions. Given that bachelor’s degree completion is relatively non-normative among delinquent adolescents, it is likely to have a strong impact on subsequent life transitions. Brand and Xie (2010) examine the heterogeneous effects of college education and observe that individuals who are least likely to obtain a college education benefit the most monetarily. Further examination of “negative selection” into conventional institutions, like work and education, and future offending would be valuable.

We theorize that the financial returns to crime can alter the timing of life transitions through perceptions, preferences and opportunity structures. While we measure incarceration and bachelor’s degree completion as indicators of opportunity structures, we are not able to directly measure perceptions and preferences. Future work could ideally study how perceptions and preferences differ both across individuals and within individuals over time and across contexts (see Hoeben & Thomas, 2019). Prior research suggests that changes in offending over the life course are associated with changes in expectations and preferences regarding the rewards and costs of crime (Thomas & Vogel, 2019). To be sure, the threat of formal sanctions might weigh heavily on some individuals’ decision to engage in crime; however, for others, the weight they give to the possibility of sanctions might vary across different life stages or by prior justice system experience.

Another important point is whether illegal earnings in adolescence is capturing general offending in adolescence or something unique. Bivariate analyses indicate that illegal earnings overlap with expressive crimes such as gun carrying and assault, but not entirely. It is difficult to isolate offending types at age 14 to 17 in the same structure as illegal earnings with the NLSY97. Recall, illegal earnings questions are asked relative to specific calendar years. However, after the baseline interview, offending questions are asked relative to the date of last interview. To capture offending at age 14 to 17, we would need to ensure that respondents are older than 14 but younger than 17 between interviews, which would induce measurement error due to the timing (and non-interviews) of annual interviews. We encourage scholars to examine how early participation in different types of crime is differentially associated with short- and long-term outcomes.

Although we approached our study within a choice framework, there are other theoretical explanations that may also account for relationships between illegal earnings and life course transitions, most notably social learning theory, whereby illegal earnings are rewarding and function as a reinforcement of behavior (Akers, 1998). Choice theories have connections to learning theories in that experiences can shape expectations of incentives and costs (Anwar & Loughran, 2011) or people can learn from others’ experiences (Stafford & Warr, 1993). Unfortunately, we are not able to disentangle to precise mechanisms. We encourage scholars to continue to study the implications of illegal earnings through a variety of theoretical lenses.

Studying the responses to sanctions has long been of key interest to criminologists. Comparatively, little has been done to investigate how rewards relate to offending, which is unfortunate because rewards may offer potential policy considerations. The limited research on the relationship between incarceration and illegal earnings shows that correctional institutions can serve as schools for crime. Nguyen et al. (2017) examined the illegal earnings of serious adolescent and young adult offenders before and after a period of incarceration and found an increase in illegal earnings post-incarceration (see also Hutcherson, 2012). Given that offenders are sensitive to and motivated by monetary rewards from crime (Paternoster & Bushway, 2009; Pezzin, 1995), incarceration may delay desistance by concurrently constraining legitimate opportunities and increasing the monetary rewards to offending. The results of the current study suggest that early experiences with illegal earnings can hasten the crime school process.

Another consideration for policy is the nature of illegal earnings. Illegal earnings is different from legal earnings—it is often cash based, anonymous and immediate (Nguyen et al., 2023). These qualities motivate instrumental offenses such as drug dealing and fuel conspicuous consumption driven by risky lifestyles and the liquidity of cash (Felson et al., 2019). The availability of cash on the streets facilitates illegal transactions. Policies that aim to reduce cash with digital transaction systems in banking and the welfare distribution systems appear promising and warrant more consideration (see Wright et al., 2017). For example, Pridemore et al. (2018) compared robbery rates across 67 countries and found that countries with higher rates of electronic systems of payment have lower rates of robbery. More research is needed to understand the crime prevention efforts of adopting digital payment methods.

In conclusion, we contribute to the study of choice and structure by examining whether early life choices are associated with later life opportunity structures and the timing of life transitions. We find some evidence that illegal earnings can constrain or open opportunity structures via incarceration and/or bachelor’s degree completion. It is likely that opportunity structures dynamically shape choices sets, which are “the set of discrete alternatives considered by an individual in the decision-making process” (Pagliara & Timmermans, 2004, p. 181). For example, incarceration can be related to both real and perceived opportunities in the labor market and in turn impact labor market outcomes. We believe a fruitful avenue to study choice and structure is examining the dynamic processes associated with opportunity structures, changing choice sets and choices. Such lines of inquiry are both theoretically and practically appealing.

Footnotes

Acknowledgements

We would like to thank Tim Barnum, Jean-Louis van Gelder, Chae Jaynes, the participants at the workshop and the anonymous reviewers for their thoughtful comments on earlier versions of this manuscript.

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.

ORCID iD

Holly Nguyen

Notes

Author Biographies

Holly Nguyen is Associate Professor of Sociology, Criminology, and Public Policy at the Pennsylvania State University. Her interests include employment and crime, rewards to offending and illicit drug markets.

Brandy R. Parker is a Ph.D. candidate in the Department of Sociology and Criminology at The Pennsylvania State University. Her current work broadly focuses on human capital and crime. Her publications have appeared in Journal of Quantitative Criminology, Justice Quarterly, and International Journal of Drug Policy.

Bill McCarthy is Dean at the School of Criminal Justice at Rutgers University, Newark. His early research focused on adolescent offending and relationships with parents, fictive family, peers, and romantic partners. Some of his more recent research examines overt discrimination and juvenile crime, while other projects use a labor perspective to examine sex work.

References

Akers

R. L.

(1998). Social learning theory and social structure: A general theory of crime and deviance. Northeastern University Press.

Allison

(2004). Event history analysis. In M. Hardy & A. Bryman (Eds.), Handbook of data analysis (pp. 369–386).

Anderson

(2000). Code of the street: Decency, violence, and the moral life of the inner city. W. W. Norton & Company Inc.

Anwar

Loughran

T. A.

(2011). Testing a Bayesian learning theory of deterrence among serious juvenile offenders. Criminology, 49(3), 667–98.

Apel

(2016). The effects of jail and prison confinement on cohabitation and marriage. The ANNALS of the American Academy of Political and Social Science, 665(1), 103–126.

Apel

(2022). Sanctions, perceptions, and crime. Annual Review of Criminology, 5, 205–227.

Apel

Bushway

Brame

Haviland

A. M.

Nagin

D. S.

Paternoster

(2007). Unpacking the relationship between adolescent employment and antisocial behavior: A matched samples comparison. Criminology, 45(1), 67–97.

Austin

P. C.

(2010). Absolute risk reductions and numbers needed to treat can be obtained from adjusted survival models for time-to-event outcomes. Journal of Clinical Epidemiology, 63(1), 46–55.

Austin

P. C.

(2014). The use of propensity score methods with survival or time-to-event outcomes: reporting measures of effect similar to those used in randomized experiments. Statistics in Medicine, 33(7), 1242–1258.

10.

Bacak

Kennedy

E. H.

(2015). Marginal structural models: An application to incarceration and marriage during young adulthood. Journal of Marriage and Family, 77(1), 112–125.

11.

Baker

D. P.

(2014). The school society: The educational transformation of global culture. Stanford University Press.

12.

Barnum

T. C.

Pogarsky

(2022). Situational peer dynamics and crime decisions. Journal of Research in Crime and Delinquency, 59(5), 535–573.

13.

Barnum

T. C.

Solomon

S. J.

(2019). Fight of flight: Integral emotions and violent intentions. Criminology, 57(4), 659–686.

14.

Becker

G. S.

(1968). Crime and punishment: An economic approach. Journal of Political Economy, 72(2), 169–217.

15.

Becker

G. S.

(1981). Altruism in the family and selfishness in the market place. Economica, 48(189), 1–15.

16.

Bird

C. E.

Rieker

P. P.

(2008). Gender and health: The effects of constrained choices and social policies. Cambridge University Press.

17.

Bouchard

Nguyen

(2010). Is it who you know, or how many that counts? Criminal networks and cost avoidance in a sample of young offenders. Justice Quarterly, 27(1), 130–158.

18.

Brand

J. E.

Xie

(2010). Who benefits most from college?: Evidence of negative selection in heterogeneous economic returns to higher education. American Sociological Review, 75(2), 273–302.

19.

Bushway

S. D.

(1998). The impact of an arrest on the job stability of young white American men. Journal of Research in Crime and Delinquency, 35(4), 454–479.

20.

Clarkberg

(1999). The price of partnering: The role of economic well-being in young adults’ first union experiences. Social Forces, 77(3), 945–968.

21.

Cleves

Gould

W. W.

Marchenko

Y. V.

(2008). An introduction to survival analysis using Stata. Stata Press.

22.

Cloward

R. A.

Ohlin

L. E.

(1960). Delinquency and opportunity: A theory of delinquent gangs. Free Press.

23.

Coleman

J. S.

(1990). Commentary: Social institutions and social theory. American Sociological Review, 55(3), 333–339.

24.

Draca

Machin

(2015). Crime and economic incentives. Annual Review of Economics, 7, 389–408.

25.

Eastwick

P. W.

Luchies

L. B.

Finkel

E. J.

Hunt

L.L.

(2014). The predictive validity of ideal partner preferences: A review and meta-analysis Psychological Bulletin, 140(3), 623–665.

26.

Ehrlich

(1973). Participation in illegitimate activities: A theoretical and empirical investigation. Journal of Political Economy, 81(3), 521–565.

27.

Elder

G. H.

Jr. (1998). The life course as developmental theory. Child Development, 69(1), 1–12.

28.

Elder

G. H.

Jr Johnson

Crosnoe

(2003). The emergence and development of life course theory. In: Mortimer

J. T.

Shanahan

M. J.

, (Eds.), Handbook of the life course (pp. 3–19). Sage.

29.

Fagan

Freeman

R. B.

(1999). Crime and work. Crime and Justice, 25, 225–290.

30.

Felson

Osgood

D. W.

Cundiff

P. R.

Wiernik

(2019). Life in the fast lane: Drugs, hedonistic lifestyles, and economic crime. Crime and Delinquency 65(9), 1292–1318.

31.

Ferraro

K. F.

Shippee

T. P.

Schafer

M. H.

(2009). Cumulative inequality theory for research on aging and the life course. In Bengston

V. L.

Gans

Pulney

N. M.

Silverstein

(Eds.), Handbook of theories of aging (pp. 413–433). Springer Publishing Company.

32.

Freeman

R. B.

(1996). Why do so many young American men commit crimes and what might we do about it? Journal of Economic Perspectives, 10(1), 25–42.

33.

Granovetter

(1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91(3), 481–510.

34.

Hagan

(1991). Destiny and drift: Subcultural preferences, status attainments, and the risks and rewards of youth. American Sociological Review, 56(5), 567–582.

35.

Heckman

J. J.

(1981). Heterogeneity and state dependence. In Rosen

(Ed.), Studies in labor markets (pp. 91–140). University of Chicago Press.

36.

Hoeben

E. M.

Thomas

K. J.

(2019). Peers and offender decision-making. Criminology & Public Policy, 18(4), 759–784.

37.

Holzer

H. J.

Lalonde

R. J.

(1999). Job change and job stability among less-skilled young workers (Discussion Paper No. 1191-99). Institute for Research on Poverty. https://www.irp.wisc.edu/publications/dps/pdfs/dp119199.pdf

38.

Horowitz

Graf

Livingston

(2019). Marriage and cohabitation in the U.S. Pew Research Center. https://www.pewresearch.org/social-trends/2019/11/06/marriage-and-cohabitation-in-the-u-s

39.

Huang

P. M.

Smock

P. J.

Manning

W. D.

Bergstrom-Lynch

C. A.

(2011). He says, she says: Gender and cohabitation. Journal of Family Issues, 32(7), 876–905.

40.

Huebner

(2007). Racial and ethnic differences in the likelihood of marriage: The effect of incarceration. Justice Quarterly, 24(1), 156–183.

41.

Hutcherson

D. T.

(2012). Crime pays: The connection between time in prison and future criminal earnings. The Prison Journal, 92(3), 315–335.

42.

Jessor

Donovan

J. E.

Costa

F. M.

(1994). Beyond adolescence: Problem behaviour and young adult development. Cambridge University Press.

43.

Kang

(2019). The transition to adulthood of contemporary delinquent adolescents. Journal of Developmental and Life Course Criminology, 5(2), 176–202.

44.

King

Nielson

(2019). Why propensity scores should not be used for matching. Political Analysis, 27(4), 435–454.

45.

LeBel

T. P.

Burnett

Maruna

Bushway

(2008). The chicken and egg of subjective and social factors in desistance from crime. European Journal of Criminology, 5(2), 131–159.

46.

Linden

Mathur

M. B.

VanderWeele

T. J.

(2020). Conducting sensitivity analysis for unmeasured confounding in observational studies using E-values: The evalue package. The Stata Journal: Promoting Communications on Statistics and Stata, 20(1), 162–175.

47.

Lochner

(2004). Education, work, and crime: A human capital approach. International Economic Review, 45(3), 811–843.

48.

Lochner

(2007). Individual perceptions of the criminal justice system. American Economic Review, 97(1), 444–460.

49.

Loewenstein

Furstenberg

(1991). Is teenage sexual behavior rational? Journal of Applied Social Psychology, 21(12), 957–986.

50.

Lonardo

R. A.

Manning

W. D.

Giordano

P. C.

Longmore

M. A.

(2010). Offending, substance use, and cohabitation in young adulthood. Sociological Forum, 25(4), 787–803.

51.

Loughran

T. A.

Nguyen

Piquero

A. R.

Fagan

(2013). The returns to criminal capital. American Sociological Review, 78(6), 925–948.

52.

Machin

Meghir

(2004). Crime and economic incentives. The Journal of Human Resources, 39(4), 958–979.

53.

Massoglia

Remster

King

R. D.

(2011). Stigma or separation? Understanding the incarceration-divorce relationship. Social Forces, 90(1), 133–155.

54.

Massoglia

Uggen

(2010). Settling down and aging out: Toward an interactionist theory of desistance and the transition to adulthood. American Journal of Sociology, 116(2), 543–582.

55.

Matsueda

R. L.

O’Neill

K. K.

Kreager

D. A.

(2020). Embeddedness, reflected appraisals, and deterrence: A symbolic interactionist theory of adolescent theft. In Serpe

R. T.

Stryker

Powell

(Eds.). Identity and symbolic interaction (pp. 239–271). Springer Publishing Company.

56.

McCarthy

(2002). New economics of sociological criminology. Annual Review of Sociology, 28, 417–442.

57.

McCarthy

Hagan

(2001). When crime pays: Capital, competence, and criminal success. Social Forces, 79(3), 1035–1060.

58.

McClendon

Kuo

J. C.-L.

Raley

(2014). Opportunities to meet: Occupational education and marriage formation in young adulthood. Demography, 51(4), 1319–1344.

59.

McGloin

J. M.

Thomas

K. J.

(2016). Incentives for collective deviance: Group size and changes in perceived risk, cost, and reward. Criminology, 54(3), 459–486.

60.

McRae

(2003). Constraints and choices in mothers’ employment careers: A consideration of Hakim’s preferences. The British Journal of Sociology, 54(3), 317–338.

61.

Mocan

H. N.

Billups

S. C.

Overland

(2005). A dynamic model of differential human capital and criminal activity. Economica, 72(288), 655–681.

62.

Morselli

Tremblay

McCarthy

(2006). Mentors and criminal achievement. Criminology, 44(1), 17–43.

63.

Mortimer

J. T.

Shanahan

M. J.

(Eds.). (2007). Handbook of the life course. Springer Science & Business Media.

64.

Nagin

D. S.

(2013). Deterrence in the twenty-first century. Crime and Justice, 42(1), 199–263.

65.

Nagin

D. S.

Paternoster

(1991). On the relationship of past to future participation in delinquency. Criminology, 29(2), 163–189.

66.

Nagin

Paternoster

(2000). Population heterogeneity and state dependence: State of the evidence and directions for future research. Journal of Quantitative Criminology, 16(2), 117–144.

67.

Nguyen

(2020). On the conceptualization of criminal capital. Journal of Research in Crime and Delinquency, 57(2), 182–216.

68.

Nguyen

Loughran

(2018). On the Measurement and Identification of Turning Points in Criminology. Annual Review of Criminology, 1, 335–358.

69.

Nguyen

Loughran

T. A.

Morselli

Ouellet

(2021). Offending frequency and responses to illegal incentives. Journal of Quantitative Criminology, 38(1), 239–265.

70.

Nguyen

Loughran

T. A.

Paternoster

Fagan

Piquero

(2017). Institutional placement and illegal earnings. Journal of Quantitative Criminology, 33(2), 207–235.

71.

Nguyen

Loughran

T. A.

Topalli

(2023). Crime, consumption, and choice: On the interchangeability of licit and illicit income. Journal of Research in Crime and Delinquency, 60(4), 416–454.

72.

Oppenheimer

V. K.

(1988). A theory of marriage timing. American Journal of Sociology, 94(3), 563–591.

73.

Oppenheimer

V. K.

(2003). Cohabiting and marriage during young men's career-development process. Demography, 40(1), 127–149.

74.

O’Rand

(2006). Stratification and the life course: Life course capital, life course risks, and social inequality. In Binstock

R. H.

Cutler

S. J.

Schulz

J. H.

George

L. K.

Hendricks

(Eds.), Handbook of aging and the social sciences (6th ed., pp. 145–162). Academic Press.

75.

Oskarson

(2005). Social structure and party choice. In Thomassen

(Ed.), The european voter: A comparative study of modern democracies (pp. 84–105). Oxford University Press.

76.

Pager

(2003). The mark of a criminal record. American Journal of Sociology, 108(5), 937–975.

77.

Pagliara

Timmermans

(2004). Choice set generation in spatial contexts: a review. The International Journal of Transportation Research, 1(3), 181–196.

78.

Paternoster

Bushway

(2009). Desistance and the “feared self”: Toward an identity theory of criminal desistance. The Journal of Criminal Law and Criminology, 99(4), 1103–1156.

79.

Paternoster

Simpson

(1996). Sanction threats and appeals to morality: Testing a rational choice model of corporate crime. Law and Society Review, 30(3), 549–584.

80.

Pezzin

L. E.

(1995). Earning prospects, matching effects, and the decision to terminate a criminal career. Journal of Quantitative Criminology, 11(1), 29–50.

81.

Piquero

Tibbetts

(1996). Specifying the direct and indirect effects of low self-control and situational factors on offenders’ decision making: Toward a more complete model of rational offending. Justice Quarterly, 13(3), 481–510.

82.

Pridemore

W. A.

Roche

S. P.

Rogers

M. L.

(2018). Cashlessness and street crime: A cross-national study of direct deposit payment and robbery rates. Justice Quarterly, 35(5), 919–939.

83.

Ripollone

J. E.

Huybrechts

K. F.

Rothman

K. J.

Ferguson

R. E.

Franklin

J. M.

(2018). Implications of the propensity score matching paradox in pharmacoepidemiology. American Journal of Epidemiology, 187(9), 1951–1961.

84.

Rosenbaum

P. R.

(2002). Over bias in observational studies. In Rosenbaum

P. R.

(Ed.), Observational studies (2nd ed., pp. 71–104). Springer Publications.

85.

Rosenbaum

P. R.

Rubin

D. B.

(1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41–55.

86.

Sampson

R. J.

Laub

J. H.

(1993). Crime in the making: Pathways and turning points through life. Harvard University Press.

87.

Sampson

R. J.

Laub

J. H.

(1997). A life-course theory of cumulative disadvantage and the stability of delinquency. In Thornberry

T. P.

(Ed.). Advances in criminological theory ( Vol.7, pp. 133–161). Transaction Publishers.

88.

Shover

Thompson

C. Y.

(1992). Age, differential expectations, and crime desistance. Criminology, 30(1), 89–104.

89.

Siennick

S. E.

Staff

(2008). Explaining the educational deficits of delinquent youths. Criminology, 46(3), 609–635.

90.

Siennick

S. E.

Staff

Osgood

D. W.

Schulenberg

J. E.

Bachman

J. G.

Van Eseltine

(2014). Partnership transitions and antisocial behavior in young adulthood: A within-person, multi-cohort analysis. Journal of Research in Crime and Delinquency, 51(6), 735–758.

91.

Smock

P. J.

(2004). The wax and wane of marriage: Prospects for marriage in the 21st century. Journal of Marriage and the Family, 66(4), 966–973.

92.

Stafford

M. C.

Warr

(1993). A reconceptualization of general and specific deterrence. Journal of Research in Crime and Delinquency, 30(2), 123–135.

93.

Statista Research Department. (2023, June 2). Educational attainment in the U.S. https://www.statista.com/statistics/184260/educational-attainment-in-the-us/

94.

Sugie

N. F.

Zatz

N. D.

Augustine

(2020). Employer aversion to criminal records: An experimental study of mechanisms. Criminology, 58(1), 5–34.

95.

Thomas

K. J.

Baumer

E. P.

Loughran

T. A.

(2022). Structural predictors of choice: Testing a multilevel rational choice theory of crime. Criminology, 60(4), 606–636.

96.

Thomas

K. J.

Vogel

(2019). Testing a rational choice model of “desistance:” Decomposing changing expectations and changing utilities. Criminology, 57(4), 687–714.

97.

Tremblay

. (1993). Searching for suitable co-offenders. In Clarke

R. V.

Felson

(Eds.), Routine activity and rational choice: Advances in criminological theory (pp. 17–36). Transaction.

98.

Turney

(2015). Liminal men: Incarceration and relationship dissolution. Social Problems, 62(4), 499–528.

99.

Uggen

Vuolo

Lageson

Ruhland

Whitham

H. K.

(2014). The edge of stigma: An experimental audit of the effects of low-level criminal records on employment. Criminology, 52(4), 627–654.

100.

U.S. Bureau of Labor Statistics. (2022, May). Education Pays, 2021. U.S. Bureau of Labor Statistics, Department of Labor. https://www.bls.gov/careeroutlook/2022/data-on-display/education-pays.htm

101.

VanderWeele

T. J.

Ding

(2017). Sensitivity analysis in observational research: Introducing the E-value. Annals of Internal Medicine, 167, 268–274.

102.

van Gelder

J. L.

de Vries

R. E

. (2012). Traits and states: Integrating personality and affect into a model of criminal decision making. Criminology, 50(3), 637–671.

103.

van Gelder

J. L.

Elffers

Reynald

Nagin

D. S

. (Eds.). (2013). Affect and cognition in criminal decision making. Routledge.

104.

Vuolo

Kadowaki

Kelly

B. C.

(2016). A multilevel test of constrained choices theory: The case of tobacco clean air restrictions. Journal of Health and Social Behavior, 57(3), 351–372.

105.

Wright

Tekin

Topalli

McClellan

Dickinson

Rosenfeld

(2017). Less cash, less crime: Evidence from the electronic benefit transfer program. The Journal of Law and Economics, 60(2), 361–383.

106.

Xie

Raymo

J. M.

Goyette

Thornton

(2003). Economic potential and entry into marriage and cohabitation. Demography, 40(2), 351–67.

107.

Western

Pettit

(2000). Incarceration and racial inequality in men’s employment. Industrial & Labor Relations Review, 54(1), 3–16.

108.

Widdowson

A. O.

Siennick

S. E

Hay

(2016). The implications of arrest for college enrollment: An analysis of long-term effects and mediating mechanisms. Criminology, 54(4), 621–652.