The Association Between Perceptions of Neighborhood Conditions and the Employment of Emerging Adults Formerly Involved in the Juvenile Justice System

Abstract

Although prior research finds that poor neighborhood conditions are negatively associated with employment, little study has focused on emerging adults who formerly had contact with the juvenile justice system and are frequently engaged in informal job markets. Using a hybrid model and three waves from panel data with formerly juvenile justice-involved emerging adults (18–24 at T₁, 20–25 at T₂, 20–26 at T₃) collected in Philadelphia and Phoenix (N = 947), we find an inter-individual increase in the degree of disorder within the neighborhood decreases weeks worked for community jobs (between-effect), whereas an intra-individual increase in neighborhood condition scores increases weeks worked for under-the-table jobs among emerging adults (within-effect). Some time-variant and invariant sociodemographic factors, including perceived opportunity for work, mental health, substance use, gang membership, race, and education, are significantly related to employment. Our findings reiterate justice-involved young people may have difficulty sustaining formal employment partially due to neighborhood conditions.

Keywords

neighborhood conditions employment formerly incarcerated emerging adults

Introduction

Employment access has been known as one of the most critical factors directly linked to individuals’ lives since it is essential in reducing poverty and supporting sustainable economic and social development (United Nations, 2005). Over the 10 years (2010–2019) before the COVID-19 outbreak, the average U.S. unemployment rate decreased from 9.6% to 3.7% (Bureau of Labor Statistics [BLS], 2020b). Beginning in February 2020, the outbreak unprecedentedly affected the U.S. labor market, where the unemployment rates spiked, causing millions to lose their jobs (BLS, 2020a; Pew Research Center [Pew], 2021). A year after the biggest job loss in U.S. history, many jobs have returned, but the labor market is still not fully recovered to pre-pandemic levels; job losses were disproportionately concentrated among the low-earning workforce (BLS, 2021).

Understanding the employment of vulnerable populations such as formerly justice-involved people, however, has been historically challenging. According to Couloute and Kopf (2018) who use data from the National Former Prisoner Survey, the unemployment rate among formerly justice-involved people is over 27%-the highest total unemployment rate in U.S. history. Former justice-involved persons need stable jobs for the same reasons as the general population since it provides many benefits, such as establishing routine activities, financially supporting families, and having better mental/physical health outcomes (Laub & Sampson, 2003). This ultimately strengthens their communities as well (Couloute & Kopf, 2018).

When explaining the employment difficulties of people from disadvantaged backgrounds, previous studies have widely focused on individual-level factors such as human capital, education, or mental health (Turney et al., 2006). Given that the unemployment rate (before the pandemic) was 10% in disadvantaged communities, which was much higher than the overall nationwide rate (“Poor Neighborhoods,” 2018), and the cycling of justice-involved people to disadvantaged communities seems undeniable (Rose & Clear, 1998), examining the effect of the perception of neighborhood conditions on the participation in the labor market by such populations needs to be further studied. Unemployment itself is not the only problem, in that residents living in neighborhoods with high social and physical disorders face fewer opportunities to work, with lower pay and qualifications (Turney et al., 2006). Justice-involved young people living in disadvantaged communities, particularly, often experience informal work and struggle to find secure jobs (Sheppard & Ricciardelli, 2020). This phenomenon is due to the fact that such communities often have different ecological and economic characteristics from advantaged neighborhoods where it is often observed, such as sizable residential density, less crimes, high levels of collective efficacy and resilience, etc. (Dupere et al., 2010). It is increasingly recognized that access to employment opportunities is not merely attributed to individual characteristics, but also attributed to adverse neighborhood conditions. This study, therefore, aims to highlight the role of residential context in employment among emerging adults with a history of involvement in the juvenile justice system.

Employment Difficulties: Explaining Social Isolation and Disorganization

The aforementioned characteristics are closely related to a concept of social isolation (Wilson, 1987), which is aligned with massive structural changes transforming the social, demographic, and economic composition of urban neighborhoods that the U.S. experienced in the late 1960s and 1970s (Casciano & Massey, 2008). Wilson (1987) defines social isolation as the “lack of contact or of sustained interactions with individuals and institutions that represent mainstream society” (p. 60). That is, residents of such isolated neighborhoods may not have the opportunity to develop mainstream norms and values as well as building social capital that could affect labor market outcomes (Casciano & Massey, 2008). Public and private sources selectively concentrated on suburban areas where Whites and some middle class Blacks migrated and ignored urban disadvantaged neighborhoods (Casciano & Massey, 2008). This has two major implications: First, the lack of businesses in such neighborhoods was associated with a shortage of jobs within the community; second, a lack of social capital negatively affected the capabilities of the local community organizations providing residents with basic employable skills and services (Casciano & Massey, 2008; Wilson, 1987). Inevitably, those living in socially and economically isolated neighborhoods, far from opportunities that promote employment (Turney et al., 2006), are disempowered from improving their quality of community life.

The concept of social isolation is intertwined with social disorganization theory (SDT), premising that a malfunction of effective social control due to a lack of social integration between members or groups characterizes a state of community (Sampson, 2012). The neighborhood, that is, socially isolated, and hence disorganized, has an inability to “realize the common values of its residents and maintain effective social controls” (Sampson, 2012, p. 37). Thus, the effect of neighborhood on employment can be explained well through SDT (Sampson & Raudenbush, 1999; Shaw & McKay, 2014). Shaw and McKay (2014) specifically theorized that neighborhood poverty, conflicting cultural values and practices, and low mobility were linked to higher levels of social disorganization, and ultimately greater risks for delinquency (Brenner et al., 2011; Lei & Beach, 2020; Madyun, 2011).

The theory, however, did not only explain delinquency outcomes, but also extended the associations between neighborhood conditions and socioeconomic status (SES), academic, and occupational achievements. For example, Tanner et al. (1999) point to a negative association between delinquency and long-term educational and occupational attainment. Among those achievements, employment, which is a key measure of SES along with income and education (American Psychological Association, 2015), deserves specific attention since social disorganization refers to a failure of maintaining social institutions, causing ineffectiveness of social or informal social control mechanisms (Kirk, 2019; Shaw & McKay, 2014). In this sense, the key premise of SDT is in line with the neighborhood effects hypothesis, as Wilson (1987) argued that social and economic conditions within neighborhoods influence a various range of socioeconomic prospects (e.g., employment prospect) and individuals’ outcomes (e.g., mental/physical health, academic achievement). Given the impact of concentrated socioeconomic disadvantage, it could be theoretically assumed that there would be higher levels of unemployment in neighborhoods with high levels of physical disorder (e.g., vacant or abandoned housing, vandalized and run-down buildings, abandoned cars, graffiti, and litter in the streets) and social disorder (e.g., drinking or taking drugs on the streets, drug dealing, hostile arguing, conflict and fighting, and high levels of police activity; Garcia, 2014). It is also possible that justice-involved individuals living in disadvantaged communities earn their livelihoods in informal work (Sheppard & Ricciardelli, 2020).

Neighborhood and Employment

Several studies attempted to investigate the association between neighborhood and employment (Alvarado, 2018; Casciano & Massey, 2008; Galster et al., 2015; Kling et al., 2007; Ludwig et al., 2013; Turney et al., 2006). A longitudinal study using data from the Fragile Families and Child Well-Being Study, Casciano and Massey (2008) found neighborhoods are related to employment, in that the likelihood of being employed among study participants living in poor neighborhoods was low, which was consistent with another longitudinal study from Alvarado (2018) using panel data from the NLSY79 and the NLSY Children and Young Adults cohorts. Galster et al. (2015) also evidenced that residing in a more socially disadvantaged neighborhood was associated with low employment prospects for young people, particularly African-Americans. Another study using data from families who participated in the Moving to Opportunity (MTO) for the fair housing demonstration program in Baltimore, Maryland with a mixed method revealed some significant employment barriers among residents residing in disadvantaged communities, including human capital issues (i.e., low education levels) and mental health problems (Turney et al., 2006). Given those findings, neighborhood conditions may intersect with key demographic factors such as race and education in being employed.

In contrast to these, some previous experimental studies provided inconsistent results. For example, a study using data from a randomized experiment examining neighborhood effects on adult economic self-sufficiency (e.g., job accessibility) from five states (California, Illinois, Maryland, Massachusetts, and New York) found no significant statistical difference between experimental groups randomly assigned in MTO who were offered vouchers lived in safer neighborhoods that had lower poverty rates than those of the control group not offered vouchers (Kling et al., 2007). The main reasons of no treatment effects in this study were aligned with what Turney et al. (2006) found, in that transportation difficulties and disrupted social networks were critical barriers to employment in the experimental group (Kling et al., 2007). Consistent with this, Ludwig et al. (2013) used the same, but longer datasets (10–15 years after baseline) than Kling et al. (2007) who used 4 to 7 years after random assignment found no detectable effect of neighborhood on economic outcomes. A gap found in some previous studies is that the employment outcome variables were measured by dichotomous employment status or number of hours worked weekly, indicating that it does not explicitly identify whether neighborhood conditions are associated with jobs in the formal sector or informal sector.

If focused on informal work as an outcome, disadvantaged neighborhood conditions could be positively associated with that type of employment. Some studies point out that socioeconomically disadvantaged neighborhoods, particularly in urban areas, have higher proportions of residents who participated in informal work (Curran, 2004; Hebert-Beirne et al., 2021), which is also aligned with the aforementioned theoretical accounts.

Other Factors and Employment

A body of literature has suggested sociodemographic factors that are associated with employment, including perceived employability (Consiglio et al., 2021), mental health (Aarons-Mele, 2020; Centers for Disease Control and Prevention, 2018; Olesen et al., 2013; Tolman et al., 2009), substance use (Okechukwu et al., 2019; Wu et al., 2003), gang membership (Gilman et al., 2014), race (BLS, 2014; Holzer, 2021), and education (BLS, 2016). These sociodemographic factors need to be controlled for minimizing the impact of confounding in the association between perceived neighborhood conditions and employment.

Current Study

Although there is a robust body of literature that provides supporting evidence for the structural influence on employment, there are still gaps, and the current study will contribute to the existing literature on neighborhood effects: Few studies have focused on emerging adults who formerly had contact with the juvenile justice system in the association between neighborhood conditions and employment; given that such young people are often unable to secure stable jobs, and frequently engage in informal job markets (Holzer et al., 2003; Sorensen & Oliver, 2002), known as under-the-table jobs, it is important to examine possible variations in both formal and informal jobs with neighborhood effects; we distinguish the within- and between-group changes of the perceived neighborhood conditions-work relationship to show that the relationship’s direction and statistical significance differ depending on the types of employment.

Using longitudinal panel data from the Pathway to Desistance study (N = 947), we seek to add to the literature on studies of residential context and to explore variations across different types of employment (e.g., community jobs as regular jobs earning at least minimum wage vs. under-the-table jobs), consequently expanding what is known about the neighborhood conditions among formerly justice-involved emerging adults. Regarding the neighborhood condition variable, Elo et al. (2009) suggest that perceptions of neighborhood conditions from survey respondents is a fair alternative to the measure of neighborhood conditions offered by administrative data sources such as census data. Thus, we aim to address our research questions, including (1) to what extent are perceived neighborhood conditions associated with employment among emerging adults involved in the juvenile justice system, after controlling for sociodemographic factors?; (2) do the patterns of different types of employment vary with perceived neighborhood conditions?

Based on theories and previous studies, we hypothesize that (1) emerging adults’ higher perceptions of worsening neighborhood conditions decrease weeks worked for community jobs over time; (2) emerging adults’ higher perceptions of worsening neighborhood conditions increase weeks worked for under-the-table jobs over time. Findings from this study can set a foundation for community-based interventions to address the needs of justice-involved emerging adults who live in poorly perceived neighborhoods.

Methods

Study Participants

The current study utilizes data from the Pathways to Desistance study (PDS; https://www.pathwaysstudy.pitt.edu/index.html), a longitudinal project focused on 1,354 juvenile offenders transitioning from adolescence to adulthood. The vast majority of offenses were felonies, but some were serious non-felony offenses, including misdemeanors, sexual, or weapon charges. The study participants completed baseline interviews from juvenile or adult courts located in Phoenix, Arizona (n = 654) and Philadelphia, Pennsylvania (n = 700) from 2000 to 2003, where one-third of adjudicated youth in each jurisdiction were enrolled. The participants had follow-up interviews at 6, 12, 18, 24, 30, 36, 48, 60, 72, and 84 months past baseline. The last follow-up interview was completed in 2010. The majority of participants were males (86.4%). Participants’ racial identity included primarily non-Hispanic Blacks (41%), followed by Hispanics (34%), non-Hispanic Whites (20%), and other racial groups (5%).

This study used data from the baseline and three follow-up interviews. Since our interest group is emerging adults, we utilized data drawn from 60 (T₁), 72 (T₂), and 84 month (T₃) follow-ups when most youth would have become emerging adults (≤18). The age of the participants ranged from 14 to 19 at baseline, 18 to 24 at T₁, 20 to 25 at T₂, and 20 to 26 at T₃. All PDS data are publicly available from the Interuniversity Consortium for Political and Social Research and were exempt from the university IRB review.

Measures

Dependent variables

Employment was defined as total weeks worked in the past year (e.g., 0–52 weeks), which was a continuous variable. We operationalized total weeks worked into three variables: (1) all community and under-the-table jobs, (2) community-only jobs, and (3) under-the-table only jobs, which was used in Lee and Kim (2022).

Time-variant independent variables

Perceived neighborhood conditions as a proxy measure of neighborhood conditions, a main independent variable, was measured from the Neighborhood Conditions Measure assessing the environment surrounding the adolescent’s home, including physical and social disorders, which was adapted from Sampson and Raudenbush (1999). On a question of how often each of the following occurs within your neighborhood, the study participants self-reported to the scale containing 21 items (e.g., cigarettes on the street, empty beer bottles on the streets or sidewalks, gang graffiti, needles, or syringes for physical disorder; people drunk or passed out, prostitutes on the streets, and people smoking crack for social disorder) to which participants responded on a 4-point Likert scale ranging from 1 = never occurs to 4 = often occurs. Higher scores represent a greater degree of physical and social disorders within the community. The scale was found to have excellent internal consistency at baseline and the follow-up time points (Cronbach’s α = .94–.96).

Control variables

Time-variant control variables

The time-variant covariates included in this analysis were perceived opportunity for work, mental health symptoms, substance use, and gang involvement. Perceived opportunity for work was assessed by calculating the mean score of 5-items (e.g., “employers around here often hire young people from this neighborhood”), each measured by a 5-point Likert scale ranging from 1 = strongly disagree to 5 = strongly agree (α = .76 at baseline). Mental health symptoms were assessed across two domains, depression and anxiety. We used mean scores of depression (α = .81) and anxiety (α = .78) measured from the Brief Symptom Inventory (BSI; Derogatis & Melisaratos, 1983) with a 5-level Likert-type scale anchored by 0 = not at all to 4 = extremely, with higher scores indicative of increased symptoms. For substance use, three types of substances, including alcohol, marijuana, and hard drugs adapted from the University of Missouri’s Alcohol and Health Study (Chassin et al., 1991) were included. The study participants responded to a 9-level Likert-type scale (0 = not at all to 8 = every day) on questions of how many times you have been drunk; used marijuana or hashish; and used illegal hard drugs (e.g., sedatives, amphetamines, cocaine, opiates) in the past year. Hard drug use was computed by the average number of times of drug use (range: 0–8). Reliability for drug use showed mostly acceptable alpha levels (α = .75, mean inter-item r = .38 for T₁; α = .78, r = .38 for T₃), except for T₂ (α = .55, r = .21). We converted them into binary variables, in that 0 (i.e., “not at all”) was set to 0 and values greater than or equal to 1 (i.e., “1–5 times”) were set to 1 for a meaningful interpretation (i.e., 0 = did not alcohol use, 1 = alcohol use). Lastly, gang involvement in the past 12 months was dichotomized to two categories, 0 = no, 1 = yes.

Time-invariant control variables

The time-invariant characteristics included in this model were racial identity (1 = Whites, 2 = Blacks, 3 = Hispanics, and 4 = other), age at T₁, gender, parental SES, and education. For race, we recoded this into a set of dummy variables—three dummy variables (i.e., k − 1) including Black, Hispanic, and other were created and included in our model to compare with White, which was a reference group. Gender was included as a dummy variable indicating female, 0 = male, 1 = female. Parental SES was computed with Hollingshead’s (1975) 4-factor index of social status, which was based on parents’ education and current occupation. Finally, education was classified into two categories, 0 = no, 1 = yes, measuring enrollment in school either currently (if not detained) or enrolled before coming to the facility.

Data Analyses

For the first step, we conducted descriptive statistics and bivariate analyses such as a repeated measures Analysis of variance (ANOVA) and Cochran’s Q test. These tests were conducted separately on time-variant variables to examine whether the means or proportions exhibit significant change over time. Then, we conducted Allison’s (2009) hybrid approach that combines fixed- and random-effects models. Although fixed-effects models offer the advantage of minimizing omitted variable bias, by controlling for unobserved time-invariant variables, coefficients for observed time-invariant variables (i.e., variables that do not vary over time) cannot be estimated. To avoid such limitations, the hybrid model, also known as between-within models, has been proposed, which allows us to estimate coefficients for time-varying and time-invariant variables, while controlling for all (i.e., measured and unmeasured) time-invariant characteristics (Williams, 2018). In other words, our analytic approach was mainly chosen to (1) control for unobserved time-invariant characteristics and (2) obtain valid estimates of the within- and between-person effects of neighborhood conditions on employment. To handle missing data, we used listwise deletion, resulting in the inclusion of 947 emerging adults out of 1,354 in the current analyses. Besides these, we tested whether the relationship between neighborhood conditions and employment could be possibly curvilinear by adding the square of neighborhood conditions in a preliminary analysis; and the neighborhood conditions-squared coefficient was revealed to be insignificant. Thus, we deleted the square of neighborhood conditions from our models and went back to a linear model (Allison, 1999). All statistical analyses were conducted with Stata 15.1.

Results

Descriptive Statistics

Table 1 displays the participants’ characteristics. On average, our study participants worked 17.92 weeks (SD = 20.22) at T₁, 20.67 weeks (SD = 20.80) at T₂, and about 20.32 weeks (SD = 21.59) at T₃ across community and under-the-table jobs. A one-way repeated measures ANOVA indicated that there was a significant effect of time on all jobs. For community jobs, they worked 15.54 weeks (SD = 20.10) at T₁, 16.89 weeks (SD = 20.20) at T₂, and about 16.22 weeks (SD = 20.80) at T₃ on average. For under-the-table jobs, the average weeks worked were about 4.76 weeks (SD = 12.58) at T₁, 5.31 weeks (SD = 12.98) at T₂, and 5.49 weeks (SD = 13.80) at T₃. Perceived neighborhood conditions had means and standard deviations (SD) of 2.29 (SD = 0.93) at T₁, 2.22 (SD = 0.83) at T₂, and 2.19 (SD = 0.79) at T₃. There was a significant effect of time on neighborhood conditions. Other descriptive statistics are reported in Table 1.

Table 1.

Descriptive Statistics.

Time-variant variables	Wave 1	Wave 2	Wave 3	Bivariate analyses
All jobs: mean (SD), range	17.92 (20.22), 0–52	20.67 (20.80), 0–52	20.32 (21.59), 0–52	F(1.93, 2,050.14) = 11.35^a,***
Community jobs: mean (SD), range	15.54 (20.10), 0–52	16.89 (20.20), 0–52	16.22 (20.80), 0–52	F(1.93, 1,024.08) = 2.00^a
Under-the-table jobs: mean (SD), range	4.76 (12.58), 0–52	5.31 (12.98), 0–52	5.49 (13.80), 0–52	F(1.96, 1,037.65) = 1.34^a
Neighborhood: mean (SD), range	2.29 (.83), 1–4	2.22 (.83), 1–4	2.19 (.79), 1–4	F(1.96, 1,499.26) = 8.36^a,***
POW: mean (SD), range	3.65 (.65), 1–5	3.65 (.70), 1–5	3.58 (.65), 1–5	F(2, 1,538) = 4.17^b,*
Depression: mean (SD), range	.37 (.61), 0–3.67	.42 (.62), 0–3.83	.36 (.62), 0–3.83	F(2, 908) = 5.46^b,**
Anxiety: mean (SD), range	.32 (.51), 0–3.50	.34 (.55), 0–3.83	.32 (.53), 0–3.50	F(2, 906) = 2.36^a,†
Alcohol: N (%)
No	506 (42.2)	452 (38.7)	426 (37.9)	χ²(2, N = 1,051) = 6.16^c,*
Yes	694 (57.8)	717 (61.3)	698 (62.1)
Marijuana: N (%)
No	757 (63.1)	730 (62.4)	739 (65.7)	χ²(2, N = 1,051) = 4.11^c
Yes	443 (36.9)	439 (37.6)	385 (34.3)
Drugs: N (%)
No	1,034 (86.2)	1,009 (86.3)	981 (87.3)	χ²(2, N = 1,051) = 2.15^c
Yes	166 (12.3)	160 (13.7)	143 (12.7)
Gang involvement: N (%)
No	1,124 (93.4)	1,104 (93.8)	1,067 (94.3)	χ²(2, N = 1,061) = 6.05^c,*
Yes	80 (6.6)	73 (6.2)	64 (5.7)
Time-invariant variables
Age: mean (SD), range at wave 1	21.02 (1.15), 18–24
Parental SES: mean (SD) at baseline	51.41 (12.30)
Education: N (%) at baseline
No	382 (28.2)
Yes	972 (71.8)

Note. POW = perceived opportunity for work.

Repeated measures ANOVA with Greenhouse-Geisser.

Repeated measures ANOVA with Sphericity assumed.

Cochran’s Q test.

†

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

Regression Results

Table 2 shows the results of the hybrid model that examined the effect of neighborhood condition on employment, while controlling for both time-variant and time-invariant covariates over time. Between-individual estimates for neighborhood conditions were found to have significant, negative associations in all jobs and in community jobs, indicating that total weeks worked in all jobs (CSE = 0.83, p < .001) and community jobs (CSE = 0.90, p < .001) decreased by 2.94 and 3.42 weeks, respectively, for each 1-unit increase in the degree of disorder within the neighborhood (between). Within-individual estimates for neighborhood conditions, reflecting fixed-effect estimates for neighborhood conditions found total weeks worked in under-the-table jobs increased by 2.03 weeks (CSE = 0.96, p < .05) for each 1-unit increase in the degree of disorder within the neighborhood (within). In contrast, total weeks worked across all jobs and total weeks worked in community jobs were found to have non-significant association with neighborhood conditions.

Table 2.

A Panel Analysis on the Association between Neighborhood Conditions and Work using Hybrid Regression Models.

	All jobs	Community-only jobs	Under-the-table only jobs
Between: neighborhood	−2.94*** (.83)	−3.42*** (.90)	1.05 (.67)
Within: neighborhood	1.15 (.93)	−.08 (.99)	2.03* (.96)
Between: POW	2.01* (.96)	2.31* (1.00)	−.76 (.74)
Within: POW	−.23 (.94)	−.21 (1.03)	1.20 (.99)
Between: depression	−2.72^† (1.60)	−1.65 (1.62)	−2.06^† (1.20)
Within: depression	−3.36** (1.23)	−3.33* (1.34)	−.18 (1.29)
Between: anxiety	.17 (1.84)	−.24 (1.88)	1.63 (1.39)
Within: anxiety	2.80* (1.29)	2.35^† (1.40)	.40 (1.36)
Between: alcohol (1 = yes)	11.65*** (1.51)	11.61*** (1.63)	.38 (1.21)
Within: alcohol (1 = yes)	2.11 (1.41)	2.05 (1.52)	−.66 (1.47)
Between: marijuana (1 = yes)	−5.52*** (1.43)	−7.54*** (1.52)	.76 (1.13)
Within: marijuana (1 = yes)	.56 (1.37)	−1.14 (1.43)	2.80* (1.39)
Between: drugs (1 = yes)	−3.87* (1.90)	−2.28 (1.97)	.01 (1.46)
Within: drugs (1 = yes)	−.31 (1.59)	−.06 (1.72)	.93 (1.67)
Between: gang (1 = yes)	−3.58 (2.79)	−1.27 (3.02)	−2.26 (2.26)
Within: gang (1 = yes)	−12.92** (4.61)	−9.98^† (5.38)	−2.32 (5.21)
Race (ref. white)
Black	−7.93*** (1.55)	−9.34*** (1.69)	.29 (1.24)
Hispanic	−3.40* (1.53)	−3.46* (1.67)	−1.20 (1.23)
Other	−1.44 (2.71)	−.99 (3.11)	−3.05 (2.29)
Age	.75 (.49)	.80 (.53)	.04 (.39)
Gender (1 = female)	−.78 (1.41)	−.82 (1.54)	−2.02^† (1.12)
Parental SES	.01 (.03)	−.02 (.03)	.04 (.03)
Education (1 = enrolled in school before detention)	2.95* (1.29)	1.84 (1.41)	1.38 (1.04)

Note. Beta coefficients were reported. Cluster standard errors were reported in parentheses. Observations with missing values were dropped (listwise deletion). POW = perceived opportunity for work.

†

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

To reduce the effects of confounding, we controlled for sociodemographic variables, guided by literature, which could be possibly associated with employment. Among them, for between-effects, perceived opportunities for work was positively associated with community jobs (β = 2.31, CSE = 1.00, p < .05). Regarding substance use, alcohol use was positively associated with community jobs (β = 11.61, CSE = 1.63, p < .001), whereas marijuana use was negatively associated with community jobs (β = −7.54, CSE = 1.52, p < .001). Drug use was negatively associated with all jobs (β = −3.87, CSE = 1.90, p < .05).

For within-effects, depression was negatively associated with community jobs (β = −3.33, CSE = 1.34, p < .05) whereas anxiety was positively associated with all jobs (β = 2.80, CSE = 1.29, p < .05). Regarding substance use, marijuana use was positively associated with under-the-table jobs (β = 2.80, CSE = 1.39, p < .05). In addition, gang involvement was negatively associated with all jobs (β = −12.92, CSE = 4.61, p < .01).

For the random-effects estimates of time-invariant predictors, both Blacks and Hispanics were significantly less likely to work in community jobs (β_Blacks = −9.34, CSE = 1.69, p < .001; β_Hispanics = −3.46, CSE = 1.67, p < .05) than White counterparts. Education was positively associated with all jobs (β = 2.95, CSE = 1.29, p < .05; Table 2).

Discussion

The main purpose of this study was to analyze the relationship between perceived neighborhood conditions and employment among emerging adults who were formerly incarcerated in the juvenile justice system using longitudinal data with three waves. Particularly, a strength of our study was to separately analyze the effect of neighborhood conditions on under-the-table jobs due to the fact that many young persons who had previously come in contact with the justice system and live in economically and socially disadvantaged communities tend to be in informal sectors (Wilson, 1987). The mean weeks worked across community and under-the-table jobs from wave 1 to wave 2 slightly increased, but remained steady from waves 2 to 3. Similar patterns were observed for community jobs and under-the-table jobs. Overall, our participants had lower weeks worked than the general population who worked more than 40 weeks per year on average over three decades (Pew, 2016). The mean perceived neighborhood condition scores slightly decreased over time.

Our findings from the hybrid models support the hypotheses, in that there are significant between-effects (i.e., inter-individual variation) of perceived neighborhood conditions on all jobs and community jobs, indicating that inter-individuals’ positive change in neighborhood conditions decreased weeks worked for all jobs and community jobs among emerging adults. Within-effects (i.e., intra-individual variation) also exist, in particular, in the association between perceived neighborhood conditions and under-the-table jobs. In other words, an intra-individual change in perceived neighborhood condition scores increased weeks worked for under-the-table jobs among emerging adults, which allows us to claim a causal effect of neighborhood perceptions on informal employment.

Neighborhood Conditions and Labor Market Participation

Regarding hypothesis 1, the between-estimator of perceived neighborhood conditions on all jobs and community-only jobs were negative and significant after controlling for both time-variant and invariant covariates. Our finding may fit Wilson’s concept and the SDT, elucidating that socioeconomically disadvantaged neighborhood conditions affects employment outcome. Many times, good neighborhood conditions/resources influence job accessibility, therefore creating economic opportunities along with social goods. However, living in disadvantaged neighborhoods with a record of juvenile justice involvement means these opportunities are not readily available, the well-being of those individuals and communities continues to diminish, in part due to systematic-level disparities, such as lower minimum wage (Ibragimov et al., 2019; Meltzer & Ghorbani, 2017) and chronic unemployment (Wilson, 1987) that are derived from disadvantaged neighborhood conditions, indicating an endless chain of circumstances. Although our findings of the between-effects may reaffirm the theory and some previous research (Alvarado, 2018; Galster et al., 2015), this paper was not able to control for the effects of other unobserved (time-varying) variables, such as geographic factors, due to data limitations. The insignificant within-effect of neighborhood conditions also suggests that unobserved confounders may play a role in the association between neighborhood conditions and employment. Thus, the between-effects of neighborhood perceptions on employment should be cautiously interpreted. Additionally, insignificant within-effect of neighborhood conditions on community jobs indicates that neighborhood conditions per se may not contribute to changing an individual’s proclivity for this type of work, which could be explained by previous studies (i.e., Ludwig et al., 2013) reported no significant treatment effects on job accessibility by adults randomly assigned in the MTO program.

Our results confirm hypothesis 2 by showing that an emerging adult has worked more weeks for under-the-table jobs over time as she/he perceives a greater degree of disorders within their community supported by the theories and previous studies (within; Curran, 2004; Hebert-Beirne et al., 2021). As theory and literature implied (Curran, 2004; Hebert-Beirne et al., 2021; Holzer et al., 2003; Wilson, 1987), the more active participation in the informal job market by justice-involved emerging adults mirrors disadvantaged neighborhood environments as a barrier where only informal work can be offered. Curran (2004) specifically argues that disadvantaged neighborhoods that underwent gentrification-induced displacement forced their residents to have informal jobs that were not recorded by the public sector.

Other Factors and Employment

Some time-variant factors are significantly associated with employment. Among them, what is more complex is about the associations between mental health, substance use, and employment where there has been no empirical consensus on the directionality of mental health and substance use on employment. For mental health, we found a negative association between depression and employment whereas a positive association between anxiety and employment was found (within). Prior studies report both depression (Lee & Lee, 2018; Perreault et al., 2017) and anxiety (Levinson & Lerner, 2009) negatively affects one’s employment status. Although our finding of anxiety contrasts with some studies (i.e., Levinson & Lerner, 2009), Aarons-Mele (2020) states that anxiety could be possibly a motivator to finding employment.

The relationship between substance use and employment is also complex with mixed results on literature. First, we found that those with alcohol use had more weeks worked in community jobs (between), which was in contrast to literature reporting chronic unemployment due to its adverse effects (Andersen, 2013). To clarify this result, future study needs to examine its reverse relationship as Virtanen et al. (2015) found work intensity was significantly associated with more alcohol use. A significant such reverse association may imply that alcohol use and employment have mutual influence to each other for this population. Second, marijuana use was negatively associated with community jobs (between) while it was also positively associated with under-the-table jobs (within). These results support previous studies (Kaestner et al., 2013; Lawn et al., 2016), in that people with a history of marijuana may be less motivated to work in the formal sector (between), but an individual who used marijuana tends to have higher labor participation in the informal sector (within; Lawn et al., 2016). Third, the negative relationship between drug use and employment (between) has been widely reported (DeSimone, 2002).

The results of perceived opportunity work (between) and gang membership (within) on employment are not surprising since prior studies widely support the results (Bishop et al., 2017; Cepeda et al., 2016; Consiglio et al., 2021; McArdle et al., 2007).

With respect to time-invariant variables, race and education have significant effects on employment. Black and Hispanic emerging adults are, specifically, less likely to work for both all jobs and community jobs than their White counterparts. Low employment among Black, Indigenous, People of Color (BIPOC) is well documented (BLS, 2014; Holzer, 2021); particularly, BIPOC who have previously come in contact with the justice system are disadvantaged in securing jobs (Western, 2006). Not surprisingly, we also find that emerging adults who enrolled in school before being incarcerated are more likely to work for all jobs than his/her counterparts, which is widely reported (BLS, 2016).

Policy Implications

Our findings reiterate some policy implications for formerly juvenile justice-involved emerging adults with the persistent conditions of participating more in informal jobs. Obtaining a clearer understanding of the negative effects of adverse neighborhood conditions on employment is critical for policy intervention. Most of all, policymakers should be aware that young individuals tend to participate less in formal jobs and more in informal jobs as they perceive a worsening of their community conditions, highlighting that poor neighborhood environments are a barrier to sustain their livelihoods. The opportunities to achieve their goals are not always present in the neighborhoods in which such emerging adults reside. Yet, they must survive. Therefore, it is necessary to seek other means of achieving these goals outside of traditional methods of gainful employment, higher education, and other markers of emerging adulthood.

Despite this reality, securing jobs in the formal sector is particularly crucial since it represents strong ties to informal social controls (Laub & Sampson, 2003). An absence of or insufficient informal social controls hinders building collective efficacy; and therefore cannot help buffer communities against the negative unforeseen consequences of adverse environments. As we have understood that space is a causal factor of residential mobility, and when policy interventions are designed under this context, major domains of neighborhood environments and employment should be focused as follows. Policy formulations need to account for how many justice-involved young people are willing to work for formal jobs; promote states to expunge juvenile criminal records that can be a structural barrier to continued education and job for this population; evaluate their social/human capital or networks for the opportunities of formal jobs; explore how potential employers have a bias against job applicants with a criminal history from such communities; anticipate how policy will help prevent stigma in these neighborhoods and their residents; assess whether public/local transportation systems are adequate, etc. (Galster, 2012; Turney et al., 2006).

Limitations

Our study has some limitations. First, as we previously discussed, we used the perceived neighborhood condition variables, which is subjective, so may make less accurate inferences. For future studies, neighborhood condition variables measured objectively using a geospatial format (i.e., GIS data) should be used to generalize the findings. Related to this, other factors related to neighborhood conditions (e.g., transportation access, neighborhood level socioeconomic status) integrated from the U.S. Census should be included to more precisely project neighborhood effects on employment. Since PDS does not provide relevant data, such as zip codes, it was not able to be matched with the census data, which limits our ability to include broader community factors. Second, the complex relationship between perceived neighborhood conditions and employment cannot be fully explored with a solely quantitative approach, indicating that mixed-methods strategies would allow for the identification of hidden contextual factors related to employment. Lastly, we used only one measurement of employment (i.e., weeks worked) and could not provide any examples of community jobs and under-the-table jobs due to data limitations. Literature has discussed the importance of measuring various types of employment (Wadsworth, 2006). So future study should use alternative operationalization for employment to provide an accurate picture of the neighborhood-employment relationship as an improvement to the current measurement.

Conclusion

This study sheds light on the effect of perceived neighborhood conditions on employment by youth transitioning to emerging adults who were formerly incarcerated in the juvenile justice system. Our results reaffirm the theory of social isolation and disorganization as well as neighborhood effects in literature. Given that many young people who were involved in the juvenile justice system have difficulty obtaining legitimate jobs, and frequently engage in informal job markets, our findings extend previous knowledge on neighborhood and employment. Additionally, when formulating a related policy to improve neighborhood conditions and employment outcomes for this populations, other proposed factors in this study should be considered. For example, policymakers need to develop and implement plans for increasing such young people’s perceived employability, improving access of mental health services and substance use treatments, preventing gang membership, reducing racial disparities in labor participation, and promoting educational opportunities.

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.

ORCID iD

Lewis H. Lee

Author Biographies

Lewis H. Lee (PhD, University of Pittsburgh) is an Assistant Professor in the School of Social Work at the University of Alabama. He mainly focuses on 1) examining micro and macro factors associated with reentry among justice-involved people and 2) identifying social determinants of health related to rural health disparities that has the potential to impact vulnerable rural populations.

Minseop Kim (PhD, University of Pennsylvania) is an Assistant Professor in the Department of Social Work at the Chinese University of Hong Kong. He has conducted research in two main areas: 1) examining associations among socio-economic factors, parenting behaviors, and child /youth development; and 2) identifying factors that affect well-being for children/youth involved with the child welfare system and juvenile justice system.

Catherine Carlson (PhD, Columbia University) is an Assistant Professor in the School of Social Work at the University of Alabama. Her research focuses on the intersection of violence and mental health, and the implementation of evidence-based interventions in low-resource settings.

Taylor Ellis (PhD, University of Alabama) is an Assistant Professor in the Department of Social Work at Jacksonville State University. Dr. Ellis’ research focuses on the intersection of arts and social work.

Karen Johnson (PhD, Columbia University) is an Assistant Professor in the School of Social Work at the University of Alabama. Dr. Johnson’s research focuses on 1) preventing HIV and decreasing co-occurring risks (e.g., intimate partner violence, homelessness, common mental health disorders); 2) implementing evidence-based interventions in real world settings (implementation science); and 3) cultural tailoring/adapting evidence-based interventions.

Angela Pretz (MSW, University of Iowa) is a PhD candidate in the School of Social Work at the University of Alabama. Her areas of research interest include 1) maternal and child health; and 2) racial disparities in the criminal justice system.

References

Aarons-Mele

(2020, May 11). Leading through anxiety. Harvard Business Review. https://hbr.org/2020/05/leading-through-anxiety

Allison

P. D.

(1999). Multiple regression: A primer. Pine Forge Press.

Allison

P. D.

(2009). Fixed effects regression models. SAGE.

Alvarado

S. E.

(2018). The impact of childhood neighborhood disadvantage on adult joblessness and income. Social Science Research, 70, 1–17. https://doi.org/10.1016/j.ssresearch.2017.10.004

American Psychological Association. (2015). Measuring socioeconomic status and subjective social status. Author. https://www.apa.org/pi/ses/resources/class/measuring-status

Andersen

(2013). Alcohol and the workplace. https://www.euro.who.int/__data/assets/pdf_file/0009/191367/8-Alcohol-and-the-workplace.pdf

Bishop

A. S.

Hill

K. G.

Gilman

A. B.

Howell

J. C.

Catalano

R. F.

Hawkins

J. D.

(2017). Developmental pathways of youth gang membership: A structural test of the social development model. Journal of Crime and Justice, 40(3), 275–296. https://doi.org/10.1080/0735648X.2017.1329781

Brenner

A. B.

Bauermeister

J. A.

Zimmerman

M. A.

(2011). Neighborhood variation in adolescent alcohol use: Examination of socioecological and social disorganization theories. Journal of Studies on Alcohol and Drugs, 72(4), 651–659.

Bureau of Labor Statistics. (2014). Labor force characteristics by race and ethnicity, 2013 (Report 1050). Author. https://www.bls.gov/opub/reports/race-and-ethnicity/archive/race_ethnicity_2013.pdf

10.

Bureau of Labor Statistics. (2016). Education matters. Author. https://www.bls.gov/careeroutlook/2016/data-on-display/education-matters.htm

11.

Bureau of Labor Statistics. (2020a). Employment recovery in the wake of the COVID-19 pandemic. Author. https://www.bls.gov/opub/mlr/2020/article/employment-recovery.htm

12.

Bureau of Labor Statistics. (2020b). Labor force statistics from the current population survey. Author. https://data.bls.gov/timeseries/LNS14000000

13.

Bureau of Labor Statistics. (2021). COVID-19 ends longest employment recovery and expansion in CES history, causing unprecedented job losses in 2020. Author. https://www.bls.gov/opub/mlr/2021/article/covid-19-ends-longest-employment-expansion-in-ces-history.htm

14.

Casciano

Massey

D. S.

(2008). Neighborhoods, employment, and welfare use: Assessing the influence of neighborhood socioeconomic composition. Social Science Research, 37(2), 544–558. https://doi.org/10.1016/j.ssresearch.2007.08.008

15.

Centers for Disease Control and Prevention. (2018). Mental health disorders and stress affect working-age Americans. Mental health in the workplace. Author. https://www.cdc.gov/workplacehealthpromotion/tools-resources/pdfs/WHRC-Mental-Health-and-Stress-in-the-Workplac-Issue-Brief-H.pdf

16.

Cepeda

Saint Onge

J. M.

Nowotny

K. M.

Valdez

(2016). Associations between long-term gang membership and informal social control processes, drug use, and delinquent behavior among Mexican American youth. International Journal of Offender Therapy and Comparative Criminology, 60(13), 1532–1548. https://doi.org/10.1177/0306624X15584985

17.

Chassin

Rogosch

Barrera

(1991). Substance use and symptomatology among adolescent children of alcoholics. Journal of Abnormal Psychology, 100, 449–463.

18.

Consiglio

Menatta

Borgogni

Alessandri

Valente

Caprara

G. V.

(2021). How youth may find jobs: The role of positivity, perceived employability, and support from employment agencies. Sustainability, 13, 9468. https://doi.org/10.3390/su13169468

19.

Couloute

Kopf

(2018). Out of prison & out of work: Unemployment among formerly incarcerated people. https://www.prisonpolicy.org/reports/outofwork.html

20.

Curran

(2004). Gentrification and the nature of work: Exploring the links in Williamsburg, Brooklyn. Environment and Planning A: Economy and Space, 36(7), 1243–1258. https://doi.org/10.1068/a36240

21.

Derogatis

Melisaratos

(1983). The Brief Symptom Inventory: An introductory report. Psychological Medicine, 13, 595–605.

22.

DeSimone

(2002). Illegal drug use and employment. Journal of Labor Economics, 20(4), 952–977. https://doi.org/10.1086/342893

23.

Dupere

Leventhal

Crosnoe

Dion

(2010). Understanding the positive role of neighborhood socioeconomic advantage in achievement: The contribution of the home, child care, and school environments. Developmental Psychology, 46(5), 1227–1244. https://doi.org/10.1037/a0020211

24.

Elo

I. T.

Mykyta

Margolis

Culhane

J. F.

(2009). Perceptions of neighborhood disorder: The Role of individual and neighborhood characteristics. Social Science Quarterly, 90(5), 1298–1320. https://doi.org/10.1111/j.1540-6237.2009.00657.x

25.

Galster

G. C.

(2012). The mechanism(s) of neighbourhood effects: Theory, evidence, and policy implications. In van Ham

Manley

Bailey

Simpson

Maclennan

(Eds.), Neighbourhood effects research: New perspectives (pp. 23–56). Springer.

26.

Galster

G. C.

Santiago

Lucero

(2015). Employment of low-income African American and Latino teens: Does neighborhood social mix matter? Housing Studies, 30(2), 192–227. https://doi.org/10.1080/02673037.2014.953447

27.

Garcia

(2014). Neighborhood disorder. In Michalos

A. C.

(Ed.), Encyclopedia of quality of life and well-being research (pp. 1–7347). Springer. https://doi.org/10.1007/978-94-007-0753-5_2751

28.

Gilman

A. B.

Hill

K. G.

Hawkins

J. D.

(2014). Long-term consequences of adolescent gang membership for adult functioning. American Journal of Public Health, 104(5), 938–945. https://doi.org/10.2105/AJPH.2013.301821

29.

Hebert-Beirne

Felner

J. K.

Berumen

Gonzalez

Chrusfield

M. M.

Pratap

Conroy

L. M.

(2021). Community resident perceptions of and experiences with precarious work at the neighborhood level: The greater Lawndale healthy work project. International Journal of Environmental Research and Public Health, 18(21), 11101. https://doi.org/10.3390/ijerph182111101

30.

Hollingshead

A. B.

(1975). Four factor index of social status [Unpublished manuscript]. Yale University.

31.

Holzer

H. J.

(2021). Why are employment rates so low among Black men? Brookings. https://www.brookings.edu/research/why-are-employment-rates-so-low-among-black-men/

32.

Holzer

H. J.

Raphael

Stoll

M. A.

(2003). Employment barriers facing ex-offenders. In Employment dimensions of reentry: Understanding the nexus between prisoner reentry and work. Urban Institute. https://www.urban.org/sites/default/files/publication/59416/410855-Employment-Barriers-Facing-Ex-Offenders.PDF

33.

Ibragimov

Beane

Friedman

S. R.

Komro

Adimora

A. A.

Edwards

J. K.

Williams

L. D.

Tempalski

Livingston

M. D.

Stall

R. D.

Wingood

G. M.

Cooper

(2019). States with higher minimum wages have lower STI rates among women: Results of an ecological study of 66 US metropolitan areas, 2003-2015. PLoS One, 14(10), e0223579. https://doi.org/10.1371/journal.pone.0223579

34.

Kaestner

Sasso

A. L.

Callison

Yarnoff

(2013). Youth employment and substance use. Social Science Research, 42(1), 169–185. https://doi.org/10.1016/j.ssresearch.2012.07.007

35.

Kirk

D. S.

(2019). The effects of neighborhood context and residential mobility on criminal persistence and desistence. In Farrington

D. P.

Kazemian

Piquero

A. R.

(Eds.), The Oxford handbook of developmental and life-course criminology (pp. 515–535). Oxford University Press.

36.

Kling

J. R.

Liebman

J. B.

Katz

L. F.

(2007). Experimental analysis of neighborhood effects. Econometrica, 75(1), 83–119.

37.

Laub

J. H.

Sampson

R. J.

(2003). Shared beginnings, divergent lives delinquent boys to age 70. Harvard University Press.

38.

Lawn

Freeman

T. P.

Pope

R. A.

Joye

Harvey

Hindocha

Mokrysz

Moss

Wall

M. B.

Bloomfield

M. A.

Das

R. K.

Morgan

C. J.

Nutt

D. J.

Curran

H. V.

(2016). Acute and chronic effects of cannabinoids on effort-related decision-making and reward learning: An evaluation of the cannabis ‘amotivational’ hypotheses. Psychopharmacology, 233(19–20), 3537–3552. https://doi.org/10.1007/s00213-016-4383-x

39.

Lee

L. H.

Kim

(2022). The effect of work on antisocial behavior among formerly incarcerated young adults: A panel analysis. Social Work Research, 46(2), 115–126. https://doi.org/10.1093/swr/svac003

40.

Lee

L. H.

Lee

(2018). Association between employment and mental health service use among justice-involved individuals. Community Mental Health Journal, 54(5), 634–640. https://doi.org/10.1007/s10597-017-0186-8

41.

Lei

M. K.

Beach

S. R. H.

(2020). Can we uncouple neighborhood disadvantage and delinquent behaviors? An experimental test of family resilience guided by the social disorganization theory of delinquent behaviors. Family Process, 59(4), 1801–1817.

42.

Levinson

Lerner

(2009). Employment among persons with past and current mood and anxiety disorders in the Israel National Health Survey. Psychiatric Services, 60(5), 655–662. https://doi.org/10.1176/ps.2009.60.5.655

43.

Ludwig

Duncan

G. J.

Gennetian

L. A.

Katz

L. F.

Kessler

R. C.

Kling

J. R.

Sanbonmatsu

(2013). Long-term neighborhood effects on low-income families: Evidence from moving to opportunity. American Economic Review, 103(3), 226–231.

44.

Madyun

N. H.

(2011). Connecting social disorganization theory to African-American outcomes to explain the achievement gap. Journal of Educational Foundations, 25, 21–35.

45.

McArdle

Waters

Briscoe

J. P.

Hall

D. T. T.

(2007). Employability during unemployment: Adaptability, career identity and human and social capital. Journal of Vocational Behavior, 71(2), 247–264.

46.

Meltzer

Ghorbani

(2017). Does gentrification increase employment opportunities in low-income neighborhoods? Regional Science and Urban Economics, 66, 52–73. https://doi.org/10.1016/j.regsciurbeco.2017.06.002

47.

Okechukwu

C. A.

Molino

Soh

(2019). Associations between marijuana use and involuntary job loss in the United States: Representative longitudinal and cross- sectional samples. Journal of Occupational and Environmental Medicine, 61(1), 21–28. https://doi.org/10.1097/JOM.0000000000001463

48.

Olesen

S. C.

Butterworth

Leach

L. S.

Kelaher

Pirkis

(2013). Mental health affects future employment as job loss affects mental health: Findings from a longitudinal population study. BMC Psychiatry, 13, 144. https://doi.org/10.1186/1471-244X-13-144

49.

Perreault

Touré

E. H.

Perreault

Caron

(2017). Employment status and mental health: Mediating roles of social support and coping strategies. The Psychiatric Quarterly, 88(3), 501–514. https://doi.org/10.1007/s11126-016-9460-0

50.

Pew Research Center. (2016, October). The state of American jobs: How the shifting economic landscape is reshaping work and society and affecting the way people think about the skills and training they need to get ahead. Author. https://www.pewresearch.org/social-trends/2016/10/06/the-state-of-american-jobs/

51.

Pew Research Center. (2021, April). U.S. labor market inches back from the COVID-19 shock, but recovery is far from complete. Author. https://www.pewresearch.org/fact-tank/2021/04/14/u-s-labor-market-inches-back-from-the-covid-19-shock-but-recovery-is-far-from-complete/

52.

Poor neighborhoods have 10% joblessness. (2018, February 4). Washington Post. https://www.washingtonpost.com/national/in-poor-neighborhoods-unemployment-is-10percent-survey-says/2018/02/04/a107828e-09f8-11e8-8b0d-891602206fb7_story.html

53.

Rose

D. R.

Clear

T. R.

(1998). Incarceration, social capital, and crime: Implications for social disorganization. Criminology, 36, 441–479.

54.

Sampson

R. J.

(2012). Great American city: Chicago and the enduring neighborhood effect. The University of Chicago Press.

55.

Sampson

R. J.

Raudenbush

S. W.

(1999). Systematic social observation of public spaces: A new look at disorder in urban neighborhoods. American Journal of Sociology, 105(3), 603–651.

56.

Shaw

C. R.

McKay

H. D.

(2014). Juvenile delinquency and urban areas. In Anderson

T. L.

(Ed.), Understanding deviance: Connecting classical and contemporary perspectives (pp. 106–127). Routledge/Taylor & Francis Group.

57.

Sheppard

Ricciardelli

(2020). Employment after prison: Navigating conditions of precarity and stigma. European Journal of Probation, 12(1), 34–52. https://doi.org/10.1177/2066220320908251

58.

Sorensen

Oliver

(2002). Policy reforms are needed to increase child support from poor fathers. Urban Institute. http://webarchive.urban.org/publications/410477.html

59.

Tanner

Davies

O’Grady

(1999). Whatever happened to yesterday’s rebels? Longitudinal effects of youth delinquency on education and employment. Social Problems, 46(2), 250–274.

60.

Tolman

R. M.

Himle

Bybee

Abelson

J. L.

Hoffman

Van Etten-Lee

(2009). Impact of social anxiety disorder on employment among women receiving welfare benefits. Psychiatric Services, 60(1), 61–66. https://doi.org/10.1176/ps.2009.60.1.61

61.

Turney

Clampet-Lundquist

Edin

Kling

J. R.

Duncan

G. J.

(2006). Neighborhood effects on barriers to employment: Results from a randomized housing mobility experiment in Baltimore (Brookings-Wharton Papers on Urban Affairs, 136–187). Brookings-Wharton.

62.

United Nations. (2005). The centrality of employment to poverty eradication: Report of the Secretary-General. Author. https://undocs.org/A/60/314

63.

Virtanen

Jokela

Nyberg

S. T.

Madsen

I. E.

Lallukka

Ahola

Alfredsson

Batty

G. D.

Bjorner

J. B.

Borritz

Burr

Casini

Clays

De Bacquer

Dragano

Erbel

Ferrie

J. E.

Fransson

E. I.

Hamer

. . . Kivimäki

(2015). Long working hours and alcohol use: Systematic review and meta- analysis of published studies and unpublished individual participant data. BMJ, 350, g7772. https://doi.org/10.1136/bmj.g7772

64.

Wadsworth

(2006). The meaning of work: Conceptualizing the deterrent effect of employment on crime among young adults. Sociological Perspectives, 49(3), 343–368.

65.

Western

(2006). Punishment and inequality in America. Russell Sage Foundation.

66.

Williams

(2018). Panel data and multilevel models for categorical outcomes: Hybrid models. https://www3.nd.edu/~rwilliam/Taiwan2018/Hybrid.pdf

67.

Wilson

W. J.

(1987). The truly disadvantaged: The inner city, the underclass, and public policy (2nd ed.). The University of Chicago Press.

68.

L. T.

Schlenger

W. E.

Galvin

D. M.

(2003). The relationship between employment and substance use among students aged 12 to 17. The Journal of Adolescent Health, 32(1), 5–15. https://doi.org/10.1016/s1054-139x(02)00447-0