View this issue as a PDF: CrimHighlightsV17N4.pdf
This issue of Criminological Highlights addresses the following questions:
- Why might it be useful for US President Trump to read this Highlight?
- Are some people’s lives made more difficult by prohibiting employers from asking job applicants if they have a criminal record?
- Are there risks to encouraging gun ownership?
- How can we predict which youths who offend as adolescents will still be committing offences in 10 years?
- Which Americans should prefer to be sentenced by a judge appointed by President Trump?
- How can police actions with youth lead youths to believe that violence is justified?
- How have changes in the law in England & Wales affected sentencing and imprisonment?
- Are youths from high socioeconomic families protected from the harmful effects of criminal justice contact?
Increases in the concentration of undocumented immigrants within states in the US are not associated with increases in violence. Indeed, if anything, increases in concentration of undocumented immigrants are associated with decreases in violent crime.
There is substantial evidence that residents of those locations lucky enough to attract large concentrations of immigrants receive a benefit: crime is reduced (see Criminological Highlights 5(4)#6, 10(6)#7, 11(1)#4, 13(6)#7, 14(1)#8, 16(1)#2, 17(1)#3. However, previous studies have examined those who are (largely) “legal” immigrants. This paper, using new estimates of the concentration of “undocumented” immigrants in the US, examines the effect of changes in their concentration on violent crime between 1990 and 2014.
“Immigration is said to reduce crime and violence by attracting immigrants with low criminal propensities, strengthening local economies and bolstering processes of informal social control.” Undocumented immigration may be different however, since “Unlike their documented counterparts, [they] are hindered from effectively forming economic and social ties as a result of their lack of legal standing in the community” (p. 372). In the US, they are denied all forms of federal assistance, are largely excluded from the formal labour market, and have limited opportunities for upward mobility.
Two different techniques have been used to estimate the number of undocumented immigrants. Essentially these approaches each subtract the number of authorized or documented immigrants from the estimates of the foreign born population in a given location. The two different methods show high correlations across years, within state (r=.93).
The analysis examines changes in the number of undocumented immigrants and violent crime, within states (and the District of Columbia), across years 1990 to 2014. Various controls are included – e.g., lawful immigration, structural disadvantage, unemployment rate, percent of the population that is young, firearms availability, police strength, and year (since, generally, crime was decreasing while undocumented immigration increased). With or without the controls, as concentrations of undocumented immigrants increased, the rate of violent crime decreased. “A standard deviation increase in undocumented immigration is associated with .37 standard deviation decrease in violent crime” (p. 384) an effect which is quite large. It is important to note that “The [favourable] effects of unauthorized immigration are in addition to the significant reductions in violent crime stemming from lawful immigration” (p. 384). It was demonstrated that the effects are not driven by undocumented immigrants selecting locations to avoid violence.
The analyses were repeated using National Crime Victimization Survey data (using 3 year averages) and the results were similar. Hence the findings do not appear to be due to differential reporting to the police. The results of the full set of analyses “seriously undermine claims that violent crime has increased as a result of undocumented immigration” (p. 390).
Conclusion: People who can’t easily be counted or identified publicly are easy targets for blame for crime, among other social problems. The findings from this study show no support whatsoever for the view that violent crime increases when undocumented immigration increases in a US state. Indeed most of the findings suggest the opposite: that the effects of increased undocumented immigration to a state (or Washington, D.C.) are the same as increases in legal immigration: crime decreases. It’s too bad that the President of the US neither looks at evidence nor subscribes to Criminological Highlights.
Reference: Light, Michael T. and Ty Miller (2018). Does Undocumented Immigration Increase Violent Crime? Criminology, 56(2), 370-401.
Prohibiting employers from asking about criminal records on job applications might appear to be a good way of giving those with criminal records a chance at getting a job. Paradoxically, these well-meaning policies appear to disadvantage certain already-disadvantaged people: young, Black or Hispanic, unskilled males.
Allowing those who have been convicted of a criminal offence a chance of getting a job is obviously not only fair, but is likely to be an important step in the reintegration process. However, it appears that some employers are reluctant even to consider those with criminal records.
Between 2004 and 2014, therefore, some cities and states instituted ‘ban the box’ (BTB) policies where applicants cannot be forced on their initial applications to indicate they have criminal records. This study took advantage of the fact that BTB policies came in at different times in different locations. Looking across time, within locations that have and do not have BTB policies, the study used large employment surveys to examine the employment rates of relatively unskilled non-Hispanic Whites, non-Hispanic Blacks, and Hispanic young men.
The possible risk of BTB policies is that employers – no longer being able to identify those with criminal records – will instead screen out (and not consider) those whom they think might have criminal records: young, unskilled, Black/Hispanic men. Hence even though BTB policies might have been instituted to help these groups, they could, paradoxically, hurt their chances.
The analysis looked at the monthly employment rates of these three groups for 18 months before and 18 months after BTB policies were instituted. In order to create comparison non-BTB locations, employment rates for the three groups were examined in locations without BTB policies for 18 months before and after the average date of the implementation of BTB policies elsewhere (October 2010).
The results were clear. For White men, there was no apparent impact of the implementation of BTB policies: employment rates did not change in locations with the implementation of BTB policies. Indeed, employment rates during the 36 month period were comparable for BTB and non-BTB locations. For Blacks and Hispanic men, however, the employment rates in BTB and non-BTB cities were comparable before the implementation of BTB. But after the BTB rules were adopted, employment rates for relatively unskilled Black and Hispanic young men decreased and were significantly lower than the employment rates for these men in non-BTB cities.
Conclusion: It appears that for vulnerable groups – young, relatively unskilled Black and Hispanic men – Ban-the-Box policies had harmful effects. These policies were designed to allow job applicants to be considered for employment rather than being immediately rejected because of a criminal record. Given that many employers are worried about hiring people who have criminal records, it would appear that when they are prohibited from asking for this information, they may avoid doing interviews with people whom they assume are likely to have criminal records. Young, unskilled Blacks are about 3.4% less likely to be employed after these policies are brought in; the comparable figure for Hispanics is 2.3%. There seemed to be no benefit (or harm) to comparable White job seekers.
Reference: Doleac, Jennifer L. and Benjamin Hansen (2016). Does “Ban the Box” Help or Hurt Low-Skilled Workers? Statistical Discrimination and Employment Outcomes When Criminal Histories are Hidden. Working Paper 22469. National Bureau of Economic Research (Cambridge, MA, USA).
In December 2012, a lone shooter entered Sandy Hook Elementary School in Newtown, Connecticut and killed 20 children and 6 adults. Five days later, the president of the US, Barack Obama, suggested that it was important to find ways of using new federal legislation to restrict access to guns. He discussed specific legislative proposals in January and in his February 2013 State of the Union address.
This paper examines firearms acquisition and accidental deaths in the immediate aftermath of the Sandy Hook shootings. “Previous research has demonstrated that accidental shootings are more likely to occur when there are more guns in the home, during routine handling of a firearm, when a gun is not stored properly, and when people are playing with guns or demonstrating their use. Children may be particularly at risk” (p. 1324).
Interest in purchasing firearms was thought to be heightened because of Obama’s suggestion of new restrictions on their purchase. In the months immediately following the Sandy Hook shootings (December 2012-April 2013) there was a dramatic spike in Google searches that included “buy gun” (e.g., “Where can I buy a gun?”) and “clean gun” (e.g., “How often should I clean my gun?”). Rates of these searches were dramatically higher than they had been at any time between 2010 and the end of 2014. The correlation between the volume of these searches was +0.72 . In addition, ‘background checks’ (necessary in some locations for purchases of a firearm from a licensed firearms dealer), a measure that had previously been shown to relate to firearms purchases, also spiked during this same period and correlated highly with the Google searches on ‘buy gun’. In addition, California sales data suggest that an additional 26,000-33,000 handguns were purchased in that state in the 6 weeks after the Sandy Hook killings. Most of these sales were by first-time purchasers. Across the US (population 316 million in 2013), it is estimated that approximately 3 million additional guns were sold during this 5 month period.
Using Vital Statistics data (which record causes of death), it was discovered that there was a spike in accidental firearms deaths during the 5-month window when purchases of firearms spiked. There were no other spikes in firearms accidental deaths in any other period between 2010 and 2014. In all, using two different methods of estimating the number of additional deaths, it would seem that there were between 57 and 66 additional firearms accidental deaths in the US in this 5-month period, with 18-22 of them being children (age 14 and under). The spikes in accidental firearms deaths occurred precisely where one would expect: in states with (estimated) large per capita sales increases.
Conclusion: It is sometimes suggested that “Guns don’t kill, people do.” As noted elsewhere (Criminological Highlights 17(3)#2) crimes carried out with firearms are more likely to result in deaths than comparable crimes carried out with other weapons. The data in this study demonstrate that there are measurable dangers from firearms even when they are not being used in relation to a violent crime. The easy access to firearms after the Sandy Hook school shootings, and the belief that access might be made more difficult, led to a spike in firearms sales in the US. This increase in firearms in ordinary homes led to increased numbers of firearms accidents, most notably, involving children. More generally, the data demonstrate the risk of accidental deaths that flow directly from firearms possession. Guns kill.
Reference: Levine, Phillip B. and Robin McKnight (8 December 2017). Firearms and Accidental Deaths: Evidence from the Aftermath of the Sandy Hook School Shooting. Science, 358, 1324-1328.
The differentiation between “life-course persistent” and “adolescent-limited” male offenders, first made in 1993, is still useful 25 years later. The need to shape policies in a manner that is consistent with our knowledge of these two very different groups is more clear than it was 25 years ago.
“Antisocial adults virtually always begin as children with antisocial misconduct, but most young people who engage in antisocial misconduct do not grow up to be antisocial adults” (p. 177). An important 1993 paper made this point. This paper reviews evidence since then and serves as a reminder that the causes of “life course persistent” and “adolescent limited” offending are very different.
Using reports from parents and teachers about children from age 5 to age 11, and combining these data with self-reports, police-reports and parent and teacher reports of antisocial behaviour in mid-adolescence, two groups of offending youths can be identified. In mid-adolescence, the anti-social behaviour rates of these two groups are generally indistinguishable. But one group had experienced early risks (e.g., parental mental health problems, single parent families, harsh discipline, neurological abnormalities, reading and memory deficits, hyperactivity, fighting, peer rejection). The other group had not. Both groups of youths who engaged in anti-social behaviour were likely to have delinquent friends.
However, by age 26, the “adolescent limited” youths, who did not experience these early risks and deficits, were much less likely to engage in violent or non-violent offending than the “life course persistent” youths. Said differently, the behaviour in mid-adolescence was, for the two groups, likely to be caused by, and under the control of, very different factors. Equally important, perhaps, is the fact that looking only at the delinquent behaviour of a youth does not allow one to determine which group a youth might fall in.
Although there is some evidence that “life course persistent behaviour is characterized by difficulties in the brain’s mental functions, particularly its verbal and executive functions… decisive tests of the hypothesis that neural abnormality should characterize life course persistent but not adolescent limited individuals are lacking” (p. 183). It has been suggested that what drives adolescent limited offending “is a gap between biological and social maturation… measured as the discrepancy between pubertal age and the amount of autonomy each teen reported” (p. 183). In young adulthood, some adolescent limited individuals “had continued difficulties, particularly related to substance abuse” but typically, in the end, have “ordinary life outcomes.” However, this delay suggests that in our society “the transition from adolescence to adulthood has elongated” (p. 183) and there is some evidence that the peak rate of crime for adolescent limited males is occurring slightly later than was the case in the 1980s.
Conclusion: Looking back at the distinction between life course persistent and adolescent limited offending, at least two conclusions might be drawn. First, since individual development appears to be one driver of serious life course persistent (male) offending there is a need for early-childhood prevention aimed at families and schools. Second, there is “the need to appreciate heterogeneity within adolescent offenders, to distinguish the few who have adverse backgrounds from the many who have ordinary backgrounds and a better prognosis” (p. 183). From society’s perspective, social interventions to reduce the number of life course persistent offenders are considerably more cost effective if they are well targeted on this relatively small group. At the same time the information about adolescent limited antisocial behaviour “provides impetus for movements to reform juvenile-justice and mental-health services in directions that are more supportive for young people” (p. 183-4).
Reference: Moffitt, Terri E. (2018). Male Antisocial Behaviour in Adolescence and Beyond. Nature: Human Behaviour, 2, 177-186.
Politics matter in sentencing. Federal judges appointed by Republican presidents in the US are not always tougher on defendants: they give harsher sentences to black offenders than Democrat-appointed judges, but more lenient sentences, overall, to women.
There are data that suggest that even in fairly strict sentencing guideline systems (e.g., the US federal guidelines) disadvantaged groups are treated more harshly. Furthermore, there are systematic differences across judges and prosecutors in the sentences that offenders receive (see Criminological Highlights 5(3)#6, 15(5)#4). This paper examines whether judges’ sentencing decisions vary in systematic ways by looking at the political party of the president who appointed the judge.
Concern about the politicization of judicial appointments in many countries (notably in the US) makes it important to see what the impact might be of having a large number of judges appointed by any given political party. One might have thought that in guideline systems like the US federal guidelines, where the ‘presumptive range’ of sentences for a given offence/criminal record combination is relatively narrow (e.g., typically the top of the range is about 20-25% higher than the bottom of the presumptive range), judges would be fairly constrained in their sentencing and the political leader responsible for their appointment would not matter. This paper suggests that this is wrong: politics matter.
This paper used data from the United States Sentencing Commission which had many characteristics of the case, but did not include the identity of the judge. Using other data obtained through freedom of information requests, the identity of the judge could be determined in over half a million cases decided between 1999 and 2015. 44% of the cases were decided by judges appointed by Democratic presidents. Democrat-appointed judges were more likely than Republican-appointed judges to be Black and/or female. The cases sentenced by Democrat-appointed judges did not differ appreciably from the cases sentenced by Republican-appointed judges.
“Republican appointed judges give black offenders an additional 3.0 months in prison compared to non-black offenders, relative to Democratic judges in the same district court… Republican judges give female offenders 2.1 fewer months in prison compared to males, relative to Democratic judges” (page 13). These results hold even when the ‘final offence level’ of the case is controlled for. This means that even though judges might not formally depart from the ‘guideline range’, blacks, relative to whites, as well as men relative to women, tend to be treated more harshly by Republican judges than Democratic judges. The gender effects – Republican judges being more lenient, relatively speaking, with women – are driven largely by violent and drug offences. The race effects – Republican judges being relatively harsh in their sentencing of blacks — are driven largely by drug and property offences. Various analytical techniques were used in the study to ensure that the results were not related to any specific method of examining the data.
In 2005, the US Supreme Court (US v. Booker) made the guidelines more advisory than mandatory though judges were required to take the guidelines into account. After this case had been decided, racial and gender disparities by judge political affiliation became larger. The difference between the sentences for black and white offenders sentenced by Republican judges increased to 4.8 extra months in prison, compared to sentences from Democratic judges.
Conclusion: Judicial ideology – in this study operationalized by whether a judge was appointed by a Republican or Democratic president — makes a difference. The precise reasons that blacks, relative to whites, are treated more harshly by Republican appointed judges and women, relative to men, are treated more leniently by Republican appointed judges are not clear. The study suggests, then, that part of the reason that sentencing disparity still exists in sentencing guideline systems is that judges are able systematically to sentence according to what apparently are their political beliefs.
Reference: Cohen, Alma and Crystal S. Yang (December 2017). Judicial Politics and Sentencing Decisions. Discussion Paper No. 943. The John M. Olin Centre for Law, Economics, and Business. Harvard University.
Police contacts with youth can have negative effects on perceptions of procedural injustice, willingness to report crimes to the police, support for the use of violence in response to problems, and reported delinquency.
There is substantial empirical support for the conclusion that fair treatment by the police can be both “a form of crime control” (p. 403) and have other beneficial effects (see Criminological Highlights 7(2)#4, 11(4)#1, 11(5)#1, 15(1)#5, 15(4)#3, 17(2)#4). This paper uses data from a longitudinal survey of Black, White, and Latino urban youths in 7 US cities to examine some of the potentially harmful effects of criminal justice contact when that contact is not experienced in a favourable light.
High school youths’ experiences in encounters they have had with the police in the previous 6 months were categorized in one of four different ways (with the most recent stop being examined if they had been stopped more than once): No police-initiated stops, police-initiated stops in which they were satisfied or very satisfied, stops in which they were neither satisfied nor dissatisfied, or police-initiated stops in which they were dissatisfied or very dissatisfied.
They were also asked not only for their perceptions of procedural justice, but also whether they would be likely to report crimes they witness to the police and their belief in the need for violent self-help (e.g., whether it is all right to use violence when another person “hit you first” or is “threatening to hurt your family”). Finally, a year later, the youths were asked about their involvement in 11 different forms of delinquent behaviour.
For all three groups of youths (White, Black, and Latino), unsatisfactory treatment by the police led to perceptions that the police generally were procedurally unjust (discourteous, disrespectful, dishonest). For Whites and Latinos, receiving neutral or unsatisfactory treatment from the police led to a lower willingness to report crime to the police. However, being stopped by the police had no impact on willingness to report crime to the police for Black youth. In general, being stopped by the police increased youths’ support for the personal use of violence, with relatively little variation due to youths’ characterization of the interaction. Youths who were stopped and, in particular, those youths who reported unsatisfactory treatment from the police were more likely to engage in delinquencies a year later.
All of these effects controlled for youths’ assessment of procedural justice more than a year before the police stop that was the subject of this study. In other words, the effects appear to be caused by the youths’ recent experiences with the police. Various demographic variables were controlled for as well (e.g., household structure, residential mobility, parental education). Personal characteristics such as impulsivity, monitoring by parents, prosocial peers, victimization as well as youths’ experiences and performance in schools and whether they had participated in a program designed to keep youths from joining gangs were, in addition, included as controls. Finally, the city the youth was in and various neighbourhood characteristics (e.g., racial mix of the neighbourhood, economic disadvantage and disorder in the school and neighbourhood) were included in a long list of factors that were statistically controlled. In other words, it seems likely that the effects that were observed were the direct result of police treatment of youths, not these other factors.
Conclusion: In general, “youth who were stopped or arrested and were dissatisfied with how they were treated had more negative outcomes [lower willingness to report crimes, increased willingness to use violence in personal disputes, and increased involvement in delinquency] when compared with those with no police contact regardless of their race/ethnicity…. For Whites and Latinos, even neutral contacts seem to weaken their relationship with the law and mainstream norms” (p. 422). For Black youths, on the other hand, it may be that they see people like them targeted by the police so regularly, that neutrally-rated police contact has little additional negative impact.
Reference: Slocum, Lee Ann and Stephanie Ann Wiley (2018). “Experience of the Expected?” Race and Ethnicity Differences in the Effects of Police Contact on Youth. Criminology, 56(2). 402-432.
The imprisonment rate in England and Wales rose substantially between 1990 and 2011 and has been relatively steady since then, notwithstanding a number of changes that have taken place in sentencing in recent decades and a 21% decrease in the number of people being sentenced between 2004 and 2014.
In recent years, there have been a number of important changes in the sentencing of criminal cases in England & Wales including the development of sentencing guidelines, the development of a new non-prison sanction (the ‘suspended sentence order’), and the development and then abolition of an ‘indeterminate sentence’ in cases not eligible for a life sentence.
In 2003, Britain’s Parliament changed the requirements for the use of a ‘suspended sentence order’ in order to increase its availability (and presumably capture some of the ‘immediate custody’ sentences that did not really need to end up in Britain’s prisons). Prior to this, the suspended sentence order was only available in exceptional circumstances and was almost never used. There was a large increase in the use of this option apparently as a result of this change in the law. Looking at the more serious category of cases (indictable offences), suspended sentences went from being used in 1% of all cases in 2003 to 14% of all cases in 2014. One might think that this would result in less use of ordinary custodial sentences. It did not. Immediate custody (ordinary prison) sentences also were more frequently used. Ordinary prison sentences were imposed in 24% of indictable offence cases in 2003 and 27% of cases in 2014. Most of the shift to suspended sentence orders came from an explicitly non-custodial choice: the ‘community sentence’.
In 2002, apparently believing in the efficiency of incapacitation as a crime control strategy, Britain’s Parliament enacted the indeterminate sentence of “Imprisonment for public protection” (IPP) which could be imposed on any person convicted of any of 153 specified offences (some with maximum sentences of only 3 years). By 2008, it became clear that it was being used more than expected. By 2012, when the indeterminate sentence was abolished, those subject to an IPP sentence constituted 44% of those serving life/indeterminate sentences and 7% of the total prison population.
English guidelines are offence specific and are structured in a very different manner from American ‘grid’ guidelines. The first guideline was issued in 2011. Judges are required to follow a formal process to arrive at a sentence, starting with an assessment of the relative seriousness of the harm that was done and the culpability of the accused. Then other factors (e.g., criminal record, a guilty plea) are considered. Hence judges “must… follow any sentencing guidelines which are relevant to the offender’s case” (p. 337), but at the same time judges are allowed to depart from the guideline “if following the guideline would be contrary to the interests of justice” (p. 337). Control of prison populations through guidelines was seen as impractical and perhaps inappropriate. Hence this is not a goal of the English guidelines.
The tension between judges’ beliefs and practices, one the one hand, and principled guidelines on the other hand can be seen by the response, in 2011, to a wave of urban riots in England. Judges responded by handing down “strikingly harsher sentences in order to deter future cases” (p. 342). England’s Sentencing Council followed judicial practice by subsequently adding the aggravating factor of “offence committed during a period of social disorder” to all subsequent guidelines. Sentence lengths in Crown Court for violent sexual offences also increased, a change that “appears unrelated to the guidelines per se” (p. 338).
Conclusion: Many changes have occurred in sentencing in England & Wales since 2000. Some of the effects are familiar (e.g., new non-prison sanctions not reducing the use of ordinary prison sentences). Some, such as the English guidelines, are less familiar. The examination of the effects of these changes on imprisonment, however, serve as a reminder of the somewhat unpredictable relationship between legal change and changes in imprisonment.
Reference: Roberts, Julian V. and Andrew Ashworth (2016). The Evolution of Sentencing Policy and Practice in England & Wales. Crime and Justice: A Review of Research, 45, 307-358.
Involvement with the criminal justice system as a high school student decreases the likelihood of having a good education and a well-paying job at age 25-32 – especially for those youths who grew up in relatively high socioeconomic status families.
There is little doubt that contact with the criminal justice system has negative impacts on people’s lives and the lives of their families long after the immediate consequences are over. This study examines how prior criminal justice experiences, ranging from having been stopped by the police to being incarcerated, affects socioeconomic attainment. In addition, it examines whether the effects, if any, vary across youths coming from different socioeconomic backgrounds.
Previous research has shown that being arrested in adolescence is associated with a reduced likelihood of obtaining a high school degree and increased likelihood of eventually being on welfare. But there is no reason to expect that the effects of criminal justice involvement are going to be uniform for all youths. One could hypothesize that the harmful effects of criminal justice contact for a youth would be smaller for advantaged youths because they have additional (family) resources to draw on to overcome the negative impact. Alternatively, one might predict the opposite: that the effects would be larger for youths from advantaged homes because criminal justice involvement is, for them, more stigmatizing or that it is less common and therefore youths from advantaged families have less experience in dealing with the consequences of this contact.
This study uses data obtained from a longitudinal study in which the first interview took place when youths were approximately age 12-19. The youths were followed and data were collected again when they were age 25-32. The main focus of this study is on their educational attainment and occupational status (ranging from being unemployed to having a high status job). Data on criminal justice involvement were collected when the youths were age 18-26 and were coded in terms of the ‘depth’ of involvement in the criminal justice system. This ranged from there being no contact to being stopped for non-traffic matters, charged, convicted, given a non-jail/prison sentence, or sentenced to jail/prison. Parental socioeconomic status was coded from the data collected when the youth was first interviewed (at age 12-19). Various other demographic factors (e.g., race/ethnicity, whether the youth lived with biological parents) were used as control factors.
The results show that as criminal justice involvement increased (from no involvement, at one end of the spectrum, to being incarcerated at the other end) the achieved socioeconomic status of the youth (at age 25-32) was reduced. This effect was dramatically larger for the youths who grew up in relatively advantaged families. Said differently, criminal justice involvement hurts advantaged youths more than it hurts less advantaged youths. But in general, the deeper the level of criminal justice involvement for all groups, the less likely they were to achieve high socioeconomic status (measured as educational attainment or employment status). One might think that youths from advantaged homes will experience the greatest loss from criminal justice involvement, but still achieve more than their lower socioeconomic status counterparts. For those who managed to make it through adolescence and early adulthood without criminal justice involvement or with low forms of involvement, parental socioeconomic status predicts the youth’s educational and employment achievements. But as involvement in the criminal justice system deepens, the differences between youths of different socioeconomic backgrounds disappears.
Conclusion: These data demonstrate that there are long-term – perhaps lifetime – negative consequences of deep levels of criminal justice involvement, in particular being convicted of a crime and incarcerated. Youths who have been convicted or incarcerated show lower educational and employment achievement when they are age 25-32. These effects are particularly pronounced for youths who come from advantaged families.
Reference: Dennison, Christopher R. and Stephen Demuth (2018). The More You Have, The More You Lose: Criminal Justice Involvement, Ascribed Socioeconomic Status, and Achieved SES. Social Problems, 65, 195-210.