Search

A very small number of young people enter youth custody between age 16 and 18 (about 4 in 1000 males), yet the consequences are severe. They spend an average of 7 months in youth custody and such incarceration has been related to negative outcomes in the longer term even if they can establish themselves in the labour market. In this paper, we evaluate whether there is a relationship between GCSE qualifications in English and maths and the probability of youth custody using administrative data in England. We are hindered in this because the majority of young people who end up in youth custody are not entered or fail their GCSEs in these subjects. Although regression results are consistent with educational achievement being a factor in why people end up in youth custody, they strongly suggest that both non-entry/low achievement and youth custody are correlated with severe vulnerabilities which are partially picked up by the explanatory variables available in administrative data (in particular indicators for special needs, disadvantage and being from some ethnic minority groups). Another interesting insight is that for many, problems only emerge (or at least become evident) in early or middle adolescence. [This paper forms part of a larger project funded by the Nuffield Foundation "Youth custody: Educational influences and labour market consequences."]

This paper studies the impact of compulsory schooling on violent behaviour and victimization in school using individual-level administrative data matching education and criminal records from Queensland (Australia). Exploiting a legislative increase in the minimum dropout age in 2006, this study defines a series of regression-discontinuity specifications to show that compulsory schooling reduces crime but increases violent behaviour in school. While police records show that property and drugs offences decrease, education records indicate that violence and victimization in school increase. Thus, prior studies that fail to consider in-school behaviour may over-estimate the short-run crime-reducing impact of compulsory education.

Using value-added models, we find that high schools impact students' self-reported socio-emotional development (SED) by enhancing social well-being and promoting hard work. Conditional on schools' test score impacts, schools that improve SED, reduce school-based arrests, and increase high-school completion, college-going, and college persistence. Schools that improve social well-being have larger effects on attendance and behavioral infractions in high school, while those that promote hard work have larger effects on GPA. Results suggest that adolescence can be a formative period for socio-emotional growth, high-school impacts on SED can be captured using self-report surveys, and SED can be fostered by schools to improve longer-run outcomes. These findings are robust to tests for plausible forms of selection.

Exposure to stressful life experiences during childhood, such as food insecurity, can have negative consequences for attainment later in life. The developmental timing of stressful events and how they influence outcomes over the life course is a critical area of research. Indeed, a more comprehensive understanding of the latter life consequences of childhood food insecurity could guide policy-makers in designing more effective social policies to reduce the severity of the poor life outcomes. This project uses data from the Panel Study of Income Dynamics to estimate the young adult impacts (as late as age 25) of food insecurity experienced in discrete childhood stages -- middle childhood (ages 5-10), early adolescence (ages 11-14), and middle adolescence (ages 15-18). It aims to identify which childhood stage-specific effects of food insecurity are most important to five young adult outcomes in two main areas -- risky sexual behaviors and criminal justice involvement. Results provide consistent evidence that the mean food security scores in middle childhood are associated with the criminal justice involvement outcome. The results are less consistent with the sexual risk taking outcomes. Middle childhood food insecurity is associated with the number of sexual partners in young adulthood, while early adolescent food insecurity is associated with the number of children in young adulthood. Results indicate that male respondents appear to be more sensitive to food insecurity than females.

Are private schools safer than public schools? Do they use the same security procedures? Do they suspend or expel students at a higher rate? Do private schools that participate in school choice programs report the same safety issues as private schools that do not participate in school choice programs? How does charter school safety compare to traditional public school safety? This report seeks to rigorously address these questions--and more--by collecting data from 618 school leaders across Indiana. We used their responses to empirically examine the relationship between private schooling and the presence of safety-related practices such as building access, security cameras, wearing of badges or picture IDs and more.

This paper examines the impact of school accountability on adult crime and economic self-sufficiency. We employ a unique source of linked administrative data from a Southern state and exploit exogenous variation generated by the state's accountability regime. Our findings indicate that a school's receipt of a lower accountability rating, at the bottom end of the ratings distribution, decreases adult criminal involvement. Accountability pressures also reduce the propensity of students' reliance on social welfare programs in adulthood and these effects persist at least until when individuals reach their early 30s. Further examination reveals that our results are consistent with an explanation related to improvements in human capital accumulation.

Prior research shows reduced criminality to be a beneficial consequence of education policies that raise the school leaving age. This paper studies how crime reductions occurred in a sequence of state-level dropout age reforms enacted between 1980 and 2010 in the United States. These reforms changed the shape of crime-age profiles, reflecting both a temporary incapacitation effect and a more sustained, longer run crime reducing effect. In contrast to the previous research looking at earlier US education reforms, crime reduction does not arise solely as a result of education improvements, and so the observed longer run effect is interpreted as dynamic incapacitation. Additional evidence based on longitudinal data combined with an education reform from a different setting in Australia corroborates the finding of dynamic incapacitation underpinning education policy-induced crime reduction.

School choice lotteries are an important tool for allocating access to high-quality and oversubscribed public schools. While prior evidence suggests that winning a school lottery decreases adult criminality, there is little evidence for how school choice lotteries impact non-lottery students who are left behind at their neighborhood school. We leverage variation in actual lottery winners conditional on expected lottery winners to link the displacement of middle school peers to adult criminal outcomes. We find that non-applicant boys are more likely to be arrested as adults when applicants from their neighborhood win the school choice lottery. These effects are concentrated among boys who are at low risk of being arrested based on observables. Finally, we confirm evidence in the literature that students who win the lottery decrease adult criminality but show that after accounting for the negative impact on the students who forego the lottery, lotteries increase overall arrests and days incarcerated for young men.

This paper summarizes high-quality research findings, which show that intrapersonal and interpersonal MESH [Mindsets, Essential Skills, & Habits] skills, such as self-control and social competence, are well-established predictors of success in academics, career, and well-being. Given the importance of these outcomes and the strength of the existing research, it is time for these competencies to be incorporated effectively into educational policy and practice as complements to existing academic and cognitive goals in order to ensure schooling works to help all students flourish.

In this paper we examine how failing to make adequate yearly progress under No Child Left Behind (NCLB), and the accountability pressure that ensues, affects various non-achievement student behaviors. Using administrative data from North Carolina and leveraging a discontinuity in the determination of school failure, we examine the causal impact of accountability pressure both on student behaviors that are incentivized by NCLB and on those that are not. We find evidence that, as NCLB intends, pressure encourages students to show up at school and to do so on time. Accountability pressure also has the unintended effect, however, of increasing the number of student misbehaviors such as suspensions, fights, and offenses reportable to law enforcement. Further, this negative response is most pronounced among minorities and low performing students, who are the most likely to be left behind. The following are appended: (1) Descriptive Statistics; (2) Offense Measures Used; (3) Supplemental Analyses; and (4) Alternate Model Specifications.

This paper asks whether improving the quality of public schools can be an effective long-run crime-prevention strategy in the U.S. Specifically, we examine the effect of school quality improvements early in children's lives on the likelihood that they are arrested as adults. We exploit quasi-experimental variation in school quality due to increases in public school funding, leveraging two natural experiments in Michigan and a novel administrative dataset linking the universe of Michigan public school students to adult criminal justice records. The first research design exploits variation in operating expenditures due to Michigan's 1994 school finance reform, Proposal A. The second design exploits variation in capital spending by leveraging close school district capital bond elections in a regression discontinuity framework. In both cases, we find that students exposed to additional funding during elementary school were substantially less likely to be arrested in adulthood. We show that the Marginal Value of Public Funds of improving school quality (via increases in funding) is greater than one, even when considering only the crime-reducing benefits. [Additional funding for this research was provided by the Early Career Scholars Grant at Policy Impacts.]

We analyze the effect of Colombia's ambitious "Free Housing" program on children's educational outcomes. The program was generous, giving free housing to beneficiaries in desirable areas. We evaluate the program by leveraging housing lotteries and linking applicants to their children. We find that public housing increases high school graduation by seven percentage points -- a seventeen percent increase relative to the control mean -- and boosts exit exam scores and college-going. Using a survey to explore mechanisms, lottery winners report better environmental conditions and shorter commute times. Their children also attend better schools and live in neighborhoods with less crime. [Additional funding for this project was provided by the Departamento Nacional de Planeacion (Contract 672-2019).]

This paper presents the first analysis of the life course outcomes through late midlife (around age 55) for the participants of the iconic Perry Preschool Project, an experimental high-quality preschool program for disadvantaged African-American children in the 1960s. We discuss the design of the experiment, compromises in and adjustments to the randomization protocol, and the extent of knowledge about departures from the initial random assignment. We account for these factors in developing conservative small-sample hypothesis tests that use approximate worst-case (least favorable) randomization null distributions. We examine how our new methods compare with standard inferential methods, which ignore essential features of the experimental setup. Widely used procedures produce misleading inferences about treatment effects. Our design-specific inferential approach can be applied to analyze a variety of compromised social and economic experiments, including those using re-randomization designs. Despite the conservative nature of our statistical tests, we find long-term treatment effects on crime, employment, health, cognitive and non-cognitive skills, and other outcomes of the Perry participants. Treatment effects are especially strong for males. Improvements in childhood home environments and parental attachment appear to be an important source of the long-term benefits of the program. [This research was supported in part by the American Bar Foundation.]

This paper examines the impact of the iconic Perry Preschool Project on the children and siblings of the original participants. The children of treated participants have fewer school suspensions, higher levels of education and employment, and lower levels of participation in crime, compared with the children of untreated participants. Impacts are especially pronounced for the children of male participants. These treatment effects are associated with improved childhood home environments. The intergenerational effects arise despite the fact that families of treated subjects live in similar or worse neighborhoods than the control families. We also find substantial positive effects of the Perry program on the siblings of participants who did not directly participate in the program, especially for male siblings. [Additional support for this research was provided by the American Bar Foundation.]

Early learning matters and can be strengthened and supported through contexts where children spend much of their time. Strong early learning experiences provide young children with opportunities to enjoy exploring their own interests and growing capabilities. Strong early learning also positively predicts well-being across a range of indicators in adulthood, including general well-being, physical and mental health, educational attainment and employment. Areas of early learning that are of particular importance include: language and literacy; numeracy and other non-verbal cognitive skills; self-regulation; emotional health, social well-being and social and emotional skills. These domains are interrelated, meaning that holistic assessment of a range of skills is necessary to understand well-being in early childhood and its implications for the future. Assessment of early learning provides an essential opportunity to reflect on whether governments, communities, schools and families are supporting the power and promise of early learning as intended.

This paper examines student exposure to school shootings in the United States since the 1999 shooting at Columbine High School. We analyze shootings that occurred during school hours on a school day and resulted in a death. These shootings are likely to be uniformly reported and have a greater potential to cause harm -- either directly or indirectly -- to enrolled students. We measure the number and characteristics of children who were exposed to them, along with measures of the economic and social environment in which these shootings occur. We distinguish between indiscriminate shootings, suicides, personal attacks and crime-related shootings. The primary finding of our analysis is the importance of separating these types of shootings. Indiscriminate shootings and suicides more commonly affect white students, schools in more rural locations, and those in locations where incomes are higher. The opposite geographic and socioeconomic patterns are apparent for personal attacks and crime-related shootings. Analyses that ignore these distinctions or focus on a particular type may provide a misleading impression of the nature of school shootings. Policy discussions regarding approaches to reducing school shootings should take these distinctions into account.

Schools face important policy tradeoffs in monitoring and managing student behavior. Strict discipline policies may stigmatize suspended students and expose them to the criminal justice system at a young age. On the other hand, strict discipline acts as a deterrent and limits harmful spillovers of misbehavior onto other students. This paper estimates the net impact of school discipline on student achievement, educational attainment and adult criminal activity. Using exogenous variation in school assignment caused by a large and sudden boundary change and a supplementary design based on principal switches, we show that schools with higher suspension rates have substantial negative long-run impacts. Students assigned to a school that has a one standard deviation higher suspension rate are 15 to 20 percent more likely to be arrested and incarcerated as adults. We also find negative impacts on educational attainment. The negative impacts of attending a high suspension school are largest for males and minorities.

We build and estimate a dynamic model of teenagers' choices of schooling and crime, incorporating four factors that may contribute to the different routes taken by different teenagers: heterogeneous endowments, unequal opportunities, uncertainties about one's own ability, and contemporaneous shocks. We estimate the model using administrative panel data from Chile that link school records with juvenile criminal records. Counterfactual policy experiments suggest that, for teenagers with disadvantaged backgrounds, interventions that combine mild improvement in their schooling opportunities with free tuition (by adding 22 USD per enrollee-year to the existing voucher) would lead to an 11% decrease in the fraction of those ever arrested by age 18 and a 17% increase in the fraction of those consistently enrolled throughout primary and secondary education. [This report was financially supported by Centre for Social Conflict and Cohesion Studies (CONICYT/ FONDAP/15130009) and partially supported by the supercomputing infrastructure of the NLHPC (ECM-02).]

Using value-added models, we find that high schools impact students' self-reported socioemotional development (SED) by enhancing social well-being and promoting hard work. Conditional on schools' test score impacts, schools that improve SED reduce school-based arrests, and increase high-school completion, college-going, and college persistence. Schools that improve social well-being have larger effects on attendance and behavioral infractions in high school, while those that promote hard work have larger effects on GPA. Importantly, school SED value-added is more predictive of school impacts on longer-run outcomes than school test-score value-added. As such, for the longer-run outcomes, using both SED and test score value-added more than doubles the variance of the explained school effect relative to using test score value-added alone. Results suggest that adolescence can be a formative period for socioemotional growth, high-school impacts on SED can be captured using self-report surveys, and SED can be fostered by schools to improve longer-run outcomes. These findings are robust to tests for plausible forms of selection.

We present the results of three large-scale randomized controlled trials (RCTs) carried out in Chicago, testing interventions to reduce crime and dropout by changing the decision-making of economically disadvantaged youth. We study a program called Becoming a Man (BAM), developed by the non-profit Youth Guidance, in two RCTs implemented in 2009-10 and 2013-15. In the two studies participation in the program reduced total arrests during the intervention period by 28-35%, reduced violent-crime arrests by 45-50%, improved school engagement, and in the first study where we have follow-up data, increased graduation rates by 12-19%. The third RCT tested a program with partially overlapping components carried out in the Cook County Juvenile Temporary Detention Center (JTDC), which reduced readmission rates to the facility by 21%. These large behavioral responses combined with modest program costs imply benefit-cost ratios for these interventions from 5-to-1 up to 30-to-1 or more. Our data on mechanisms are not ideal, but we find no positive evidence that these effects are due to changes in emotional intelligence or social skills, self-control or "grit," or a generic mentoring effect. We find suggestive support for the hypothesis that the programs work by helping youth slow down and reflect on whether their automatic thoughts and behaviors are well suited to the situation they are in, or whether the situation could be construed differently. [A full list of sponsors of this project can be found on the NBER web site: http://www.nber.org/papers/w21178.ack.]

This paper examines how elementary-aged peers affect cognitive and non-cognitive outcomes from adolescence to adulthood. We identify effects by exploiting within-school and within-neighborhood variation in the proportion of peers with an arrested parent. Results indicate exposure to these peers reduces achievement and increases antisocial behavior during middle and high school. More importantly, we estimate that a five percentage point increase in school and neighborhood crime-prone peers increases arrest rates at age 19-21 by 6.5 and 2.6 percent, respectively. Additional evidence suggests these effects are due to attending school with crime-prone peers, rather than living in the same neighborhood.

High school vocational education has a controversial history in the United States, largely due to a perceived tradeoff between teaching readily deployable occupational skills versus shunting mostly disadvantaged students away from the educational and career flexibility afforded by general academic courses. We study the effects of a nationwide high school reform in Norway that aimed to move beyond this tradeoff. Reform 94, implemented in one step in the fall of 1994, integrated more general education into the vocational track, offered vocational students a pathway to college through a supplementary semester of academic courses, and sought to improve the quality of the vocational track through greater access to apprenticeships. We identify the impacts of the reform through a difference-in-discontinuity research design, comparing students born just before and after the reform's birthdate eligibility cutoff to students born around the same cutoff in placebo years. Linking multiple administrative registries covering the entire Norwegian population, we find that the reform substantially increased initial enrollment in the vocational track, but with different subsequent outcomes for different groups. More men complete the vocational track at the expense of academic diplomas, but this has no detectable impact on college-going and leads to reduced criminal activity and higher earnings in adulthood, especially among disadvantaged men. For disadvantaged women, the initial surge in vocational enrollment leads to fewer high school dropouts and more vocational degrees with the college-prep supplement, and hence an increase in the share of college-eligible women; however, this translates into only small and insignificant increases in college completion and adult earnings. We show that men overwhelmingly pursue vocational education in higher-paying skilled trade fields, while women almost exclusively pursue vocational education in lower-paying service-based fields, which helps in interpreting some of these results. Overall, the reform succeeded at improving social mobility, particularly among men, but it somewhat exacerbated the gender gap in adult earnings.