Going beyond LATE

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. We derive nonparametric sharp bounds on average treatment effects with an instrumental variable (IV) and use them to evaluate the effectiveness of the Job Corps (JC) training program for disadvantaged youth. We concentrate on the population average treatment effect (ATE) and the average treatment effect on the treated (ATT), which are parameters not point identified with an IV under heterogeneous treatment effects. The main assumptions employed to bound the ATE and ATT are monotonicity in the treatment of the average outcomes of specified subpopulations, and mean dominance assumptions across the potential outcomes of these subpopulations. Importantly, the direction of the mean dominance assumptions can be informed from data, and some of our bounds do not require an outcome with bounded support. We employ these bounds to assess the effectiveness of the JC program using data from a randomized social experiment with non-compliance (a common feature of social experiments). Our empirical results indicate that the effect of JC on eligible applicants (the target population) four years after randomization is to increase weekly earnings and employment by at least $24.61 and 4.3 percentage points, respectively, and to decrease yearly dependence on public welfare benefits by at least $84.29. Furthermore, the effect of JC on participants (the treated population) is to increase weekly earnings by between $28.67 and $43.47, increase employment by between 4.9 and 9.3 percentage points, and decrease public benefits received by between $108.72 and $140.29. Our results also point to positive average effects of JC on the labor market outcomes of those individuals who decide not to enroll in JC regardless of their treatment assignment (the so-called never takers), suggesting that these individuals would indeed benefit from participating in JC. Terms of use: Documents in JEL Classification:J30, C13, C21

Improving the outcomes of youth with medical limitations: Evidence from the National Job Corps Study

“…NJCS analyses indicated that Job Corps participation led to substantial short‐run increases in receipt of education/training services and decreases in arrest rates, as well as medium‐run improvements in labor market outcomes and decreases in receipt of public benefits (Chen et al, 2018; Schochet et al, 2001). In addition, youth who participated in Job Corps for longer periods also realized larger gains in terms of postparticipation earnings (Flores et al, 2012).…”

“…Job Corps is one of the largest youth programs outside the traditional K–12 schooling system. The National Job Corps Study (NJCS), a large‐scale experimental evaluation conducted in the mid‐1990s, found that Job Corps helped improve medium‐term (but not long‐term) work and earnings outcomes for youth and helped reduce both criminal activity and receipt of certain forms of public assistance (Chen et al, 2018; Schochet et al, 2001, 2008). Further, longer participation in academic and vocational instruction in Job Corps (which averaged around 7 months for youth in the NJCS) was linked to larger increases in postparticipation earnings (Flores et al, 2012).…”

Discordant relaxations of misspecified models

“…It is worth noting that discrete relaxations of misspecified models have been entertained in various existing papers; see, for instance, Manski and Pepper (2000, 2009), Blundell et al (2007), Kreider, Pepper, Gundersen, and Jolliffe (2012), Chen, Flores, and Flores‐Lagunes (2018), Kédagni (2023), Mourifié, Henry, and Méango (2020), among many others. In these papers, when the initial model is too stringent, they suggested alternative weaker assumptions that are believed to be more compatible with the empirical application under scrutiny and for which the identified/outer set is not empty.…”