Data and Code for: "Cognitive behavior therapy reduces crime and violence over 10 years: Experimental evidence"

Principal Investigator(s): View help for Principal Investigator(s) Christopher Blattman, University of Chicago. Harris School of Public Policy; Sebastian Chaskel, Instiglio; Julian C. Jamison, University of Exeter and Global Priorities Institute; Margaret Sheridan, University of North Carolina-Chapel Hill

Version: View help for Version V1

Name	File Type	Size	Last Modified
STYL_10Y_replication			10/25/2023 03:26:PM

Project Citation:

Blattman, Christopher, Chaskel, Sebastian, Jamison, Julian C. , and Sheridan, Margaret . Data and Code for: “Cognitive behavior therapy reduces crime and violence over 10 years: Experimental evidence.” Nashville, TN: American Economic Association [publisher], 2023. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2023-11-27. https://doi.org/10.3886/E184306V1

Project Description

Summary:

Data for "Cognitive behavior therapy reduces crime and violence over 10 years: Experimental evidence".

Paper Abstract:
"Several small, short-term, or non-experimental studies show that cognitive behavioral-informed interventions reduce antisocial behaviors over 1–2 years, but there is little research on persistence. We followed 999 high-risk men in Liberia 10 years after randomization into: 8 weeks of low-cost, nonspecialist-led therapy; $200 cash; both; or neither. A decade later, antisocial behaviors (such as robbery and drug-selling) fell 0.2 standard deviations from therapy alone---significantly greater than the 1-year impacts. Meanwhile, men who received therapy plus cash were 0.25 standard deviations less antisocial---similar to their 1-year results. In both cases, impacts were concentrated in men exhibiting highest baseline risk."

Funding Sources: National Science Foundation (SES 191939); Wellspring Philanthropic Fund

Scope of Project

Subject Terms: cognitive behavioral therapy; cash transfers; crime; Randomized Control Trial; violence; mental health; Africa; field experiments

JEL Classification:
      D83 Search; Learning; Information and Knowledge; Communication; Belief; Unawareness
      K42 Illegal Behavior and the Enforcement of Law
      O15 Economic Development: Human Resources; Human Development; Income Distribution; Migration
      O17 Formal and Informal Sectors; Shadow Economy; Institutional Arrangements

Geographic Coverage: Liberia

Time Period(s): 2012 – 2022

Collection Date(s): 2012 – 2022

Methodology

Response Rate:

All data are self-reported survey outcomes. As one of the world's poorest countries, Liberia does not have administrative data on arrests, crimes, or violence. The main advantage of this strategy is unusually rich outcome measures. A disadvantage is potential bias in self-reported results, discussed below.

We attempted to survey each subject 7 times: (i) at baseline; (ii & iii) 2 and 5 weeks after the grants; (iv & v) 12 and 13 months after grants, and roughly (vi & vii) 114 and 115 months after grants. We ran pairs of surveys to reduce noise in outcomes with low autocorrelation, such as earnings or criminal activity. We average these pairs into "1-month'', "1-year'', and "10-year'' outcomes.

Subjects typically had no fixed address and lived under aliases or clandestinely. By collecting social network and contact information, and through intensive tracking, we located most surviving respondents. After 1 year, we interviewed 947 of the 999 (95%). After 10 years, we surveyed 833 (83% of the full sample), plus the friends or relatives of 103 who had died. Therefore we have outcomes for 93% of known survivors. Of the remaining 63: 14 refused the survey; 9 were unreachable (out of country); 7 were imprisoned; and 33 could not be found.

Sampling:

The study recruited 10 years ago 999 men actively involved in crime and violence. They were age 25 on average. We focused on five mixed-income residential neighborhoods in Monrovia with large populations of high-risk men.

All recruitment was handled by NEPI. In each neighborhood, certain places, groups, and professions had reputations for crime and violence involvement, and recruiters targeted these locations and people.

We tried to minimize general equilibrium effects and spillovers between treatment and control group members. We worked in neighborhoods with tens of thousands of residents, recruiting less than 1 percent of adult men. NEPI recruiters were also instructed to approach just 1 out of every 7-10 potentially high-risk subjects they identified on the street. This avoided more than 10 percent of high risk men being treated in a neighborhood.

Related Publications

This study is un-published. See below for other available versions.

Published Versions

Export Metadata

Dublin Core

DDI 2.5

Report a Problem

Found a serious problem with the data, such as disclosure risk or copyrighted content? Let us know.

This material is distributed exactly as it arrived from the data depositor. ICPSR has not checked or processed this material. Users should consult the investigator(s) if further information is desired.