A Comprehensive Evaluation of a Drug Market Intervention Training Cohort in Roanoke, Virginia; Jacksonville, Florida; and Guntersville, Alabama, 2011-2013. (ICPSR 36322)

Time Period(s) View help for Time Period(s)

Date of Collection View help for Date of Collection

Data Collection Notes View help for Data Collection Notes

Study Purpose View help for Study Purpose

The purpose of the study was to follow the progress of the seven sites that were trained on the DMI program. Specifically:

1. How the program was implemented.
2. How the program affected the targeted drug market.
3. Whether the program affected crime and disorder.
4. Whether the program improved police-community relations.
5. How much the program cost.

Study Design View help for Study Design

The Drug Market Intervention (DMI) program involves five phases (1) planning; (2) targeting the drug market; (3) working with the community; (4) preparing for the call-in (the call-in is a large community meeting, generally led by a high-level law enforcement representative, where offenders are confronted with their illegal and antisocial behavior); and (5) post-call-in enforcement and community building.

Seven sites applied for and participated in the training: Montgomery County, Maryland; Flint, Michigan; Guntersville, Alabama; Lake County, Indiana; Jacksonville, Florida; New Orleans, Louisiana; and Roanoke, Virginia. Four sites Montgomery County, Flint, Guntersville, and Roanoke implemented DMI and did a call-in within a year of the initiative's start. Roanoke conducted two call-ins and Flint conducted at least three call-ins. Three sites Lake County, Jacksonville, and New Orleans did not implement the program due to political, budgetary, and staff issues.

The study carried out the following activities to evaluate the Drug Market Intervention (DMI) program :

1. Program implementation and fidelity assessments.
2. Interviews with program staff, key informants, hard drug users, and program experts.
3. Focus groups with community members at three, six, and fifteen months post call-in.
4. Community surveys pre and post intervention.
5. Survey of hard drug users in a neighborhood targeted by the DMI.
6. Quasi-experimental crime analyses using administrative data.
7. Estimation of program costs using a micro-costing approach.
8. Development of a new method for estimating how a reduction in crime affects the cost of policing and court services.

Sample View help for Sample

Drug Market Activity datasets :

Respondent driven sampling (RDS) was used to recruit a regional sample of drug users 18 years or older in Jacksonville, Florida between August and November, 2011 (n=203) and in Roanoke, Virginia between May and July, 2012 (n=212). Jacksonville did not fully implement DMI so only a pre-intervention survey was carried out and due to timing issues only a post-intervention survey was carried out in Roanoke.

Community Perceptions Survey datasets :

The target population for the community survey was non-institutionalized persons age 18 and older, living in households in targeted areas of Guntersville, Alabama and Jacksonville, Florida with households drawn from the US Postal Service Computerized Delivery Sequence File.

Community Pre-Survey: On November 11, 2011 2,729 households in Guntersville, and 1,605 households in Jacksonville were mailed a survey packet. Reminder post cards where posted five days after initial mailing. On November 23, 2011, a second mailing was sent to all non-responding households. For non-responsive households for which a telephone number could be obtained, telephone interviews were conducted between November 30, 2011 to December 18, 2011. This yielded n=1,008 adult interviews in Guntersville, Alabama, and n=455 interviews in Jacksonville, Florida.

Community Post-Survey: On November 5, 2012 2,729 selected households in Guntersville, Alabama were mailed a survey packet. Reminder post cards where posted four days after initial mailing. On November 30, 2012, a second mailing was sent to all non-responding households. For non-responsive households for which a telephone number could be obtained, telephone interviews were conducted between December 12, 2012 to January 10, 2013. This yielded n=691 adult interviews in Guntersville, Alabama. Out of the 691 households responding to the Community Post-Survey, 480 also responded to the Community Pre-Survey.

Time Method View help for Time Method

Universe View help for Universe

Drug Market Activity datasets : Adult users of hard drugs in the past month before the survey, who lived in Roanoke, Virginia and Jacksonville, Florida in 2011 and 2012.

Community Perceptions Survey datasets : Adult community members living near DMI target areas in Jacksonville, Florida and Guntersville, Alabama in 2011 and 2012.

Unit(s) of Observation View help for Unit(s) of Observation

Data Source View help for Data Source

Data Type(s) View help for Data Type(s)

Mode of Data Collection View help for Mode of Data Collection

Description of Variables View help for Description of Variables

This study is composed of the following seven datasets.

Drug Market Activity datasets :

The Drug Market Activity datasets were sampled using the Respondent Driven Sampling (RDS) method. Interviewees were given coupons to recruit other prospective interviewees for the survey.

RDS Main Survey Jacksonville dataset (n=203) contains the following 489 variables:

• Interviewee's demographics, employment, education, household, residence, medical care history, alcohol/drug and mental health treatment history, interactions with law enforcement (incidents of police brutality, stopping and questioning, running from the police), perceptions on how race affects police interactions, arrests and charges details, and opinions and experiences on crime victimization.
• Interviewee's detailed alcohol and drug (illegal and prescription) usage history, drug purchase locations and times, drug transaction methods, amounts paid and method of payment, drug availability and quality for purchase, redistribution of purchases, and reasons for purchasing at a specific location.
• Interviewee's own survey respondent coupon number, survey coupon numbers given to the interviewee to distribute, knowledge about other drug users, relationship and drug use knowledge of persons who they have transacted coupons with, and RDS weights for gender and drug use.

RDS Follow up Survey Jacksonville dataset (n=81) contains 83 variables on the interviewee's distribution of the coupons, such as demographics of people who accepted/declined coupons, relationships and level of interaction with people who accepted/declined coupons, reasons for accepting/declining coupons, and drug habits of people who accepted/declined coupons. Also, contains variables on the interviewee's opinions and experiences on crime victimization.

RDS Main Survey Roanoke dataset (n=212) contains the following 922 variables:

• Interviewee's demographics, employment, education, household, residence, medical care history, alcohol/drug and mental health treatment history, interactions with law enforcement (incidents of police brutality, stopping and questioning, running from the police), perceptions on how race affects police interactions, arrests and charges details, and opinions and experiences on crime victimization.
• Interviewee's detailed alcohol and drug (illegal and prescription) usage history, drug purchase locations and times, drug transaction methods, amounts paid and method of payment, drug availability and quality for purchase, redistribution of purchases, reasons for purchasing at a specific location, alternative drug usage when preferred drug was unavailable or too expensive, noticed changes in drugs market in the past 30 days and their opinions on the change, and changes in their drug use and its reasons in the past 30 days.
• Interviewee's own survey respondent coupon number, survey coupon numbers given to the interviewee to distribute, knowledge about other drug users, relationship and drug use knowledge of persons who they have transacted coupons with, and RDS weights.

RDS Follow up Survey Roanoke dataset (n=59) contains 83 variables on the interviewee's distribution of the coupons, such as demographics of people who accepted/declined coupons, relationships and level of interaction with people who accepted/declined coupons, reasons for accepting/declining coupons, and drug habits of people who accepted/declined coupons. Also, contains variables on the interviewee's opinions and experiences on crime victimization.

Community Perceptions Survey datasets:

The Jacksonville (n=455) and Guntersville (n=1008) Community Pre-Survey datasets contain 104 variables representing survey questions of residents residing in the neighborhoods. Variables included are :

• Demographic information: Variables identifying race, gender, age, household composition, time lived in the neighborhood, employment situation, and level of education.
• Neighborhood information: Variables identifying perceptions on the safety of the neighborhood, census block, definition of a neighborhood, types and severity of problems in the neighborhood, frequency of observed drug transactions and usage, changes in drug activity and crime in the past six months, and opinions on if problems are caused by residents, non-residents, or something else. Also, variables describing resident's perceptions of their neighborhood, such as to what degree they are a close knit community, there are adults in the neighborhood that children can look up to, and parks and playgrounds closest to where they live are safe at night.
• Neighborhood policing information: Variables describing resident's perception of the police, such as their effectiveness, fairness, trustworthiness, and whether people's race and ethnic backgrounds affects their policing.

Community Post-Survey Guntersville (n=691) dataset contains 150 variables representing survey questions of residents residing in the neighborhood. Variables included are:

• Demographic information: Variables identifying race, gender, age, household composition, time lived in the neighborhood, employment situation, and level of education.
• Neighborhood information: Variables identifying perceptions on the safety of the neighborhood, census block, definition of a neighborhood, types and severity of problems in the neighborhood, frequency of observed drug transactions and usage, changes in drug activity and crime in the past year, and opinions on if problems are caused by residents, non-residents, or something else. Also, variables describing resident's perceptions of their neighborhood, such as to what degree they are a close knit community, there are adults in the neighborhood that children can look up to, and parks and playgrounds closest to where they live are safe at night.
• Neighborhood policing information: Variables describing resident's perception of the police, such as their effectiveness, fairness, trustworthiness, and whether people's race and ethnic backgrounds affects their policing. Also, variables on resident's perceptions on how Down with Drugs, Drug Court, and Second Chance programs have affected drug activity and crime in their neighborhoods.

Response Rates View help for Response Rates

Community Pre-Survey Jacksonville: 36.9%

Community Pre-Survey Guntersville: 28.3%

Community Post-Survey Guntersville: 25.3%

Presence of Common Scales View help for Presence of Common Scales

A Likert-type scale was used.

Original Release Date View help for Original Release Date

2016-09-27

Version History View help for Version History

Weight View help for Weight

Drug Market Activity datasets :

RDS weights were designed to correct for differences in respondent network size (also referred to as "degree") and transition probabilities across groups (i.e., the probability that a person will differentially recruit from groups with characteristics different from one's own, such as race, age, crack users v. non crack-users, etc.) during recruitment.

Please see the User Guide: DMI Data Guide for Public Use for full details.