HOTSPOTS- Sucre
Last registered on August 08, 2014

Pre-Trial

Trial Information
General Information
Title
HOTSPOTS- Sucre
RCT ID
AEARCTR-0000469
Initial registration date
August 08, 2014
Last updated
August 08, 2014 9:30 AM EDT
Location(s)
Primary Investigator
Affiliation
CAF - Development Bank of Latin America
Other Primary Investigator(s)
PI Affiliation
Stanford University
PI Affiliation
CAF/IESA - Primary invertigator
Additional Trial Information
Status
On going
Start date
2013-08-01
End date
2015-03-31
Secondary IDs
Abstract
A number of studies show that crime is often concentrated in a reduced number of high crime areas, furthermore, that it is highly concentrated in a small fraction of street segments, usually termed crime hotspots. Such concentration suggests that focusing police patrol time in these hotspots should be an effective strategy for preventing crime, and a number of experimental studies in the US and the UK have provided evidence to this effect. In an attempt to tackle violent criminal activity in the municipality of Sucre in Caracas, Venezuela, the mayor’s office and the municipal police decided to run a randomized control trial to evaluate the impact of a hotspots policing strategy on the incidence of homicide. Police are required to deliver four 15 minute visits to each hotspot over every 24 hour period, sometimes simply under a guardianship role and other times stopping and searching. The program is currently being implemented.
External Link(s)
Registration Citation
Citation
CAF, Pilar, Dorothy Kronick and Daniel Ortega. 2014. "HOTSPOTS- Sucre." AEA RCT Registry. August 08. https://www.socialscienceregistry.org/trials/469/history/2336
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Interventions
Intervention(s)
Sucre is one of four municipalities of Venezuelan capital city, Caracas. Roughly 1 million people live there. According to the latest data available, 80% of homicides in Sucre occur in 6% of its street segments. Using geo-referenced homicide data, we identified 113 hotspots in July 2013: street segments with at least two homicides in 2012 and that met one of the following criteria:

• Always hot (45): segments that were among the 160 deadliest segments in each of the previous three years.
• Heating (8): segments among the 160 deadliest in t (June 2012 – June 2013), but not in t-1 (June 2011 – June 2012), with at least 1 homicide in t-1.
• Cooling (36): segments among the 160 deadliest in t-1, but not in t, with some homicides in t.
• Newly hot (23): segments among the 100 deadliest in t, but that could not be included in any of the other three categories and with at least two homicides in t.

Field visits to the identified segments with the local patrol police led to a definition of 92 relevant “intervention areas”, or policy –hotspots, mostly as a result of two hot street segments being too close to one another to represent separate hotspots, or corners where it made sense to park a patrol car and from where two hot segments were in straight line of sight. 46 spots were randomly assigned to treatment and 46 to control (see below). Two months into the implementation of the program, there were significant compliance problems (police were not delivering the required dosage) so the number of treatment hotspots was reduced to 34.
Police patrols are expected to make visits using an unpredictable pattern. Also, during each visit, police are expected to stop and search, to guard and to build cooperative relationships with the community.

Police patrols were equipped with GPS so that their movements can be monitored. A geofence around each hotspot helped precisely clock the time of entry and exit to each hotspot. Reports on police compliance are produced three times a day and sent to the patrol supervisors, and a weekly report on the implementation of the program is sent to the supervising authorities. In order to get police personnel to conform with these protocols, the administrators established a system of incentives consisting of in-kind payments to the 20 officers with the best performance. Specifically, performance is being evaluated along a number of dimensions: compliance with time and number of visits to the hotspots, variability in the hours at which visits occur and that their GPS equipment has not been tampered with or damaged
Intervention Start Date
2013-09-01
Intervention End Date
2014-12-31
Primary Outcomes
Primary Outcomes (end points)
Number of homicides at the hotspot level
Primary Outcomes (explanation)
Homicides are considered to have occured within a hotspot if the geographical coordinates recorded where the deseased person was found fall within the geofence defined for each street segment
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
Using geo-referenced homicide data as well as a visual inspection of the field along with the local police, 92 intervention areas were identified. To evaluate whether targeted and systematic patrol has an effect on crime levels, half of these hotspots were randomly assigned to the treatment group and half to the control group. Random assignment was performed by first pairing spots according to homicide levels and trends, population density and a socioeconomic index, and then randomizing within each pair. Geo-referenced homicide data before, during and after the program will be used to measure impacts. 12 of the original 46 treatment hotspots were dropped from the intervention due to logistical constraints, so the study currently has 34 treated and 34 control hotspots.
Experimental Design Details
Using geo-referenced homicide data as well as a visual inspection of the field along with the local police, 92 intervention areas were identified. To evaluate whether targeted and systematic patrol has an effect on crime levels, half of these hotspots were randomly assigned to the treatment group and half to the control group. Random assignment was performed by first pairing spots according to homicide levels and trends, population density and a socioeconomic index, and then randomizing within each pair. Geo-referenced homicide data before, during and after the program will be used to measure impacts. 12 of the original 46 treatment hotspots were dropped from the intervention due to logistical constraints, so the study currently has 34 treated and 34 control hotspots.
Randomization Method
For each pair, randomization was done in an office with a computer
Randomization Unit
Hotspot
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
92 hotspots

Sample size: planned number of observations
92 hotspots
Sample size (or number of clusters) by treatment arms
46 hotspots in treatment group and 46 hotspots in control group
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
IRB Approval Date
IRB Approval Number
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
No
Is data collection complete?
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers