Download StudyGeneral

Study Overview

Evaluating the Impact of Road Safety Initiatives in Liberia
Study is 3ie funded:
Study ID:
Initial Registration Date:
Last Update Date:
Study Status:
In Development

The proposed impact evaluation (IE) seeks to understand the incentives of the different actors on the road (drivers, passengers, and police officers) as well as the role played by the environment (road design) and, subsequently, to learn which road safety policies would be the most effective at reducing road traffic crashes in Liberia.  This impact evaluation will assess the impact of different interventions targeted to drivers and passengers and lay the groundwork for future evaluation of speed enforcement and road design interventions. Because careless driving is one of the leading causes of reported crashes, we focus on interventions intended to nudge drivers’ behavior.  First, we will test the behavior of drivers and their incentives to adopt safe behavior on the road through driver education and compensation. This intervention allows us to test whether drivers know the behavior they should adopt to avoid crashes but lack incentives to do so, or if they are unable to adopt a safe behavior because they do not know which behaviors lead to higher probabilities of crashes.  Second, we will test the role that passengers play in limiting the risky choices of the driver through a passenger empowerment intervention. To evaluate these interventions, the project requires the development of a road incident data management system with data collected by GPS tracking devices, speed cameras, surveys, and/or smartphones for measures of speed and driving behavior and crashes.

Registration Citation:

Bougna, Theophile, Golvine de Rochambeau, Impact evaluation of Road Safety Initiatives in Liberia

Additional Keywords:
Road Safety, Transportation, Liberia, Behavioral Microeconomic, Impact Evaluation
Secondary ID Number(s):

Principal Investigator(s)

Name of First PI:
Theophile Bougna
The World Bank Group
Name of Second PI:
Golvine de Rochambeau
Sciences Po

Study Sponsor

The world Bank Group (DIME) and FCDO
Study Sponsor Location:
United States

Research Partner

Name of Partner Institution:
Infrastructure Implementation Unit, Ministry of Public Works, Liberia
Type of Organization:
Government agency (eg., statistics office, Ministry of Health)

Intervention Overview


The impact evaluation will focus on the Suakoko Highway and the 100-km+ planned rehabilitation from Ganta to Zwedru.

  1. Driver Education and Rewards: A training program will be rolled out with the goals of educating drivers about the dangers of speeding and risky driving and providing them with a set of actions they can take to protect themselves, their passengers, and pedestrians. In addition, drivers will be randomly selected to be part of a rewards program in one of two ways: rewards conditional on adopting safe behavior, or rewards conditional on having no crashes. 
  2. Passenger Empowerment: The second component of the project will focus on how passengers can influence safe driving behavior. Passengers will be contacted at transport hubs and given information on safe driving behavior. They will also be randomly reminded of this during their trip through SMS messages, meant to mimic public awareness campaigns, or will be given a device to monitor vehicle speed in real-time during the trip.
  3. Speed Enforcement: Automated speed control, such as mobile cameras, coupled with police posts have been shown to be very effective at reducing speeds.
  4. Road Design: During the course of the study, the Liberian Ministry of Public Works plans to implement a number of physical road safety interventions along the corridor. These interventions may include row markers, signage, crash barriers, raised pavement markers, and bridge safety enhancements. 


Theory of Change:

The main and only input is the improvement and maintenance of the Suakoko Highway. Besides the increase in mobility, the other outputs are related to the increase in traffic along the highway. This theory of change will only focus on activities, intermediates, and long-term outcomes. Our theory of change is grounded by two theoretical frameworks: the framework of the theory of planned behaviour (TPB) developed by Icek Ajzen (1985) and the Behaviour change techniques. The focus is not so much on conscious decisions that involve weighing up the pros and cons of different decisions as on behaviors that are more subconscious or automatic. We are planning to test explicit attitude measures in the education and training of drivers. Since we are interested in both the short and long-term impact of the intervention, the ultimate goal is to link driving behavior with health and road safety outcomes. This IE will allow us to understand the forces at stake and the incentives of actors involved so that the most effective interventions can be identified and prioritized


Multiple Treatment Arms Evaluated?

Implementing Agency

Name of Organization:
Infrastructure Implementation Unit, Ministry of Public Works, Liberia
Type of Organization:
Public Sector, e.g. Government Agency or Ministry

Program Funder

Name of Organization:
World Bank, GRSF, and FCDO
Type of Organization:
Foreign or Multilateral Aid Agency

Intervention Timing

Intervention or Program Started at time of Registration?
Start Date:
End Date:
Evaluation Method

Evaluation Method Overview

Primary (or First) Evaluation Method:
Randomized control trial
Other (not Listed) Method:
Additional Evaluation Method (If Any):
Other (not Listed) Method:

Method Details

Details of Evaluation Approach:

The intervention will be rolled out across markets along the corridor, and the associations within each union (motorcycle-taxis, car-taxis, and trucks), and opportunities to expand the intervention to additional locations will be explored.

The main equation we want to estimate is the following:


Where Treatmentj is a variable that takes the value 1 if driver i has been selected to receive treatment j (and where different treatments correspond to the different treatment arms detailed in the past sections), and Yit is the outcome of interest. Here, the outcomes of interest are driving behavior (for example speed or sudden braking) as well as outputs related to crashes (whether they had a crash in the past or the seriousness of the crash). 

This equation will allow us to estimate the average treatment effect, as well as the intent-to-treat effect (by instrumenting the take-up of treatment by the assignment to treatment). Since the take-up of the education treatment will not be 100%, it is important to compute both effects.  Since different groups of drivers might have different effects from treatment, we will cluster standard errors at the transport-hub level or, in the case where drivers are recruited through unions, at the union level.

Outcomes (Endpoints):

Average speed of vehicles as well as maximum speed

Number of heavy brakes

Number of crashes, number of road injuries and casualties

Motorcycle driver is wearing helmet

Number of times passengers speak-up to the driver

Level of education of drivers regarding safety notions

Number of trips completed and average revenue of drivers,

Safety perception of passengers

Unit of Analysis:
Taxi (drivers and passengers)

The primary research question this project seeks to address is: what are the most effective measures at improving road safety in this context? We will focus on specific behavioral and physical road safety interventions that are tentatively planned in Liberia. The secondary questions that arise are, therefore:

1.         What is the role of the lack of driver education in explaining high crash rates?  Do drivers respond to rewards?

2.         Can passengers play a role in reducing crash rates?

3.         How does speed enforcement shape driving behavior?

4.         What is the impact of road design on safe driving behavior?

Unit of Intervention or Assignment:
Vehicle (driver and passenger behavior)
Number of Clusters in Sample:
Not available
Number of Individuals in Sample:
Size of Treatment, Control, or Comparison Subsamples:
treatment: 1,000, controls: 1,000

Supplementary Files

Analysis Plan:
Other Documents:

Outcomes Data

New surveys: 1. GPS trackers send all driving behavior information to an online serve 2. Data from speed cameras 3. Smartphone • Multiple datasets will be used to cross-reference each other. For example, the Liberia National Police reports will be cross-checked with data from GPS trackers, and data from interviews of passengers after their trip will be cross-checked from smartphone applications.
Data Already Collected?
Data Previously Used?
Data Access:
Data Obtained by the Study Researchers?
Data Approval Process:
Approval Status:

Treatment Assignment Data

Participation or Assignment Information:
Data Obtained by the Study Researchers?
Data Previously Used?
Data Access:
Data Obtained by the Study Researchers?
Data Approval Process:
Approval Status:

Data Analysis

Data Analysis Status:

Study Materials

Upload Study Materials:

Registration Category

Registration Category:
Prospective, Category 1: Data for measuring impacts have not been collected

Completion Overview

Intervention Completion Date:
Data Collection Completion Date:
Unit of Analysis:
Clusters in Final Sample:
Total Observations in Final Sample:
Size of Treatment, Control, or Comparison Subsamples:


Preliminary Report:
Preliminary Report URL:
Summary of Findings:
Paper Summary:
Paper Citation:

Data Availability

Data Availability (Primary Data):
Date of Data Availability:
Data URL or Contact:
Access procedure:

Other Materials

Survey Instrument Links or Contact:
Program Files:
Program Files Links or Contact:
External Link:
External Link Description:
Description of Changes:

Study Stopped