International Gender Balancing Reforms in Postconflict Countries: Lab-in-the-Field Evidence from the Liberian National Police Force

Although the study investigated several aspects of police work, only the effects on unit cohesion and the participation and influence of women were statistically significant.

• Unit cohesion in terms of task cohesion (defined as the similarity of individual and group preferences in the experimental tasks) increases as the number of women in the group increases, but unit cohesion in terms of social cohesion does not:
  • Task cohesion increased as the number of women in the group increased, although it was only statistically significant for the all-women group (17% alignment).
  • Social cohesion, defined as collegiality shown during the experimental tasks, was not affected by the number of women in the group.
• There was no evidence that the presence and participation of women affected the group’s ability to detect sexual and gender-based violence in the crime scene experimental task.
• There was no evidence that the presence and participation of women affected the group’s beliefs on gender norms, but it did affect women´s participation and influence in the experimental tasks.
  • Adding more women increased the participation of both men and women (compared to groups with no women, officers in 4-women groups participated 38.5% more and women in all women groups 48% more).
  • Adding more women increased the influence of women in the group.

Whereas having more women increased unit task cohesion and the participation and influence of women, there is no evidence to suggest that adding more women increases group or individual ability to detect sexual and gender-based violence or improves men’s beliefs about women’s role in the security force in the short term. Competence as a police officer was found to be a significant determinant of both task cohesion and participation in the experimental tasks, and predicted the correct identification of sexual and gender-based violence in the crime scene task. Future efforts should focus not only on gender balancing, but also on providing specific trainings to increase the competence of police officers.

The hypotheses were tested with a lab-in-the-field experiment including 612 LNP officers divided into 102 groups of six people with varied gender composition: no women, two women, four women, and six (all) women. Each group was given five tasks and observed by a trained Liberian enumerator. In the first activity, each group was given money and asked how much money they would place into a lottery pool, both individually and as a group after reaching consensus. In the second activity, each group was given money and asked how much they would donate to another group in the room, both individually and as a group after reaching consensus. The third activity was cooperative- the group had to build a tower out of newsprint and masking tape, and the group with the largest tower won money to be split evenly among group members.

In the fourth activity, the group was shown a picture of a crime scene and told to memorize details. They then privately answered individual questions and reconvened as a group to reach consensus answers. In the fifth activity, the same picture of a crime was used. Participants guessed privately the type of crime they thought the photo depicted and why, then reconvened to reach a group consensus. Finally, participants completed an exit questionnaire to survey participant’s feelings during the team activities and their beliefs about gender norms in the LNP. The authors compared the individual answers to the group consensus in each activity, and also used data observed by the enumerators and the exit questionnaire to measure unit cohesion, operational effectiveness in detecting sexual violence, and organizational beliefs about gender norms. Control variables included demographics and subjects´ competency as police officers based on answers to cognitive score questions (basic reasoning and mathematical skills), a memory test, and answers to the crime scene activity.