Can Transferring Policing Power to Community Members Improve Legal Protection in Papua New Guinea?
- Rural population
- Police
- Citizen satisfaction
- Access to justice
- Attitudes and norms
- Crime reporting
- Community policing
- Diversifying security institutions
- Police skills training
How does the expansion of state policing powers affect the power dynamics of traditional societies? The researcher used a randomized evaluation in Papua New Guinea to study the impact of the Community Auxiliary Police (CAP), a program that devolves policing powers to carefully selected community members. Results suggest increased state presence may have widened the gap between men and women’s preferences for state versus customary authorities but did not reduce crime.
Policy issue
In contexts where state authorities are inaccessible to citizens to resolve interpersonal disputes, citizens sometimes instead rely on customary authorities, such as chiefs, religious leaders, and clan heads. However, in contrast to state courts and police, decisions made by customary authorities are typically not based on written legal code and formal sentencing guidelines but are instead presided over by a small group of people who often hold the political power in a given community. Therefore, while resolution practices of customary authorities may provide more accessible, grassroots justice, they can also reinforce existing power and social inequalities. For instance, in 2015, in almost every region of Africa, both men and women reported seeing traditional leaders as significantly more biased against women compared to the police and courts1.
Central governments have sought to enforce state legal codes in remote areas, possibly disrupting or even displacing traditional forms of justice. While state legal codes and processes may intend to provide more equitable rule than traditional authorities, in practice results can be mixed due to a variety of challenges, including the low accountability of state police to local communities. This study examined an attempt to overcome such challenges by transferring the state’s policing power to carefully selected community members in remote areas, where the centralized state police force is often inaccessible and customary authorities tend to predominate. As the state expands its capacity to address interpersonal conflict by increasing the availability of local police, how does this affect the role played by customary authorities? How does this change the local citizens’ behavior or perceptions of community police and local crime?
Context of the evaluation
Before gaining independence in 1975, the Autonomous Region of Bougainville (AROB) had a history of German and Australian colonial rule that focused heavily on extractive industries and met with only limited success in expanding state services to citizens in remote areas. After a decade of civil war sparked by conflict over copper mining and demands for independence, the minimal state services available in Bougainville towns were mostly withdrawn, and the state police forces that were left became active participants in the widespread crime and violence. In particular, women experienced especially severe violence. A 2013 report by the United Nations found that 60 percent of Bougainvillean men self-reported having sexually assaulted women, and of the 2,385 incidents of assault or property crime reported by respondents, 35 percent constituted man-on-woman crime2.
Due to this traumatic history with state police as well as the remote geography separating villages from central police stations3, citizens in Bougainville do not often rely on the central police force when they are victim of a crime. Instead, families, clans, and bikmen (“big men”) or “chiefs” play a large role in maintaining social order through dispute arbitration. Often, disputes are resolved through in-person meetings and an exchange of material goods as compensation for any wrongdoing. Anecdotal evidence, however, suggests that compensations have not been commensurate with the severity of incidences and outcomes may have been biased against women. Criminal incidences are also rarely reported; fewer than one-quarter of incidences measured in this evaluation were reported to either state or customary authorities.
The Community Auxiliary Police (CAP) program was established in 2005, with support from the New Zealand Police, to provide remote villages with community police officers to serve within the villages they lived. The CAP program aims to deliver legal protection to vulnerable or disadvantaged citizens in remote communities, where customary solutions to problems of social order tended to dominate. The CAP program transfers policing powers to well-respected individuals who are not directly related to local chiefs and accountable to their communities. The New Zealand Police helps recruit, manage, and train CAP officers. Countering gender-based violence is at the center of CAP efforts4, including an emphasis on hiring women and GBV-focused trainings. CAP officers learn to gather evidence for prosecution of gender-based crimes, and sociological training on gender-based power inequalities. Prior to start of this evaluation, CAP officers served in 350 of Bougainville’s roughly 2,000 villages.
Details of the intervention
In partnership with the Bougainville and New Zealand Police, the researcher used a randomized evaluation to study the impact of the Community Auxiliary Police (CAP) program on citizen behaviors, perceptions, and attitudes towards state and customary authorities.
The initial vetting process for new CAP applicants involved an intensive interview and literacy testing process to determine eligibility. After this preliminary selection process, 45 candidates from 39 different villages remained between which the police recruiters were indifferent. The researcher worked with the police to design a lottery system to randomly recruit the remaining 17 CAP officers. Selected recruits worked in their home villages. Thus, the villages whose candidates were randomly selected for recruitment formed the intervention group (17 villages), while the remaining villages whose candidates were not selected formed the comparison group (22 villages). In intervention villages that received CAP, a uniformed officer5 served full-time in a policing role, conducting arrests, levying fines, and investigating crimes.
The researcher conducted surveys with 1,383 men and women approximately eight months after the CAP officers had been working in the intervention villages to examine the impact on changes in behavior and attitudes around police and conflict resolution. They collected information about criminal incidences, including reporting and resolution, pertaining to all members of the household of the respondent, as well as about perceptions around crime in the community.
In addition to receiving ethical review and approvals from an institutional review board, the researcher made efforts to address and account for ethical questions by asking victimization questions in a way that made it impossible to identify perpetrators or third party victims beyond their gender and membership in the household – no individuals were ever personally identified, unless the respondent chose to identify as a victim, and in those cases the questions did not record the identity of the alleged perpetrator. In light of high gender-based violence rates in the region, women enumerators were trained to carry a walkie talkie and always travel with at least one other enumerator. Additionally, some areas and candidates were selected outside of the lottery based on a needs assessment by the police. Due to limited recruitment spots, some areas were guaranteed not to receive recruits, regardless of the lottery. Finally, the lottery was conducted separately for different regions to ensure all regions received a proportional share of the recruits. Changes in candidate availability after randomization created mismatches between lottery selection and hiring outcomes in five of the fifteen regions. While it does not necessarily bias the results, it does weaken the ability to detect impact. For more on the researcher’s discussion of ethical considerations, see pages 24 to 26.
Results and policy lessons
Men and women of Bougainville perceived the intervention and quality of community policing differently. The presence of police increased men’s support for local chiefs. Police were more likely to be involved in the resolution of conflict that women had with men. However, changes in attitudes towards both state and customary authorities did not have a detectable effect on underlying incidences of crime.
Perceptions of police intervention: Overall, women were more likely than men to report positive experiences with the police, and the presence of a CAP officer widened the pre-existing difference in how men and women perceived police treatment and quality. In villages where CAP officers were present, women were 7.3 percentage points (11.6 percent) more likely to believe that the police treated them in the same way as men (up from 63 percent), and 10.4 percentage points (16.3 percent) more likely to believe that their problem was taken seriously (up from 64 percent). In contrast, while men were overall less likely than women to agree that they were treated fairly, this gap in perceived discrimination increased in villages where CAP officers were present. Men in these villages were 28.2 percentage points less likely than women to agree that they were treated fairly compared to women (up from a 16 percentage point difference).
Attitudes towards chief involvement in dispute resolution: The presence of a CAP officer strengthened men’s support for involving the chief in dispute resolution but did not significantly impact women’s support of the chief. In villages where CAP officers were present, men’s support for customary authorities in dispute resolution increased by approximately 6.5 percentage points (11 percent) from an average of 59 percent in villages without CAP officers. This suggests that men’s experience with CAP officers led them to vest more authority in the local chief.
Effects on reported crime: Researchers found no evidence that crime decreased due to the presence of the CAP officers. It could be that CAP officers could not deter crime due to the nature of most commonly committed offenses (violence and property crime often involving alcohol), but the inability to detect an effect may also be due to the study’s limited statistical power.
Perceptions of state: Increased access to police services did not impact trust in the government, loyalty to the local chief, and knowledge about the government. The results suggest that the expansion of state authorities in remote areas did not replace customary authorities and their resolution practices. Instead, among those who may traditionally benefit from customary authorities, namely men, positive attitudes toward customary authorities increased. The presence of police may have encouraged women to report conflict with men, which in turn incentivized men to draw upon local chiefs to protect their interests. To complement these findings, in a historical analysis of over 30,000 case records from 2005 to 2009, the researcher finds that the presence of at least one female police officer increased both instances of police and chief involvement in cases of violence against women. Taken together, the findings suggest the existence of numerous alternative channels of resolving disputes may limit the ability of the state to enforce its own legal code.