Tutor Lesson

Behavioral science links ideas about psychology and social relationships to practical programs concerning health, finances, collective welfare, security, and peace. Here, the central application is the prevention of political violence and conflict.

• Personality traits
• Earlier experiences
• Beliefs and expectations
• Emotional reactions
• Existing relationships and prejudices

The situation includes pressures and features of the environment that can influence what people do:

• Pressure from other people
• Stress and insecurity
• Rules and institutional processes
• How easy or difficult program access is
• The physical layout of buildings
• Whether information is available
• Chances to interact with others

Policy can usually alter a situation more directly than it can alter personality or a lifetime of experience. It might simplify enrollment, lower stress, open a path to cooperation, improve access to healthy choices, or shape a space so that people meet rather than remain separated.

Because policies are meant to influence people, designers must understand how people respond to fear, mistrust, expectations, and difficult procedures. A logically appealing policy can fail when those responses are ignored.

1. Identify the behavior or relationship that should change.
2. Find the psychological or social process blocking that change.
3. Choose a situation that could address the process.
4. Decide what evidence would indicate success.

Consider a community where mistrust and rumors sustain conflict. A program might connect respected leaders, organize a shared project for groups that seldom meet, or make visible cooperation demonstrate that working together is socially acceptable.

• Program design uses theory to select the situation that will be created.
• Program evaluation checks whether that situation caused the predicted changes.

A program should do more than pursue a worthwhile aim. It can also test whether a proposed explanation works in that context and serve as a pilot for future policy.

Policies remain informed guesses until evidence shows what they produced. A sensible intervention can still fail for several reasons:

• Its theory may not fit local conditions.
• The intervention may end too soon.
• Participants may receive too little exposure.
• External political events may interfere.
• Participants may change without the wider community changing.
• The intended results may be measured badly.

Familiarity is the proposed mechanism. Knowing someone from another group may revise beliefs about that person and perhaps about the broader group.

• Negative stereotypes
• Prejudice
• Fear
• Social distance
• Discriminatory conduct

Programs have applied this idea to boundaries involving religion, ethnicity, nationality, political identity, occupation, and histories of conflict. Their practical task is to let divided groups meet, talk, and work together.

Contact has appeared stronger across differences without long histories of severe hostility; disability-related divisions were one example in the research discussed. Results for religious and conflict-based divisions were generally positive but smaller because deep violence and mistrust are difficult to transform through a brief intervention.

• Attitudes capture what people believe or feel.
• Behaviors capture what people actually do.

An evaluation should measure both. A program may produce valuable behavioral gains even when attitudinal change is not statistically significant.

Contact also may stay inside its structured setting. In an Iraq study, cooperation between Muslim and Christian young men on a soccer field did not transfer to other interactions.

• The period of contact may have been too brief.
• Interaction opportunities elsewhere may have been scarce.
• Community norms may have remained unchanged.
• Personal change may not have influenced other people.
• Wider transfer may require attitudinal change.
• The program may have been too detached from everyday life.

The study took place in Nigeria's Middle Belt, where conflict joined livelihood divisions with religious ones. Farmers were mostly Christian and pastoralist herders mostly Muslim, though neither category was absolute. At the time, roughly 7,000 people had died over five years, and the estimated economic cost was $13 billion.

1. Train leaders in peaceful conflict resolution.
2. Create joint farmer–herder committees.

Each committee brought together about eight leaders from each side to complete projects benefiting both communities.

• Views of the other group would improve.
• People would experience greater security.
• Peaceful conduct toward the other group would increase.

Fifteen eligible communities were identified, but resources covered ten. Random assignment gave ten communities the intervention and left five as initial comparison sites. Every site paired a farming community with a herding community, and about one hundred residents per community were surveyed.

• Committee members with direct exposure
• Other residents of intervention communities
• People living in comparison communities

These exposure levels made it possible to examine direct effects as well as community spillover.

• Surveys covering attitudes, trust, security, and reported actions
• A public-goods game intended to capture cooperation
• Direct counts at markets and social gatherings

Local researchers counted farmers and herders at markets and events. These observations captured behavior outside committee activities instead of relying solely on self-reports.

Stated willingness to interact changed little, but observation told a different story. Because markets were located in farming communities, farmer attendance stayed fairly stable. Pastoralists continued visiting markets in intervention areas while their attendance declined in comparison areas.

The intervention therefore appears to have preserved ordinary contact during surrounding violence. Attendance at some social events also improved, although contact did not rise in every form.

Residents of intervention communities felt substantially safer. Perceived insecurity matters because feeling unsafe can discourage travel, market attendance, social contact, and cooperation across groups.

Comfort with and trust in the other group moved positively, though not dramatically. Near the final survey, Benue State adopted a no-grazing law that harmed pastoralists more than farmers and was followed by greater violence. Intervention sites improved modestly, whereas comparison sites worsened.

Committee participants changed most. Other intervention-community residents moved less, yet more than comparison residents. This gradient is consistent with partial community spillover.

Because leaders cooperated visibly on shared projects, they may have signaled that farmer–herder cooperation was both possible and accepted. Public action can change norms even when private beliefs move slowly.

• Only a small number of communities participated, reducing precision and limiting generalization to other farmer–herder conflicts.
• Committee volunteers selected themselves after individual random assignment proved unacceptable; they may have been more cooperative from the outset.
• The no-grazing law triggered violence and displacement shortly before the final survey.
• Some pastoralists were later surveyed elsewhere, but those located may have differed from those who could not be found, making the final pastoralist sample partly one of convenience.

Local staff and organizations were central to implementation and research. A local partner led data collection and helped develop and pilot the survey, while program managers checked that measures and procedures were acceptable to participating communities.

Several years of work and established relationships in Nigeria also made the structured RCT more feasible and sensitive to local conditions. Rigorous design and local involvement can reinforce each other: local knowledge improves an evaluation's quality, interpretation, and acceptability.

The first study attributed the main effects to the joint committees, yet participants felt safer without reporting improved dispute resolution. That puzzle motivated a later RCT isolating mediation training.

• 88 communities in total
• 44 communities receiving mediation training
• 44 comparison communities
• About 2,000 baseline respondents
• About 4,000 midline respondents
• A final measurement roughly three years later

Violence-related measures improved at both midline and endline, with some becoming slightly better over time. Economic interaction had risen by midline and rose much more by endline. Trust was not significantly better at midline but was significantly higher at endline.

1. Mediation lowers violence.
2. Lower violence makes economic exchange safer.
3. Economic exchange creates opportunities to cooperate.
4. Successful exchanges strengthen trust.
5. More trust supports further peaceful interaction.

A short evaluation can miss a sequence in which conduct changes first, security follows, interaction expands, and trust or economic benefits emerge later.

• Evaluation funding often stops early.
• Academic incentives favor first studies over later checks.
• Returning to communities can be difficult.
• Political circumstances can shift.
• Participants may relocate.
• Later work may merely show that an earlier effect endured or vanished.

Women served on committees in the first Nigeria study and received mediation training in the second, yet the studies did not detect different community-level effects by gender. That finding does not make gender irrelevant.

In some contexts, women have less public access or need household permission to participate. In certain Muslim households, researchers interviewed a husband before interviewing a woman, which could shape women's participation and exposure.

• Who is able to participate
• How much public interaction is available
• Whether household approval is required
• Which mediation roles society accepts
• Whether results transfer to another setting

Familiarity and trust may operate similarly across genders, while social conditions determine whether women and men have equal chances to experience contact.

Some research placed refugees and host-community members in the same classes or training, but physical proximity does not establish that meaningful interaction occurred.

A study of internally displaced people and host communities in Afghanistan found no improved relations. One possible reason was that participants already shared a similar background and lacked a pronounced social divide.

Designers should therefore assess the initial relationship: Is there a meaningful division to address, and are conditions safe enough for contact to help?

A Guatemala City gang-violence program used community-driven development. Its theory proposed that collective decisions and shared projects would increase cohesion, which would then reduce violence.

The account sounded reasonable, but prior evidence did not strongly support it; much community-driven-development research has not found reliable increases in social cohesion. The program nevertheless tested whether lessons from efforts to prevent young men from joining violent extremist movements could transfer to gang violence.

• A theory that appears plausible
• A theory backed by evidence
• A theory tested where it is meant to operate

Research has little practical influence if it never reaches those who design, finance, and deliver programs. Partnerships can connect academic researchers, government agencies, legislative staff, international development organizations, local practitioners, and community groups.

• Turn findings into concrete design decisions.
• Adjust interventions to local circumstances.
• Select suitable outcomes.
• Plan evaluation from the program's beginning.
• Read results accurately.
• Choose whether to expand or modify a program.

A written synthesis can report what is known without telling an officer exactly how to act. Direct collaboration helps translate research into program choices. Evidence should shape design at the outset, not appear only after implementation is complete.

1. Both personal characteristics and situations shape behavior.
2. Policymakers can usually alter situations more directly than personalities.
3. Behavioral theory identifies settings that may promote constructive action.
4. Programs establish those settings in actual communities.
5. Evaluation checks for the predicted outcomes.
6. Research examines participant effects and wider spillover.
7. Later follow-up tests whether changes endure.
8. The findings inform subsequent programs and public policy.