Investigating The Role Of Socio-Psychological Factors In The Perpetuation Of Security Sector Violence, A Case Study Of The Colombian National Police

In post-conflict societies, police forces often remain embedded in institutional cultures shaped by militarized logics and wartime mentalities. While formal reforms may seek to democratize the police, change in practice is often slow and uneven. In Colombia, the Policía Nacional has historically operated at the blurred boundaries of military functions, particularly in regions marked by conflict, displacement, and high levels of state fragility. Following the 2016 peace agreement with the FARC-EP, and especially after the nationwide protests in 2021, the legitimacy of the police came under renewed scrutiny. The protests revealed deep tensions between law enforcement and civilian populations—particularly those perceived as marginalized or politically dissenting. Widespread accounts of excessive force, racialized violence, and impunity further eroded public trust. Despite government-led efforts to advance a police reform agenda—including changes to uniforms, discourse, and training—the transformation of police behavior remains partial and contested. This study investigates the psychological and institutional determinants of Colombian police officers’ willingness to protect civilians. It builds on Ortiz-Ayala (2025), who introduced the concept of willingness to protect in the context of post-conflict military forces and extends it to policing. We focus on how officers respond to civilians with different social and political attributes—such as political ideology, race, gender, and territorial origin—using a randomized survey experiment embedded in a nationally distributed questionnaire. Theoretically, we draw on González (2017, 2020), who conceptualizes policing in unequal democracies as fundamentally political. In such contexts, protection is selectively distributed based on perceived social hierarchies and ideologies. We complement this with insights from the Multidimensional Existential Threat (MET) model (Hirschberger et al., 2016), identity fusion theory (Swann et al., 2009; Swann et al., 2012), and research on moral disengagement (Bandura, 2002, 2016; Jackson & Gaertner, 2010), to explore how perceived threat, emotional responses, and institutional identification shape behavioral orientations. The project also incorporates recent literature on post-authoritarian policing (Blair & Morse, 2021; González, 2020), security sector reform (Tiscornia, 2019, 2023), and state violence and accountability (Trejo et al., 2018). It aims to shed light on how individual characteristics, emotional responses, and perceptions of civilian’s shape police officers’ attitudes toward protection, human rights, transitional justice, and institutional reform. By combining experimental and observational measures, the project examines how these factors intersect with actual behavior in the field, offering a novel perspective on the micro-foundations of police conduct in post-peace agreement settings. In doing so, it addresses the broader challenges of transforming security institutions in contexts marked by historical violence, deep social inequalities, and enduring ideological polarization. The empirical strategy combines two complementary causal components. First, we implement a randomized survey experiment in which officers evaluate hypothetical scenarios describing civilians who vary across key social and political attributes (political ideology, race, gender, territorial origin, and association with armed actors). These vignette dimensions are randomly assigned, allowing us to identify how civilian characteristics causally affect officers’ willingness to protect, perceived threat, emotional reactions, and support for rights-respecting conduct. Second, we exploit the phased rollout of newly mandated professional training courses within the Colombian National Police. Because officers complete the training at different points in time, we leverage this temporal variation to construct a treatment group (officers who have already completed the course) and a control group (officers who have not yet participated at the time of the survey). This staggered implementation provides quasi-experimental leverage to estimate the causal effect of professional training on officers’ role orientation (protective vs. warrior mindsets), their respect for human rights, and their general disposition toward civilian protection. Crucially, the vignette experiment is embedded within both groups. By randomly assigning identical scenarios to trained and untrained officers, we can test not only whether training shifts average professional orientations, but also whether it moderates selective protection biases. In other words, the design allows us to assess whether training reduces differential willingness to protect civilians based on ideology, race, or territorial background, thereby addressing the micro-foundations of institutional transformation.