The State That Forges Organized Criminal Groups

• In Event: Lightning Round: Graduate Student Research: Ideas, Institutions, Coalitions & Movements

Abstract

Summary
Under what conditions does the State make criminal groups stronger? I explore why, how and when law enforcement agents deviate from organizational mandates. I hypothesize that individual-level incentives and organizational culture explain their behavior. I study this in city-level case studies in Brazil, integrating large-n quantitative analysis with careful fieldwork.

Description
Under what conditions does law enforcement make organized criminal groups (OCGs) stronger? In Latin American cities, members of law enforcement agencies often face competing individual and organizational incentives when engaging with violent actors. The individual choices they make can have sweeping consequences: shaping where and how criminal groups exercise governance (Lessing, 2021) and the levels of violence following police raids (Magaloni et al. 2020). Indeed, police officers can go so far as to form their own vigilante groups (Paes Manso 2020; Bateson 2020). A prosecutor may decide not to press charges against murderous police officers, prioritizing this officer’s discretion to act in face of violence. Prison police may let criminal leaders exert governance to reduce lethal violence inside the prison system.

Previous research has found that, faced with criminal violence, States can cooperate, coexist or contest (Arias and Barnes 2017; Duran-Martinez 2015; Moncada 2016). Yet these works predominantly focus on State-level characteristics and arrangements (Solar 2018) and don’t differentiate between agencies, their mandates and the individuals that compose them, much less how these two levels interact.

This project contributes to our understanding of criminal governance persistence by exploring the other side of this state-society relationship: determinants of state actors' behavior. Members of the same agency treat groups differently and agencies have different attitudes towards the same OCG behavior. State capacity (Bergman 2018; Duran-Martinez 2015), OCG behavior (Cruz 2011; Dias 2009; Lessing 2017; Skarbek 2011) and democratic mechanisms (González 2019) alone do not explain such patterns.

My theory is that individual-level behavior and organizational-level incentives in these agencies shape individuals' resort to use of force, discretion and deviation in the face of OCGs. Wilson (2019) highlights the challenge of studying bureaucratic motivation. I build on his insight and explore how organizational culture shapes incentive structures, conditioning behavior and agency performance in combating organized criminal groups.

I hypothesize that mission-oriented organizations are more likely to either formally or informally reward members based on outcome than on processes (Suzuki and Hur 2020; Bhavnani and Lee 2017; Tsai 2007; Honig 2019; Khan, Khwaja, and Olken 2019; Brierley 2021; Scharpf and Gläßel 2020; Pierskalla and Sacks 2020; Pepinsky, Pierskalla, and Sacks 2017). The implication is that these incentives will encourage both compliance and mission-oriented deviation (Bertrand et al. 2020). These mechanisms also encourage rent-seeking behavior, reinforcing deviation.

By focusing on the agencies involved in brokering State-OCG relations, I build on the insight that cartel-state interaction shapes violence (Lessing 2015; Blattman et al. 2021; Moncada 2021; González 2019). A core contribution is to show how different types of bureaucrats can shape violence and OCG persistence in diverging ways. I explore both organizational (Peters, n.d.; Wilson 2019) and actor-level characteristics (Brehm and Gates 1999; Gailmard and Patty 2007) that first inform and then reproduce such behavior in following cohorts.

Design and Methods
In order to test these theories, I propose running city-level analysis in Brazilian cities selected according to differences in violent criminal state-society relations. Having worked in Brazilian non-profits for over 8 years, I have already collected data on police operations in Rio de Janeiro and constructed original datasets on judicial criminal procedures. I have also established working relationships with officials in aforementioned agencies. I intend to study Rio de Janeiro, São Paulo, and Manaus or Porto Alegre, large cities that vary in the number of recognized criminal groups present and the capacity of law enforcement agencies.

As a mixed-methods project, I foresee using semi-structured interviews to identify causal mechanisms and foresee (i) collecting data on police and prison guard behavior, leveraging quasi-experimental methods in their analysis; and (ii) generating data on prosecutors, due to scarcity of publicly available data on their behavior, fielding an experimental survey.

Author

• Ana Paula Pellegrino, Georgetown University