In spite of a large-scale police professionalization program implemented by the federal government of Mexico over the past six years, efforts to mitigate crime and generate citizen trust have produced wildly uneven results across the country. There is scant evidence about what works in law enforcement practices, and a poor understanding of the institutional, social and contextual dynamics that drive variation in police efficacy and citizen trust. This project partners with Mexico’s federal and local police agencies in their efforts to professionalize and improve their law enforcement capacity, while strengthening the rule of law and ensuring human rights.
With the use of surveys in several states in Mexico, focus groups, and interviews with police officers, police chiefs, and federal authorities, the goal of this project is to understand how the police functions from the inside; how differences in police organization and practices across states and municipalities affect police performance, and how institutional climate, training, and practices impact violent crime and professionalization.
The goal of this project is to a) describe the dynamics of violence in Mexico; b) the evolution of criminal groups, their dynamics across time, and map their territorial presence; and c) identify the way organize criminal groups (OCGs) “rule” the territories they control, including how they relate state officials and the communities where they operate. To do so, the Povgov Lab is using text analysis (on different sources, printed and digital), GIS, Neural Networks, and several data processing techniques.
We know very little about the mechanisms through which citizen trust in the police is established. Civil society, the private sector and even policy-makers themselves lack the necessary evidence that comes from detailed case studies and cross-regional comparisons to inform their monitoring and oversight strategies.
The Povgov Lab is using sub-national cases (i.e., the four biggest Metropolitan areas in Mexico) to understand how citizens’ attitudes, values, and preferences impact their perceptions of police and their ability to effectively protect them from criminal predation and whether citizens might fear police as much as criminals. In addition, we are diagnosing victimization, crime incidence, risk factors, criminal governance orders and territoriality, as well implementing both vignettes and list experiments in which respondents were asked to judge short text descriptions of (fictional) scenarios related to law enforcement behavior.
At the Poverty, Violence, and Governance Lab (Povgov), we are starting a new stream of research aimed at answering the question of how legally sponsored temporary migration can help families and communities escape cycles of poverty and violence.
Mexicans and Central Americans are estimated to account for roughly 7.6 million (around 67%) of the unauthorized immigrants in the United States. Many of them migrate to escape violence. This kind of migration brings a myriad of risks as migrants face dangerous environments as well as predation by organized crime groups and rent-seeking government officials. Upon arriving in the United States, they are at risk of deportation at any moment, thus exposing them to the risk of being returned to the environment that sent them abroad in the first place. Additionally, deportations impose stresses on families, inflict childhood traumas, and may themselves involve human rights violations.
We seek to explore the ways in which legally sponsored temporary migration can serve as an alternate approach to addressing the social problems that often motivate this form of immigration. We are particularly interested in understanding the effects of large increases in family monthly incomes (an average of US $3,000, which places the household in the top decile of the Mexican income distribution) on household poverty, human capital, health, and other family-level outcomes. We also want to understand how this large increase in household income can be mobilized to create local public goods and increases in political accountability. Finally, we are particularly interested in understanding ways to insulate these resources from predation by organized criminal groups.
While there is considerable research on the community effects of remittances, results remain mixed on a wide range of outcomes. Some argue that remittances erode governance, while others find evidence that it improves governance. Results with respect to issues ranging from pro-democratic attitudes to communal welfare outcomes are equally unclear.
The literature on remittances and migration faces important problems in isolating the causal impact of immigration due to the difficulty of finding a valid counterfactual. Over the course of the next five years, we will be able to study approximately 20,000 temporary workers who will be granted visas to work in the U.S. Of those, we estimate that, conservatively, between two and three thousand will be newcomers for whom we will be able to randomize the allocation of their first visa. We plan to carry out this intervention by recruiting workers from poor, marginalized communities. Some of these communities are located within the most violent areas in Mexico where the state has been captured by criminal groups – it is comparable to an intervention in insurgent-controlled territory during a civil war.
We hope to use this project to set up a longitudinal study that can follow subjects across generations and serve as a resource for researchers from diverse disciplines. We plan to track outcomes concerning health, human capital, social mobility, and entrepreneurship as well as the durability of these outcomes. We recognize that our study induces stark inequalities in poor communities and will examine the ways that communities respond to this challenge –whether this is a source of conflict or whether these new resources can be mobilized for the public good.
The Path from Science to Impact
This study embeds a new perspective in our broader research agenda that examines the mitigation of criminal violence. By advancing our understanding of how communities facing violent challenges respond to this influx of resources, we can better understand the available policy options intended to mitigate those problems. This question is relevant not only to Mexico –foreign aid programs often face similar problem whereby resources transferred to poor and violent communities may be expropriated by violent actors.
As of this year, the United States operates one of the largest immigration court systems in the world, with an estimated 830,000 cases pending in 68 courthouses across the country. As immigration court judges are often the deciding actors separating a foreign national's admission to or removal from the U.S., their decisions are frequently matters of life or death. Against the backdrop of a rapidly changing immigration landscape, Povgov seeks to better understand 1) the legal and non-legal considerations that enter into judicial decision-making, and 2) the long-term consequences of immigration court decisions on migrants’ well-being and home countries.
Disparities in adjudicative outcomes across the U.S. immigration courts are well-documented—while some judges exhibit asylum grant rates of over 90 percent, others have grant rates in the single digits. These variations have led some scholars to conclude that asylum and removal outcomes are less dependent on the legal facts of the case in question and more on the identity of the presiding immigration judge. While present research has extensively documented these discrepancies across time and space, they provide little insight intowhysuch disparities exist in the first place. To shed light on this question, we conduct a sentiment analysis of immigration court transcripts to determine whether recurrent themes and the manner in which they are considered in the courtroom setting vary according to judge-specific characteristics. This analysis will be combined with additional statistical approaches—such as predictive modeling and machine learning—to address other relevant considerations, including whether judges’ language usage during hearings can predict case outcomes.
Given how powerful nationality is as a determinant of asylum and removal decisions, we also examine how immigration court outcomes can affect migrants’ home countries and long-term well-being. The U.S. deported more than 267,000 migrants in fiscal year 2019 alone; yet, the effects of and dynamics underlying such mass deportations are not wholly understood. For instance, deportations of criminal offenders may facilitate an exporting of criminal capital that further exacerbates violence and crime in their home countries. This, in turn, may generate a cyclical effect whereby increases in crime and victimization incentivize more individuals to migrate to the U.S. Even more, immigrants deported back to their home countries often report experiencing heightened violence—ranging from extortion to kidnapping—at the hands of gangs and other criminal organizations. Through innovative mixed methods designs, we seek to understand how these dynamics function in concert to affect violence against migrants and overall immigration flows.