Econometrics

3-26-2021

Chamberlain Seminar

Vira Semenova

Better Lee Bounds

ATE with large number of covariates.

4-8-2021

Harvard/MIT Seminars in Econometrics

Ismael Mourifie

Layered Policy Analysis in Program Evaluation using the MTE

MTE using copula - not sensitive to endogeneity.

4-9-2021

Chamberlain Seminar

Alex Torgovitsky

“Instrumental Variables with Multiple Instruments”

The Causal Interpretation of Two-Stage Least Squares with Multiple Instrumental Variables

Policy Evaluation with Multiple Instrumental Variables

The accompanying Stata module can be found on his personal website.

Causal Inference

  • 4-1-2021

Harvard/MIT Seminar in Econometrics

Dmitry Arkhangelsky

On Policy Evaluation with Aggregate Time-Series Shocks

  • 5-7-2021

Chamberlain Seminar

Anna Mikusheva

Inference with Many Weak Instruments

We develop a concept of weak identification in linear IV models in which the number of instruments can grow at the same rate or slower than the sample size. We propose a jackknifed version of the classical weak identification-robust Anderson-Rubin (AR) test statistic. Large-sample inference based on the jackknifed AR is valid under heteroscedasticity and weak identication. The feasible version of this statistic uses a novel variance estimator. The test has uniformly correct size and good power properties. We also develop a pre-test for weak identification that is related to the size property of a Wald test based on the Jackknife Instrumental Variable Estimator. This new pre-test is valid under heteroscedasticity and with many instruments.

  • 6-4-2021

Chamberlain Seminar

Shu Shen

Dynamic Regression Discontinuity Under Treatment Effect Heterogeneity

  • Regression discontinuity (RD) is a popular tool for the analysis of economic policies or treatment interventions. This paper extends the classic static RD model to a dynamic framework, where observations are eligible for repeated RD eligibility tests and, therefore, treatments. Such kinds of dynamics complicate the identification and estimation of longer-term average treatment effects. Previous empirical research with such designs typically ignored the dynamics in the model or adopted restrictive identifying assumptions. This paper lays out the dynamic RD model using the potential outcome framework and proposes nonparametric and semiparametric identification and estimation strategies under various assumptions. The proposed methods are applied to revisit the effect of local school bonds in the seminal study of Cellini et al. (2010). Estimated effects are qualitatively similar or somewhat larger than reported in Cellini et al. (2010).

Networks

4-15-2021

Harvard/MIT Seminar in Econometrics

Bryan Graham

An optimal test for strategic interaction in social and economic network formation beteen heterogeneous agents

Python package for the simulation is available on the same page.

4-26-2021

BC Econometrics Seminar

Susanne Schennach

Independent Nonlinear Component Analysis

Primarily applied to regularization of control variables.

Crime

4-1-2021

Econ of Crime Seminar

Mike LaForest

The Effects of Parole Conditions and Early Release at the Margin

How the effects of parole conditions and early release affect recidivism.

3-25-2021

ClassACTForum

Racism In the Criminal System: Communities Fight for Justice

4-15-2021

Hanna Hoover

Civil Rights Restoration and Recidivism

Florida is currently one of only a handful of states that does not restore civil rights upon completion of a sentence for felons. Instead, persons with a prior felony conviction must apply for civil rights restoration through a constitutionally authorized process known as clemency. Civil rights revocation not only revokes the right to vote, but also denies eligibility of certain occupational licenses and state-funded scholarships. Over the past 26 years, the Rules of Executive Clemency have been amended such that approval became automatic for those eligible between 2007 to 2011. Outside of this time period, however, clemency required an application, hearing, and a mandatory waiting period. It is unknown, a priori, if such policy changes influenced labor market outcomes, voting behavior, or educational attainment, any of which may affect incentives for convicted felons to re-offend. As an aggregate measure of these channels, I analyze the casual link between civil rights restoration and rates of recidivism. To establish causality, this paper uses changes in the Rules of Executive Clemency affecting ex-felons’ ability to restore their civil rights in order to identify the impact of civil rights restoration on rates of recidivism.

4-29-2021

Tinna Laufey Asgeirsdottir

Monetizing victim suffering due to violence

Abstract: Efficiency calculations of violence prevention are likely to be severely biased if the hard-to-measure value of utility reductions due to victimization are not included. We measure the monetary compensation needed to offset the welfare loss associated with being subjected to violence, by applying the compensating-income-variation method to data from an Icelandic health-and-lifestyle survey carried out in 2017. We examine differences in the monetary compensation needed by (i) types of violence, (ii) duration since the exposure, and (iii) the relationship with the perpetrator. Our results show that the cost associated with the welfare loss of violence is highest when the violence psychological, the perpetrator is a spouse or an ex-spouse, and when the violence happened recently.

5-20-2021

Jorge García Hombrados

Recidivism and Neighborhood Institutions: Evidence from the Rise of the Evangelical Church in Chile

Abstract: Rehabilitating convicted criminals is challenging and an important share of convicted criminals return to prison only a few years after their release. Thus, finding effective ways of encouraging crime desistance has become an important policy goal to reduce crime and incarceration rates. This paper provides causal evidence that the local institutions of the neighborhood where individuals return after prison matter. Specifically, we show that the opening of an Evangelical church reduces twelve-months re-incarceration rates among property crime offenders by more than 10 percentage points. This effect represents a drop of 16% in the probability of returning to prison for this group of individuals. We find smaller and less precise effects for more severe types of crime. We discuss two classes of mechanisms that could explain our results: religiosity and social support. We provide evidence that the social support provided by evangelical churches is an important driver of our results. This suggests that non-religious local institutions could also play an important role in the rehabilitation of former inmates.

5-27-2021

Teresa Molina

Persuasive agenda-setting: Rodrigo Duterte’s inauguration speech and drugs in the Philippines

  • Use Google Search data

6-3-2021

Brian Knight

Crime and Gender Segregation: Evidence from the Bogota Lockdown

Education and Crime

7-15-2021

Jason Baron

More School Funding, Less Crime?

Given the enormous annual costs of crime to society, a large literature has emerged that aims to identify effective crime prevention strategies. While researchers have long focused on enforcement and punishment as a crime prevention strategy, a growing literature emphasizes the potential efficiency gains of policy interventions that prevent the development of criminals in the first place. This debate is particularly salient today, as budget-constrained cities across the country face increasing calls to allocate the marginal dollar toward social programs and away from law enforcement. In this paper, we ask whether it is possible to reduce crime rates by increasing the amount of funding to a particularly important social program—public education. Despite its enormous policy implications, surprisingly little is known about the relationship between public school funding and criminal activity. Our empirical strategy examines whether public school children exposed to more funding during early grades are less likely to engage in criminal activity as adults. Specifically, we exploit plausibly exogenous variation in school funding driven by Michigan’s 1994 school finance reform, Proposal A, as well as a novel source of administrative records which link statewide data on the universe of individual public school records and adult arrests. We find that primary school students exposed to 10 per-cent ($1,000) more operational spending each year for four years experienced a 15% reduction in the probability of ever being arrested as adults. This effect was driven by reductions in the likelihood of ever being arrested for a violent, property, and drug crime, and was entirely concentrated in baseline low-income school districts. Additional funding during primary school also led to large reductions in the probability of ever being imprisoned. We show that the cost-effectiveness of public school funding as a crime-prevention strategy is similar to other early interventions such as Head Start and the Perry Preschool Project. Exploring mechanisms, we find that school districts targeted the marginal dollar toward instruction, support services for students, and the operation and maintenance of infrastructure. As a result, primary school students exposed to additional funding faced smaller class sizes and higher-paid teachers. These students had higher test scores in elementary school, higher attendance rates in middle school, were less likely to repeat a middle school grade and be placed in a juvenile detention center, and were more likely to graduate high school and college. However, improvements in all of these intermediate outcomes can explain only about 30% of the overall effect, which suggests that previous knowledge of how additional funding impacts educational attainment and test scores does not entirely predict how, or even whether, school funding will impact adult crime. Furthermore, these results suggest that additional funding might impact later-in-life crime through non-cognitive outcomes such as attendance and juvenile crime, which highlights the importance of not focusing exclusively on cognitive outcomes when evaluating the overall effects of funding.

Policy and Crime

4-22-2021

Marta Troya-Martinez

Mobbed Up: Control of Labor Unions as Cartel Enforcement

After select counties enforced Racketeer Influenced and Corrupt Organizations Act (RICO), they see a statistically significant rise in the number of new establishments and employment. The paper attributes it to the successful removal of mafia influence in these locations.

8-5-2021

Ashna Arora (University of Chicago)

Econ of Crime seminar

The Impact of Specialized Prosecution on the Safety of Domestic Violence Victims

Little is known about how to effectively reduce re-victimization among those who experience domestic violence, despite decades of investment in specialized programs run by law enforcement, prosecutors, and court systems across the United States. We study the effects of increasing prosecutor capacity and victim-focused wraparound services on cases deemed to be at highest risk of re-victimization in Cook County, Illinois. Prosecutors vary in their tendency to include cases in the program, which we use as an instrumental variable to show that specialized prosecution lowers the likelihood of homicide for those on the margin of inclusion. We discuss several mechanisms that may be driving these results, including the impact of the program on the likelihood of incarceration and the receipt of victim-focused services such as advocacy and civil litigation.

Policing and Crime

6-10-2021

Emily Weisburst

Do Police Make Too Many Arrests?

We ask whether reductions in arrests increase crime, a central question in the current debate about police reform in the United States. Identifying this causal effect is difficult, as changes in arrests are nonrandom and often reflect or coincide with changes in crime. We study changes in arrests after line-of-duty deaths of officers, events that are likely to impact peer police behavior through increased fear of job risk or emotional distress but are unlikely to directly affect civilian criminal activity or community trust. We document that officer deaths are not preceded by changes in crime within cities, indicating that their timing is plausibly exogenous. These deaths lead to short-term but significant reductions in arrests of all types, with the largest reductions in more discretionary arrests for lower level offenses. Despite this drop in arrest activity, we find no evidence of an increase in crime. Moreover, there does not appear to be a threshold level or duration of reduced arrest activity past which crime increases. Our findings suggest that police departments may be able to reduce marginal enforcement without incurring public safety costs.

7-21-2021

Aaron Chalfin

NBER SI Crime

The Professional Motivations of Police Officers

How do public sector workers balance their pro-social motivations with private interests? In this study of police officers, we exploit two institutional features that change the implicit cost of making an arrest: arrests made near the end of an officer’s shift are more likely to require overtime work, and arrests made on days when an officer “moonlights” at an off-duty job after their shift have a higher opportunity cost. We document two consequences for officer behavior. First, contrary to popular wisdom, officers reduce arrests near the end of their shift, and the quality of arrests increases. We argue that these patterns are driven by officer preferences rather than departmental policy, fatigue, or incapacitation from earlier arrests. Second, officers further reduce late-shift arrests on days in which they moonlight after work, suggesting that they are, in fact, modestly responsive to financial incentives. Using these results, we estimate a dynamic model that identifies an officer’s implied trade-off between private and pro-social motivations. We find that police officers exhibit high pro-social motivation towards their work. In contrast to prior research showing that law enforcement outcomes are sensitive to financial incentives at an institutional level, the behavior of individual officers — the “street-level bureaucrats” who enforce the law — is not meaningfully distorted by monetary considerations.

Others