The Effects of DNA Databases on the Deterrence and Detection of Offenders
Anne Sofie Tegner Anker, Jennifer Doleac & Rasmus Landersø
University of Virginia Working Paper, December 2017
This paper studies how collecting offender DNA profiles affects offenders’ later recidivism and likelihood of getting caught by exploiting a large expansion of Denmark’s DNA database. We find that DNA profiling increases detection probability and reduces recidivism within the following year by as much as 43%. We estimate the elasticity of criminal behavior with respect to the probability of detection to be -1.7, implying that a 1% higher detection probability reduces crime by almost 2%. We also find that DNA profiling changes non-criminal behavior: profiled offenders are more likely to engage in a stable relationship, and live with their children.
The Digital Sin City: An Empirical Study of Craigslist's Impact on Prostitution Trends
Jason Chan, Probal Mojumder & Anindya Ghose
University of Minnesota Working Paper, October 2017
The Internet facilitates information flow between sex workers and buyers, making it easier to set up paid sexual transactions online. Despite the illegality of selling sexual services online, Section 230 of Communications Decency Act shields websites from liability for unlawful postings by third parties. Consequently, the websites like Craigslist have become a haven for prostitution-related ads. With increasing number of prostitution-related sites launched over time, it is imperative to understand the link between these sites and prostitution trends. Specifically, in this paper, we quantify the economic impact of Craigslist’s entry on prostitution incidence, and identify potential pathways in which the website affects the sex industry. Using a national panel data for 1,796 U.S. counties from 1999 to 2008, our analyses suggest that entry of Craigslist to a county leads to a 17.58 percent increase in prostitution cases. In addition, the analyses reveal that a majority of prostitution on Craigslist are induced by organized vice groups, in addition to voluntary participation by smaller set of independent providers. Further, we find site entry has a stronger impact in counties with past history of prostitution and produces spillover effects in neighboring locations that are not directly served by Craigslist. Sex workers providing niche sexual services are found to increase with site entry. We find that the increase in prostitution arrests does not catch up with the growth in prostitution trends brought in by Craigslist. Finally, we find complementarity effects between erotic and casual sex ads in leading to the increase of prostitution. Our results contribute broadly to the emerging literature on the societal challenges associated with online intermediaries and Internet penetration, and serve to provide guidelines for policy makers in regulating the sex industry in the Internet era.
Crime and Public Housing: A General Equilibrium Analysis
Boston University Working Paper, November 2017
Did the demolition of 22,000 units of public housing in Chicago cause reductions in city-wide crime? I use time series variation to demonstrate how standard identification strategies which exploit cross-sectional variation in distance to demolition fail to capture general equilibrium spillovers. This idea leads to a non-parametric statistical test that is useful for diagnosing control group contamination in difference-in-difference designs with many treatment periods. Point estimates from the time series which incorporate both the direct and indirect effect of the demolitions indicate that in the short run the average demolition increased city-wide crime by 0.5% per month relative to baseline, with no evidence of long run reductions. Last, I provide suggestive evidence that the spillovers are mediated by demolition-induced migration across gang territorial boundaries.
Explaining Gun Deaths: Gun Control, Mental Illness, and Policymaking in the American States
Jacob Smith & Jonathan Spiegler
Policy Studies Journal, forthcoming
Seeking to test two commonly proposed solutions to gun deaths in the United States, we examine the extent to which (1) tougher gun control laws, (2) greater access to mental health services, and (3) a combination of both approaches affect the rate of gun deaths in American states. We find that tougher gun control laws, as well as a combination of both approaches, are associated with a lower overall rate of gun deaths, and with a lower rate of nonsuicide gun deaths, while only tougher gun control laws are significantly associated with a reduction in the rate of gun-related suicides. Our findings serve as an initial guide to policymakers seeking to reduce the rate of gun deaths in their states.
Ecological Networks and Urban Crime: The Structure of Shared Routine Activity Locations and Neighborhood-Level Informal Control Capacity
Christopher Browning et al.
Criminology, November 2017, Pages 754–778
By drawing on the work of Jacobs (1961), we hypothesize that public contact among neighborhood residents while engaged in day-to-day routines, captured by the aggregate network structure of shared local exposure, is consequential for crime. Neighborhoods in which residents come into contact more extensively in the course of conventional routines will exhibit higher levels of public familiarity, trust, and collective efficacy with implications for the informal social control of crime. We employ the concept of ecological (“eco-”) networks — networks linking households within neighborhoods through shared activity locations — to formalize the notion of overlapping routines. By using microsimulations of household travel patterns to construct census tract-level eco-networks for Columbus, OH, we examine the hypothesis that eco-network intensity (the probability that households tied through one location in a neighborhood eco-network will also be tied through another visited location) is negatively associated with tract-level crime rates (N = 192). Fitted spatial autoregressive models offer evidence that neighborhoods with higher intensity eco-networks exhibit lower levels of violent and property crime. In contrast, a higher prevalence of nonresident visitors to a given tract is positively associated with property crime. The results of these analyses hold the potential to enrich insight into the ecological processes that shape variation in neighborhood crime.
Analysis of the Strength of Legal Firearms Restrictions for Perpetrators of Domestic Violence and Their Association With Intimate Partner Homicide
April Zeoli et al.
American Journal of Epidemiology, forthcoming
In this research, we estimate the association of firearm restrictions for domestic violence offenders with intimate partner homicides (IPHs), based on the strength of the policies. We posit that the association of firearm laws with IPHs depends on the laws’: 1) breadth of coverage of high-risk individuals and situations restricted; 2) power to compel firearm surrender or removal from prohibited persons; and 3) systems of accountability that prevent prohibited persons from obtaining guns. We conducted a quantitative policy evaluation using annual state-level data from 1980 through 2013 for 45 US states. Based on the results of a series of robust negative binomial regression models with state fixed effects, domestic violence restraining order firearm prohibition laws are associated with 9% reductions in IPH. Statistically significant protective associations were evident only when restraining order prohibitions covered dating partners (−10%) and ex parte orders (−12%). Laws prohibiting access to those convicted of non-specific violent misdemeanors were associated with a 23% reduction in IPH rates; there was no association when prohibitions were limited to domestic violence. Permit-to-purchase laws were associated with 10% reductions in IPHs. These findings should inform policymakers considering laws to maximize protections against intimate partner homicide.
Aggression and sleep: A daylight saving time natural experiment on the effect of mild sleep loss and gain on assaults
Rebecca Umbach, Adrian Raine & Greg Ridgeway
Journal of Experimental Criminology, December 2017, Pages 439–453
Methods: Using National Incidence Based Reporting System data and city-reported data from Chicago, New York, Philadelphia, and Los Angeles, we calculated the difference in assault rates on the Monday immediately following daylight saving time (DST) as compared to the Monday a week later using a Poisson quasi-maximum likelihood estimator model. The same analyses were performed to examine effects of the return to standard time in the fall. We employed several falsification checks.
Results: There were 2.9% fewer (95% CI: –4.2%, −1.6%, p < 0.0001) assaults immediately following DST, when we lose an hour, as compared to a week later. In contrast, there was a 2.8% rise in assaults immediately following the return to standard time, when an hour is gained, as compared to a week later (95% CI: 1.5%, 4.2%, p < 0.0001). Multiple falsification analyses suggest the spring findings to be robust, while the evidence to support the fall findings is weaker.
Conclusions: This study suggests that mild and short-term changes in sleep do significantly affect rates of assault. Specifically, there is support for the theory that mild sleepiness possibly associated with an hour loss of sleep results in reduced assaults. This contradicts the simple inverse relationship currently suggested by most of the correlational literature. This study and the mixed findings presented by experimental studies indicate that measurement variability of both sleep and aggression may result in conflicting findings.
Outlaw and economics: Biker gangs and club goods
Rationality and Society, forthcoming
Today, outlaw motorcycle gangs are best known for their involvement in an international criminal network dealing in narcotics, human trafficking, and arms smuggling. Law enforcement agencies in three continents have identified groups like the Hells Angels Motorcycle Club, the Outlaws Motorcycle Club, and the Bandidos Motorcycle Club as a major threat to public safety. Before their descent into organized crime, outlaw bikers captured the imagination of the American public due to their peculiar look and outrageous behavior. They dressed in dirty sleeveless leather jackets and Nazi paraphernalia, their arms covered in tattoos of Nazi and White-supremacist symbolism. They drove highly customized, loud, and heavy American bikes — almost always Harley-Davidsons — and despised Japanese vehicles. They were notorious for their erratic behavior, in particular, the propensity to use violence in an idiosyncratic way when interacting with non-bikers and the public display of nudity and sexual practices. Unlike standard treatments of outlaw bikers, which draw from criminology, sociology, and psychology, I propose an explanation for these seemingly irrational and certainly odd practices rooted on the economic approach. Following the literature on the economic theory of religious sects, I argue that these odd practices served as effective obstacles to the ability of outlaw bikers to free ride on the club goods provided by these organizations.
Foreshadowing targeted violence: Assessing leakage of intent by public mass murderers
James Silver, John Horgan & Paul Gill
Aggression and Violent Behavior, January–February 2018, Pages 94–100
The idea that identifiable behaviors presage violence is a core concept in the threat assessment literature. Especially meaningful from an operational perspective is “leakage”, which concerns whether offenders intentionally or unintentionally reveal insights into their thoughts or feelings that suggest impending targeted violence. Previous research has generally been limited to assessing the prevalence of leakage in various offender populations. The present study more thoroughly describes leakage in a sample of 115 public mass murderers in the U.S. whose activities took place between 1990 and 2014. We disaggregate leakage into three distinct forms (written statements, verbal statements to the public, verbal statements to family/friends), and examine these in relation to theorized correlates of leakage. The only significant predictor of leakage we found is the presence of a grievance, specifically a grievance against a person or entity, as opposed to a grievance against a category of people or a grievance against an idea, movement or religion. We discuss implications of these results as well as directions for future research.
New Evidence on the Causal Impact of Traffic Safety Laws on Drunk Driving and Traffic Fatalities
Nicholas Wright & La-Troy Lee
Georgia State University Working Paper, November 2017
In the United States, about 28 lives are lost daily in motor vehicle accidents that involve an alcohol-impaired driver. The conventional wisdom is that these accidents can be prevented through the use of strict traffic laws that are robustly enforced, though no consensus exists on the causal impact of these laws in reducing motor vehicle-related fatalities. This paper exploits quasi-random variation in state-level driving and road safety restrictions to estimate the causal effect of select traffic laws on the number of fatal accidents and fatal accidents involving a drunk driver. In this paper, we employ the contiguous-border county-pair approach. This is causally identified from the discontinuities in policy treatments among homogeneous contiguous counties that are separated by a shared state border. This approach addresses the econometric issues created due to spatial heterogeneity that may have biased several studies in the literature. The analysis reveals that the laws related to accident prevention, such as having a good graduated licensing system, Pigovian beer taxes, and primary seatbelt enforcement, are the most effective in reducing motor vehicle-related fatalities. Using these estimated coefficients, simple simulations suggest that policymakers have been utilizing existing traffic laws sub-optimally, saving only 17% of the lives lost to motor vehicle crashes.
Looking Down the Barrel of a Loaded Gun: The Effect of Mandatory Handgun Purchase Delays on Homicide and Suicide
Griffin Edwards et al.
Economic Journal, forthcoming
We exploit within-state variation across time in both the existence and length of statutory delays — both explicit wait periods and delays created by licensing requirements — between the purchase and delivery of a firearm to examine the effect of purchase delays on homicides and suicides. We find that the existence of a purchase delay reduces firearm related suicides by between 2 to 5 percent with no statistically significant increase in non-firearm suicides. Purchase delays are not associated with statistically significant changes in homicide rates.
The role of verbal intelligence in becoming a successful criminal: Results from a longitudinal sample
Cashen Boccio, Kevin Beaver & Joseph Schwartz
Intelligence, January–February 2018, Pages 24-31
Intelligence has been linked with success across a wide array of life domains. To date, however, relatively little research has examined whether intelligence may predict criminal success — that is, engaging in criminal behaviors, but escaping detection and arrest. The current study addresses this gap in the literature by examining the associations among verbal intelligence, criminal involvement, and criminal justice processing (i.e., arrest) using data from the National Longitudinal Study of Adolescent to Adult Health (Add Health). Our findings reveal that verbal intelligence is associated with criminal justice processing, wherein individuals with higher verbal intelligence scores are more likely to avoid arrest for criminal behavior when compared with individuals with comparatively lower verbal intelligence scores. We discuss the implications of these findings for future research.
A Partner in Crime: Assortative Matching and Bias in the Crime Market
Journal of Economic Behavior & Organization, forthcoming
In the market for criminals the Prisoner's Dilemma creates an incentive for every criminal to find a partner with a higher likelihood of success in evading arrest. If no one is arrested, no one will have to talk to the police about the partnership and be at fault for the arrest of the other partner. This incentive leads to an equilibrium pattern of Positive Assortative Matching (PAM) on the likelihood of success. Using individual matched report-arrest data from the National Incident Based Reporting System and a novel empirical strategy, I find that offenders generally match according to PAM on the probability of success. I find that matching between criminals of different race can be subject to search frictions such as availability of potential partners. Matching between genders is likely biased, as females have a higher probability of success than males, contrary to unconditional success ranking. This bias persists in groups of three but diminishes in matches between older criminals, consistent with notions of discrimination. The results are of interest to policy makers who want to impede criminals from organizing.