Reducing gun violence

Taking into account both potential implementation challenges and expected efficacy, investing in police-led targeted enforcement directed at places and persons at high risk for gun crime (e.g.,"hot spot" policing; gang enforcement) would reduce gun violence. - participant responses

Participant	Vote	Confidence	Comment
Stephen Billings University of Colorado - Boulder	Agree	8
Shawn Bushway RAND Corporation	Agree	8
Aaron Chalfin University of Pennsylvania	Strongly Agree	8	Gun violence is highly concentrated among a small number of people and places and so highly-targeted approaches can potentially pay large dividends. Implemented poorly, this approach is potentially accompanied by high costs. But, when it comes to improving public safety, it is also the strategy that has the greatest evidence behind it.
Megan Stevenson University of Virginia	Neutral/No Opinion	1	Gun violence is not an area of expertise for me.
Randi Hjalmarsson University of Gothenberg	Agree	7
Anna Bindler University of Cologne	Agree	8
Paul Heaton University of Pennsylvania	Agree	5	There is solid evidence in favor of well-executed "hot spots" type interventions, but not all law enforcement agencies may be able to execute these approaches sufficiently well to reduce violence.
Rosanna Smart RAND Corporation	Agree	9
Michael Makowsky Clemson University	Disagree	4	While I remain persuaded that US cities are largely under-policed, and I have no trouble imagining specific "hot spots" where targeted enforcement may lead to beneficial reductions in violence, I am skeptical of narrowly-targeted policing as an effective policy. Absent concomitant increases in policing coordinated with community involvement, my expectation would be significant displacement of violence to neighboring areas, following idiosyncratic patterns where new sets of unexpected interaction
David Kirk University of Oxford	Agree	6
David Weisburd George Mason University	Strongly Agree	10
Patrick Sharkey Princeton University	Agree	5	The first part of the question is what's crucial. Some of these programs have been shown to be effective, but my sense is that in less professional departments they would not be effective.
David Harding University of California - Berkeley	Neutral/No Opinion	5
John MacDonald University of Pennsylvania	Strongly Agree	10
Giovanni Mastrobuoni Collegio Carlo Alberto	Agree	5
Benjamin Hansen University of Oregon	Agree	7	In addition to the wealth of literature about police employment levels and violent crime, a recent paper by Chalfin, Laforest, and Kaplan forthcoming in JPAM studies targeted "gang takedowns" finding robust evidence they reduce local gang violence.
Monica Deza City University of New York	Strongly Agree	9	Targeting the policy to the high risk group is crucial for cost-efficiency, but it requires training on determinants of being part of the target group
Anna Harvey New York University	Agree	7	See Chalfin, LaForest and Kaplan, "Can Precision Policing Reduce Gun Violence," JPAM 2021; multiple RCTs of "hot spot" policing.
Robert Apel Rutgers University	Disagree	3	Generally speaking, there is a lot of experimental evidence about hot spots interventions, although it is worth pointing out the overall effect sizes are pretty small and the interventions are highly variable--so variable, in my view, we cannot say with certainty what police should actually *do* in hot spots aside from increasing visible presence. Evaluations devoted specifically to gun violence are few and non-experimental (ergo, endogenous sample selection and endogenous treatment effects), an
Paolo Pinotti Bocconi University	Disagree	5	I believe targeting at places and persons, respectively, have very different implications for crime -- the latter being much more effective than the former
William Horrace Syracuse University	Neutral/No Opinion	5
Jillian Carr Purdue University	Neutral/No Opinion	5	There is a lot of variation in the types of programs that could fit this criteria. Some could be successful (some gang enforcement strategies seem to work) but others have been unsuccessful.
Morgan Williams, Jr. New York University	Agree	8
Jens Ludwig University of Chicago	Agree	7	We have an accumulating body of evidence from randomized experiments and 'natural experiments' suggesting these types of focused strategies can reduce gun violence. What we know much less about is what the consequences are for enforcement harms and community well-being, as well as community attitudes towards police or government more generally.
David Abrams University of Pennsylvania	Agree	5	Much more confident in interventions targeted based on individual affiliation and history than location. Not so much on hot spot policing.
Stephen Machin London School of Economics	Agree	7	If done effectively - and it is likely to be expensive, so costs of ineffective targeting could be high.
John Donohue Stanford University	Strongly Agree	9
Jennifer Doleac Texas A&M University	Agree	7	There is more evidence on this than other popular strategies. But targeted police interventions always come with a risk of social costs (particularly if that enforcement is perceived as harassment by individuals/communities).
Manudeep Bhuller University of Oslo	Agree	7
Santiago Tobón Universidad EAFIT	Neutral/No Opinion	10	It depends on the context and the drivers of gun violence. An example from Colombia. In Medellín, a city where most homicides are instrumental, hot spots policing did not impact homicides. In Bogotá, a city where most homicides are emotional, hot spots policing did impact homicides.
Emily Owens University of California - Irvine	Agree	6	The NAS reviewed this in 2018. Multiple RCTs of place based policing policies in the US have found evidence of localized short run crime reductions, and not found strong evidence of adverse spillovers. At the same time, jurisdiction-level, longer run evaluations haven't really been done. The existence of benefits also does not mean there are no costs.
John Pepper University of Virginia	Agree	7	There is a large body well-designed research showing consistent evidence. But, there is significant heterogeneity, the evaluations generally only provide short run measures, and external validity is a concern.
Kevin Schnepel Simon Fraser University	Agree	5
Charles Loeffler University of Pennsylvania	Strongly Agree	8
Greg Midgette University of Maryland	Agree	8	Hot spots policing can reduce violent crime, and studies of potential displacement to neighboring areas are promising. Heavier targeted enforcement without accompanying community outreach might generate intangible costs to the community that may impede investigations, and some tactics are more effective against violence than others.
Ayobami Laniyonu University of Toronto, St. George	Agree	4
Allison Harris Yale University	Disagree	8
Amanda Agan Rutgers University	Agree	4

Taking into account both potential implementation challenges and expected efficacy, investing in police-led focused deterrence programs (clearly communicating “carrots and sticks” to local residents identified as high risk, followed by targeted surveillance and enforcement with some community-based support for those who desist from crime) would reduce gun violence. - participant responses

Participant	Vote	Confidence	Comment
Stephen Billings University of Colorado - Boulder	Neutral/No Opinion	7
Shawn Bushway RAND Corporation	Agree	7
Aaron Chalfin University of Pennsylvania	Agree	5	The available evidence for these approaches is mostly positive but it is exclusively quasi-experimental. There's likely to be wide variation in effectiveness depending on the quality of implementation.
Megan Stevenson University of Virginia	Neutral/No Opinion	1
Randi Hjalmarsson University of Gothenberg	Agree	7
Anna Bindler University of Cologne	Agree	8
Paul Heaton University of Pennsylvania	Neutral/No Opinion	8
Rosanna Smart RAND Corporation	Agree	7
Michael Makowsky Clemson University	Disagree	4	I am skeptical of the capacity, at this stage, to produce targeted deterrence and surveillance programs that communities would trust not to be riddled with bias and abuse. I would be far more sanguine of such programs implemented by organizations parallel to, but not inclusive of, the police.
David Kirk University of Oxford	Agree	3	Depends on implementation, and whether question is about short-run or long-term reductions.
David Weisburd George Mason University	Agree	9
Patrick Sharkey Princeton University	Neutral/No Opinion	5	The evidence base is very mixed. I designed and carried out an RCT to evaluate a program in NY State that did not affect crime, but I am persuaded that programs in Boston and Chicago were effective.
David Harding University of California - Berkeley	Agree	8
John MacDonald University of Pennsylvania	Agree	8	This approach has more challenges to implementation because it requires coordination across multiple agencies and believable messaging.
Giovanni Mastrobuoni Collegio Carlo Alberto	Agree	4
Benjamin Hansen University of Oregon	Agree	7	Programs like this in NYC, evaluated by Sharkey, through RCTs, have been show to mainly affect technical violations among recent paroles, rather than reducing overall observed violence. However, such studies need to address the potential for competing risks in their evaluation designs. A working paper by Raphael and Skogg do find a targeted monitoring program in California, Armed and Prohibited Persons System (APPS), introduced in 2006 substantially reduced gun violence.
Monica Deza City University of New York	Agree	6	Awareness would decrease gun violence to some extent, but violent crimes are usually a result of "spur of the moment." I think gun availability may play a larger role than awareness of the consequences of using gunf
Anna Harvey New York University	Neutral/No Opinion	5	Focused deterrence programs vary widely in their design and implementation and are rarely randomized, so it's hard to have a well-defined expectation of whether these programs reduce gun violence. A randomized trial of an FD program in NYS (Sharkey 2021) reported no effects on gun violence arrests.
Robert Apel Rutgers University	Disagree	3	Evaluation is immensely difficult because of explicit sampling on risk, violation of the non-interference assumption of SUTVA when the intervention target is groups, and the question of the appropriate unit of analysis for measuring outcomes. There is not yet any experimental evidence, as far as I am aware. The quasi-experimental evidence suggests effectiveness on violence especially when the target is gangs as opposed to individuals, but while promising, this must await more carefully designed
Paolo Pinotti Bocconi University	Agree	5
William Horrace Syracuse University	Neutral/No Opinion	6
Jillian Carr Purdue University	Agree	7	There appears to be support for this in the literature. The successful programs with which I'm familiar were generally police-led, but also heavily involved the community.
Morgan Williams, Jr. New York University	Agree	7
Jens Ludwig University of Chicago	Neutral/No Opinion	2	I can see the logic behind these sorts of strategies but they have been challenging to evaluate, and from what I understand can also be challenging to implement with fidelity in a sustained way. More and better evidence on these strategies would be very important for policy because if they worked, they create the potential to reduce gun violence without some of the collateral costs of more traditional policing strategies.
David Abrams University of Pennsylvania	Agree	7
Stephen Machin London School of Economics	Agree	7	Again could prove hard to coordinate effectively and therefore costly if coordination and communication issues ensue.
John Donohue Stanford University	Agree	5
Jennifer Doleac Texas A&M University	Disagree	5	The evidence on focused deterrence strategies is pretty weak, though it sounds good in theory ("carrots & sticks " = exactly the strategy an economist might design). Targeted surveillance by police can be harmful if done poorly, so we need to be sure that benefits outweigh any social costs. The only rigorous study on this strategy is an RCT by Pat Sharkey - found no effect on arrests.
Manudeep Bhuller University of Oslo	Agree	5
Santiago Tobón Universidad EAFIT	Neutral/No Opinion	8	The broad idea of focused deterrence is reasonable, but we need more evidence. These interventions might even have adverse spillovers, inducing violence in other places/groups/people
Emily Owens University of California - Irvine	Agree	3	The NAS reviewed this in 2018. The evidence on the impact of person-based deterrence strategies isn't as strong as the place-based ones; many quasi-experimental studies do find reductions in crime (specifically gang violence and crime associated with drug markets) in affected areas, larger crime reducing effects are generally found in studies with weaker designs. I think implementation really matters. The existence of benefits also does not mean there are no costs.
John Pepper University of Virginia	Agree	4	There is some recent evidence supporting this conclusion, but it is limited.
Kevin Schnepel Simon Fraser University	Agree	5
Charles Loeffler University of Pennsylvania	Strongly Agree	9
Greg Midgette University of Maryland	Agree	6	A fairly comprehensive and recent review by Braga and colleagues (2018) found the programs to be effective on average. That said, relationships between some police departments and communities are so fraught that the coordination these programs require might be a major challenge to implementation and effectiveness.
Ayobami Laniyonu University of Toronto, St. George	Neutral/No Opinion	5
Allison Harris Yale University	Disagree	8
Amanda Agan Rutgers University	Neutral/No Opinion	5	Some (weak) evidence that focused deterrence programs reduce violence (see e.g. Braga et al. 2019), recent RCT evidence shows no reduction. Increased police contact for targeted individuals is a cost.

Taking into account both potential implementation challenges and expected efficacy, investing in purely community-led violence-interruption programs (community-based outreach workers try to mediate and prevent conflict, without police involvement) would reduce gun violence. - participant responses

Participant	Vote	Confidence	Comment
Stephen Billings University of Colorado - Boulder	Neutral/No Opinion	6
Shawn Bushway RAND Corporation	Disagree	7
Aaron Chalfin University of Pennsylvania	Neutral/No Opinion	3	These approaches have been around for many years. The available evidence, all quasi-experimental, is very mixed.. The quality of implementation figures to be very important here and so scalability is an eternal question. As long as the programs are inexpensive, it makes sense to continue to support innovation in this area because if the approach works, it's a wonderful idea. But this is a hard approach to count on based on what we know today.
Megan Stevenson University of Virginia	Neutral/No Opinion	1
Randi Hjalmarsson University of Gothenberg	Neutral/No Opinion	5
Anna Bindler University of Cologne	Neutral/No Opinion	5
Paul Heaton University of Pennsylvania	Neutral/No Opinion	5
Rosanna Smart RAND Corporation	Neutral/No Opinion	2
Michael Makowsky Clemson University	Agree	6	I think such programs could be highly effective and would likely be complementary to "on-foot" increases in police presence. Trust in community policing would likely benefit from additional channels to mitigate violence without direct police involvement, reducing the number of interactions where police are forced to act outside of their expertise, potentially exacerbating problems and eroding trust. Such community programs would likely be complementary with "procedural justice" programs, helping
David Kirk University of Oxford	Neutral/No Opinion	5
David Weisburd George Mason University	Disagree	8
Patrick Sharkey Princeton University	Neutral/No Opinion	5	Some programs that fall under this description have been extremely effective, others have been disasters. The answer depends mostly on whether the organization running the program is funded at a level that allows for high-quality training, low turnover, and long-term sustainability.
David Harding University of California - Berkeley	Agree	5
John MacDonald University of Pennsylvania	Neutral/No Opinion	6	The evidence is pretty thin on the efficacy of violence interruption programs on their own. At the same time they are worth trying for they provide community-lead action to try and curb violence in neighborhoods.
Giovanni Mastrobuoni Collegio Carlo Alberto	Agree	5
Benjamin Hansen University of Oregon	Neutral/No Opinion	5	While some evaluations of Operation Cease-Fire have suggested they could have promise, replications and extensions have been far more mixed, and communities with said interventions have not shown long run reductions in violence and as youth homicides increased after initially dropping.
Monica Deza City University of New York	Strongly Disagree	10	A community needs a police department. There is a case to be made that the police could be trained better, but there is no substitute for policing.
Anna Harvey New York University	Neutral/No Opinion	5	Violence interruption programs vary widely in their design and implementation and are rarely randomized, so it's hard to have a well-defined expectation of whether these programs reduce gun violence. The Chicago READI program is currently being evaluated, but final results are not yet available.
Robert Apel Rutgers University	Agree	5	Evaluations of violence interruption programs face similar challenges as focused deterrence, and the quality of evidence is comparably unexceptional but promising. But unlike police-led interventions, I would predict a number of positive spillovers stemming from the direct investment in local organizations (which is linked with lower violent crime rates, as strong quasi-experimental evidence indicates), employment of justice-impacted individuals as interrupters (who would otherwise face employme
Paolo Pinotti Bocconi University	Neutral/No Opinion	5
William Horrace Syracuse University	Agree	6
Jillian Carr Purdue University	Agree	8	I've seem some examples of these programs that seem to work well. The ones I'm thinking of were rolled out with other initiatives as well, so it's not easy to say which programs were the most effective.
Morgan Williams, Jr. New York University	Neutral/No Opinion	8
Jens Ludwig University of Chicago	Neutral/No Opinion	2	As with focused deterrence programs, it's not hard to see and appreciate the logic behind these strategies. But these community-led violence interruption strategies are difficult to evaluate, and we don't have much convincing evidence on them one way or the other yet in my opinion. I hope that the US Department of Justice uses the major new investments in these types of programs that is part of the Recovery Plan as an opportunity to generate much better , rigorous evidence on their impacts, by f
David Abrams University of Pennsylvania	Agree	7
Stephen Machin London School of Economics	Disagree	6	On such a serious and dangerous issue of gun violence, community outreach alone without police involvement would seem difficult to justify, but this is a very hard issue that may well be more of less suitable for different settings depending on a number of factors, including the issue of how well community/police relations function and historical place specific relevant factors.
John Donohue Stanford University	Disagree	4	I have a relatively low level of confidence in my opinion here, but my sense is that some of the studies suggesting these programs work are just illustrating regression to the mean after crime spikes led to their implementation.
Jennifer Doleac Texas A&M University	Neutral/No Opinion	5	The evidence on this strategy (Cure Violence and similar interventions) is extremely weak. We basically have no idea if it works. But even the correlational studies aren't consistently finding beneficial effects, which should give us pause. Communities that want to implement this strategy should do so in a way that enables them to tell if the investment is worth it.
Manudeep Bhuller University of Oslo	Neutral/No Opinion	3
Santiago Tobón Universidad EAFIT	Neutral/No Opinion	8	As with focused deterrence, the broad idea of violence interruptors is reasonable, but we also need more evidence in this case. There are challenges also in how some community members might result injured as a result of their participation in the program
Emily Owens University of California - Irvine	Neutral/No Opinion	1	Maybe - I know less about the evidence base for this specific set of strategies. I am aware of case studies that find very promising reductions in violence without police involvement, and at least one reasonable quasi experimental evaluation of the impact of nonprofits on violence.
John Pepper University of Virginia	Neutral/No Opinion	8	While this may be true, the existing evidence does not support this conclusion.
Kevin Schnepel Simon Fraser University	Agree	5
Charles Loeffler University of Pennsylvania	Agree	5
Greg Midgette University of Maryland	Neutral/No Opinion	9	The evidence I'm aware of is mixed but promising. It seems that the impact of these programs can depend a lot on the bundle of actions and groups comprising implementation and the setting in which they're run. Some versions may work better than others, and the evidence is sparse on what's preferable.
Ayobami Laniyonu University of Toronto, St. George	Agree	6
Allison Harris Yale University	Agree	8
Amanda Agan Rutgers University	Neutral/No Opinion	5	Evidence on community-based programs like Cure Violence is weak and mixed (e.g. Butts et al. 2015)