The paper below"Preventing the use of deadly force:the relationship between police agency policies and the rates of office-involved gun deaths",and answer the following questions:

1. What is the research question?
2. What dependent variables and independent variables are used?
3. Which variables are significant?Pick two significant results and interpret them(say something about what do the results mean in English).Do the results make sense to you?
4. (optional)which part of the research article is difficult to understand?

Home Tools Research paper.pdf x ? H 1 /9 Cleveland State University Search 'Highlight' Preventing the Use of Deadly Force: The Relationship between Police Agency Policies Lo Export PDF A and Rates of Officer-Involved Gun Deaths Adobe Export PDF Convert PDF Files to Word lay T. Jennings ka posisocicul or Excel Online Abstract: Killings of civilians by police officers have become a matter of intense public concern in the United States. research le bow in the Arnelle Streets High-profile deaths, especially those of black citizens, have caused outrage and sparked the Black Lives Matter move- Institute for Civic Life at the University of lexis al Again. His research leones on the ment with calls for dramatic changes in how police agencies operate. However, little systematic research exists to answer relationships between mitutors and the Select PDF File questions about which policies should be ended or put in place to reduce these deaths. The authors leverage a large capacity of cillems to effectively panic; the data set of gun deaths by police officers in the United States, combined with agency-level policy data and community in polises , he's mark drewes on the lieich al Research paper.pdf X demographic data, to examine whether certain policies are associated with lower or higher rates of officer-involved gun political communication, deaths. Findings show that one policy the requirement that officers file a report when they point their guns at people E-mail: jepiercings danthaneasedu bus do not fire-is associated with significantly lower rates of gun deaths. Convert to Meghan L. Rubado is assistant Practitioner Points pootensor of public administration and cup management in the Maine Goodman Microsoft Word (".docx) . Police agencies that require officers to file a report when they point their guns but do not shoot at civilians Levin Colege al Urban Alfalis at Cleveland had significantly lower rates of gun deaths by police officers. Roughly half of agencies required such reports State University. Har rocoach caamines how insfluians shape the decision- in 2013 Document Language: making process and affect local policy . This requirement may cause officers to use more caution when considering using force by pointing their Burcomes. Her primary interests are uhas English (U.S.) Change guns and may demonstrate to officers an agency commitment to best practices that protect civilians from service delivery local governance, and unnecessary use of force. E-mail: mabon scaphio sdu * The reporting requirement for displaying a gun is not associated with higher rates of gun deaths of police officers, suggesting that the policy does not endanger officers. While community policing and racial representation on police forces have been tied to improved outcomes in several areas, we do not find that these factors are associated with lower rates of police-involved gun deaths, Convert Fatal shootings of civilians by police officers in the past several years have drawn the attention our hypotheses on a large sample of U.S. cities and counties. We find that one policy-the requirement View Converted Files of the news media, protesters, and investigators that police officers file a report when they point their from all levels of government. Public outcry from gun at someone but do not fire-is associated with individuals and organizations, including the Black systematically lower rates of gun deaths of civilians. Lives Matter movement, has called for major changes Edit PDF in police agencies that result in reductions in these Our analysis relies on a unique data set compiled deaths, especially those of black citizens. But which by a nonprofit organization that has attempted to police agency policies are associated with higher catalog every instance of police-involved gun death incidence of fatal police-involved shootings? What since 2000. These are cases in which a police officer To Create PDF changes could public officials make to reduce these fired at a civilian, resulting in death. This database fatalities? These questions have received little in the project, called Fatal Encounters, is crowdsourced and way of systematic research by scholars. Given the fact-checked by the nonprofit. Fatal Encounters and Comment continual loss of civilian lives in these incidents and several less organized databases like it have sprung the recent unrest that has followed them in cities from up in recent years as a citizen solution to the lack of Cleveland to Minneapolis, Baltimore to Ferguson, reliable government data. The federal government these questions warrant attention from scholars, does not mandate collection of these data in a public Convert, edit and e-sign PDF public leaders, and American society in general database. The Federal Bureau of Investigation (FBI) forms & agreements does collect information on police-involved deaths of Public Administration Review. We suggest that certain policy choices by local officials civilians, but these data are systematically incomplete vol.27, 13s.Z, pp.217-226. 2017 by may reduce the rates of these deaths, and we test because of voluntary self-reporting by state and local The American Society for Public Adrink luvion. DOE 10,1 1 1 14at 12738, Free 7-Day Trial Preventing the Use of Deadly Force: The Relationship between Police Agency Policies and Rates of Officer-Involved Gun Deaths 217Home Tools Research paper.pdf x ? H 2 /9 agencies (Burghart 2014: Davis and Lowery 2015). We merge the We interviewed leaders of several police agencies about this policy Fatal Encounters data with agency-level policy variables gathered by to better understand how it works and why some agencies put the U.S. Bureau of Justice to examine whether particular policies it in place while others do not. Our research suggests that the or department features are associated with higher or lower rate paperwork requirement may act as a deterrent to unnecessary use of gun deaths by police officers. Understanding which agency of force because of the added demand on officers' time, the implicit variables are associated with reduced incidence may provide police recommendation from leadership of avoiding unnecessary gun agencies with options to reduce their rates of police-involved deaths draws, and a general commitment to best practices among agencies and help direct policy change by local and state legislative bodies. that have this policy, Co ations, we also find Additionally, it will provide a foundation for future academic that several control variables affected gun death rates by police. For research on an extremely important, although neglected, topic, example, agencies situated in communities with higher levels of economic inequality and lower education levels had more police- Existing scholarly research suggests that the use of force by police involved gun deaths per capita. is affected by factors at multiple levels, including the individual, agency, and community. For example, racial diversity within The Literature: What Predicts the Use of Police Force, communities (Jacobs and O'Brien 1998) and violent crime rates Especially Deadly Force? Smith 2003) have been found to affect the frequency of fatal Past research on what influences officers to use force against police-involved shootings. City-level policies such as the presence civilians typically focuses on one of three major sets of explanatory of a citizen review board have also been associated with police variables or some combination of these: personal or psychological, brutality complaints (Smith and Holmes 2003). Other factors organizational, and situational (Friedrich 1980; Worden 1996). contributing to the use of force more generally include the racial or Personal or psychological explanations center on individual ethnic minority status of the citizen(s) involved, education level of officers and their backgrounds, personalities, and demographic the police officer, income of the police officer, and training efforts characteristics. Organizational explanations focus on policies, by the police agency. Our work here focuses on agency policies structures, and cultures of police agencies that may alter individuals' and characteristics that are associated with relatively low or high tendencies to use force, Situational explanations place emphasis rates of gun deaths by police, although we control for important on the contextual circumstances of an event in which force might community characteristics as well. Recent deaths of civilians by be used, such as the neighborhood, time of day, and race of the police have drawn public attention to strained relations between civilian(s) involved. government and citizens, particularly citizens of color. We seek to illuminate potential Recent deaths of civilians In a two-year study of 80 new police recruits policy options that may help reduce the in an unnamed city, Oberfield (2012) finds rates of these deaths-a normatively positive by police have drawn public that while organizational features, such as outcome from the perspective of leaders, attention to strained rela- training and culture, can have significant police, and citizens. tions between government and effects on officers' attitudes about using force, citizens, particularly citizens of these effects are short-lived, and the officers Much of the past research on the relationship color, return to their previous views. Meanwhile, between police use of force and agency- of personal characteristics were significant community-level factors has relied either on and lasting predictors of views about using case studies of a single police agency (e.g., McElvain and Kposowa force. These included the recruits' views on the use of force prior 2008; Oberfield 2012) or on FBI data on police-involved to entering the police academy, family income, status as former homicides, creating concern regarding the external validity of military, age, and race. At the last point in the survey, for example, findings. This article represents a significant contribution, as we officers who had entered the academy thinking that the use of force leverage a far more complete database of police-involved gun was effective were significantly more likely to hold that view, as were deaths to test which agency policies are associated with fatal police white officers. At various points in the study, higher income was shootings of civilians. The article reviews the literature, lays ou associated with lower levels of agreement with the view that the use hypotheses related to police-involved gun deaths, discusses the of force is effective. This finding is in line with previous research data used for the analysis, and reports results from statistical that supports a relationship between individual characteristics of models. police officers and their behavior. We hypothesize that agencies with certain policies will have lower Among those characteristics associated with higher incidence of rates of gun deaths. These include hiring police forces that better using force are low levels of education and experience of police match the racial makeup of communities, requiring paperwork of officers (Terrill and Mastrofski 2002; Worden 1996), positive views officers who draw their guns and point them at civilians, assigning of the use of force (Oberfield 2012; Worden 1996), and negative regular beats to officers, and requiring more community policing impressions of citizens (Worden 1996), Research on contextual training. Of these policies, we find that only one has a significant or situational factors associated with the use of force finds that relationship with the rate of gun deaths: the requirement that characteristics of the citizens matter. For example, officers were officers document incidents in which they draw their firearms. more likely to use force when citizens were young or poor (Terrill Agencies that required this documentation had lower rates of and Mastrofski 2002), racial or ethnic minorities (Alpert and officer-involved gun deaths on average. This finding was remarkably Dunham 2004; Fyfe 1982; Garner and Maxwell 2002; Meyer robust and quite substantively significant. 1980; Worden 1996), located in disadvantaged neighborhoodsHome Tools Research paper.pdf x ? H 3 /9 (Terrill and Reisig 2003), or viewed as disrespectful or antagonistic homicides by police (Bailey 1996; Smith 2004). That is, agencies (Friedrich 1980; Gallo, Collyer, and Gallagher 2008). Other that required more training hours had higher rates of police- research has shown that heterogeneous communities suffer from involved deaths. Smith (2004) notes that more systematic study lower levels of trust and increased levels of intergroup conflict that of organizational predictors of police killings, particularly with may elevate police-citizen conflict (Putnam 2007; Sampson and respect to training, is needed before drawing conclusions about Groves 1989). the meaning of this finding. Fyfe (1982) finds some evidence that restrictive shooting policies by agencies decreased the number of The Role of Race police-involved deaths in New York City, While nearly all agencies While race has been at the center of many recent conflicts between require documentation of instances when a firearm is discharged, officers and protesters in Ferguson, Baltimore, and New York, only 46 percent require documentation of instances when the previous research has not found that more racially diverse agencies firearm is displayed but not fired (U.S. Department of Justice have different types of arrest behavior (Riksheim and Chermak 2013). 1993), lower instances of police brutality (Cao and Huang 2000), or fewer instances of police-involved deaths (Smith 2003). However, It has become common for police agencies to introduce a variety research has consistently shown that minorities, particularly African of procedures and policies to foster a more positive relationship Americans, have significantly higher rates of interaction with law with communities. Community policing policies and training enforcement (Baumgartner et al. 2016; Harris 1999; Provine are generally considered to decrease crime and strengthen the 2007). Several studies have found that the share of a community's relationship between the community and law enforcement population that is black is associated with higher incidence of agencies (Cordner 2014), While it would be expected that these police-involved deaths and other outcomes, such as traffic stops that positive relationships would result in fewer fatal encounters result in a search (Petrocelli, Piquero, and Smith 2003; Smith 2003, berween police and the community, there is no empirical evidence 2004; Smith and Holmes 2003). These findings support what some to support that hypothesis. Furthermore, it is unclear to what scholars refer to as the "racial threat" hypothesis, which is drawn degree community policing tactics are being fully implemented. from the conflict theory of law and posits that police officers use The 2013 Law Enforcement Management and Administrative force as a response to a perceived threat from racial or economic Statistics show that a large proportion of agencies train recruits groups that are viewed as threats to the existing social order in community policing tactics, but a smaller proportion provide (Harring et al. 1977; Smith 2003), training for nonrookies. Specifically, the share of agencies providing at least half of their recruits with this training is 65 A suggested connection, particularly after the Ferguson protests, is percent, while the share providing the training to at least half of all the relationship between the racial composition of the community full-time officers is 49 percent. and the agency serving the community. Theobald and Haider- Markel (2009) do find a strong connection between the race of the Hypotheses arresting officer and the race of the citizen when considering the We focus our analysis on agency-level policies that we expect may be legitimacy of the arrest. When a white officer arrests a black citizen, associated with the rate of gun deaths by police officers. First, given black citizens are significantly less likely to expectations about the racial match of citizens view the arrest as legitimate. In a study of When a white officer arrests a and police officers and the past finding that Welsh and English police forces, Hong (2017) black citizen, black citizens are black citizens view arrests of black citizens finds that an increase in the proportion of by white officers as less legitimate (Theobald officers who are ethnic minorities is associated significantly less likely to view and Haider-Markel 2009), we expect that with a subsequent reduction in officer the arrest as legitimate. departments with police rosters that are misconduct. Black citizens, Hong finds, make disproportionately white compared with their fewer misconduct complaints against ethnically diverse police forces. citizenry will be associated with more strained community relations and conflict. This increased conflict and reduced legitimacy is The Role of Agency Policies expected to produce higher rates of gun deaths by police. Rosters Evidence about the association between agency policies and the that better march the racial composition of the community served use of force is less developed, but several scholars have found that are expected to have lower rates of these fatalities. police agency decisions can make a difference, In a study of criminal justice reforms after 1950, Walker (1993) finds that agency policies Hypothesis 1: Police agencies with a racial composition of aimed at controlling the discretion of police officers can significantly police officers that better matches the community served will reduce instances of officer-involved shootings. A study on police- have lower rates of gun deaths, involved deaths shows that agencies with more stringent recruitment guidelines had lower incidence of use of deadly force (Willits and By implementing a policy in which officers are required to complete Nowacki 2014). Another shows that officers working in more highly a report any time their firearm is displayed (but not discharged), bureaucratized police agencies were more likely to use "reasonable" agencies are more strictly regulating the use of firearms by officers. force, although there was no relationship between bureaucratization According to our interviews with leaders of police agencies, such levels and use of "unreasonable" force (Worden 1996). policies mean that when officers point their guns at civilians, even momentarily, they must file reports documenting that their guns A perplexing finding in two studies suggests that the required were displayed and explaining the circumstances. These policies number of hours of field training is positively associated with are often nested within agencies' "use of force" policies on the Preventing the Use of Deadly Force: The Relationship between Police Agency Policies and Rates of Officer-Involved Gun Deaths 219Home Tools Research paper.pdf x H ? 4 /9 grounds that pointing a gun at a person is a use of force (personal minimizing the use of force and therefore better able to reduce communications, May June 2016). police-involved deaths. We believe the mechanism connecting firearm display Hypothesis 2: Agencies with a policy requiring a written documentation and officer-involved citizen deaths is a two-step report for instances of the display of an officer's firearm will process. The first step is the documentation requirement's role in have lower rates of police-involved deaths. decreasing the frequency with which officers draw their firearms. The second step is increased officer-involved deaths once a firearm Community policing policies and training are expected to has been displayed by an officer. So, these policies are expected to reduce crime and foster improved relationships between agencies lead to lower rates of gun deaths through their effect of discouraging and the communities they serve (Cordner 2014). We expect unnecessary gun draws by officers. this positive relationship to lead to better interactions between officers and civilians, fewer violent conflicts involving police, We expect the policy to discourage the drawing of officers' guns in and, therefore, reduced incidence of police-involved gun deaths. two key ways. First, by simply instituting the policy and requiring Community policing focuses on building connections between officers to document the display of their police agencies and individuals and groups firearms, policy makers and agency heads in the communities. The idea is to build are recommending that officers take more Community policing focuses on partnerships and trust so that the police seriously the acc of drawing their weapons building connections between agencies and communities are working and pointing them at civilians. Research on police agencies and individuals together for better neighborhoods, If the the implementation of policies requiring and groups in the communities. intended goal is realized, officers will have changes in documentation standards ha more positive connections in the community, shown that people interpret such policies making incidents that result in gun death less as recommendations from policy makers (Johnson and Goldstein likely. Community policing policies are associated with police beats, 2003; Madrian and Shea 2001). In the case of firearms, while each in which officers are assigned to regular neighborhoods for long officer may interpret this implicit recommendation differently durations. Under these conditions, we expect officers to know the requiring documentation when a firearm is displayed becomes a de community members better, to feel more connected to them, and facto policy discouraging the unnecessary use of firearms. therefore to be more likely to hesitate to use deadly force in certain situations. Second, in decision making. people give more weight to potential losses than to potential gains (Tversky and Kahneman 1991) Hypothesis 3: Requiring community policing training will Humans tend to be loss adverse (Thaler and Sunstein 2008). lead to lower rates of police-involved deaths. Most people view paperwork as both a loss of time and a negative experience. Viewing the documentation as a punishment or a Hypothesis 4: Assigning officers to regular bears will lead to loss of something valuable (time), we can see that the policy may lower rates of police-involved deaths. make officers susceptible to loss aversion. If an officer views the documentation as a loss, then the officer is less likely to draw his or Data and Method her firearm because this action will lead to a certain loss, even if the This article incorporates three data sources: the Fatal Encounters potential gains (increased personal safety of the officer) are more Project (FE), the U.S. Bureau of Justice Statistics' Law Enforcement valuable. Management and Administrative Statistics (LEMAS), and the U.S. Census Bureau's American Community Survey, Data from the three It is also possible that the inclusion of this policy is most common sources were combined to create a single data set in which the unit among agencies that stay current with best practices and trends of analysis is the law enforcement agency. The Fatal Encounters within the policing profession, In fact, in our interviews with Project is a nonprofit organization that has been tracking police- police agency leaders, several noted that their agencies had added involved deaths, with data going from 2000 through the end of this policy in order to maintain or achieve certification by an 2015. Through a combination of paid researchers, volunteers, outside organization or to adopt the recommended policy of a and Internet crowdsourcing, FE has collected records on more consulting firm. For example, the Rome Police Department in than 10,000 police-involved deaths in that 15-year time period. Georgia has required reports for pointing a gun without firing All instances are sourced through media reports, Freedom of for at least 25 years and maintains this policy and other best Information Act requests, or agency reporting.' While this data set practices as part of its commitment to earning certification by cannot claim to be complete or without some errors, it is the most the Commission on Accreditation for Law Enforcement Agencies complete data set of police-involved deaths in the United States. (M. Tyson, private communication, May 31, 2016). In another While the FBI does collect data on police-involved deaths, the case, the Flagstaff Police Department each year tailors state- reporting by agencies is voluntary and some states do not report specific template law enforcement policy produced by Lexipol. data at all (U.S Department of Justice 2015). which includes this policy of mandatory reporting for pointing a firearm (D. Musselman, private communication, May 31, For a variety of reasons, the data presented in this article do not 2016). Agencies adhering to a broad set of best practices, such as represent the total number of police-involved deaths in the FE the firearm display documentation requirement, may also foster database. First, only gun-related deaths are included in our data set, the development of a police force that is more committed to as the other types of deaths are less clearly defined as intentional 220 Public Administration Review . March April 2017? Home Tools Research paper.pdf x H 5 /9 uses of deadly force. For instance, many people in the data set died funded by the U.S. Department of Justice and produced by the as a result of being Tasered. It is difficult to assume that the officer Inter-University Consortium for Political and Social Research, chose to use deadly force in those instances because typically a Taser provides organizational data for more than 3,000 law enforcement does not kill someone. Even murkier were cases of vehicular deaths agencies, including all those with 100 or more sworn officers and and asphyxiation while in police custody. Next, cases in which a representative sample of smaller agencies. The data set allowed multiple law enforcement agencies were involved in the death had for construction of variables on agency policy regarding required to be eliminated. Because the unit of analysis is the agency, these documentation for display of firearm, training in community deaths were not included because of an attribution problem. As we policing, and assignment of officers to a regular beat. It also were primarily interested in local agencies, state and federal agency. provided data on racial composition of the agency, which were used related deaths were also eliminated. In a small number of instances to create a variable for the degree to which the agencies match their in which the agency responsible for the death in the FE data could communities. not be matched with an agency in the LEMAS, those agencies were not included in our data set. Finally, for the analysis presented in To rest hypothesis 1, our model includes a measure of white this article, only agencies serving communities with populations representation in the agency compared with white representation over 25,000 have been included." This leaves us with a total of 5,141 in the community. This variable simply takes the percentage of the police-involved deaths across 1,107 law enforcement agencies from community that is white and subtracts the percentage of full-time 2000 to 2015. Our data set is not time serial; rather, it is cross- officers who are white. So, when the number is positive, it indicates sectional with a count of police-involved deaths over the 15-year time that whites are underrepresented in the agency, and when it is period, standardized by population. The data include police agencies negative, whites are overrepresented in the agency. Agencies with for communities of various sizes, including the nation's largest cities negative values, then, have police forces that are less racially diverse and counties, as well as small cities and other municipalities. than the communities they serve. To test hypothesis 2, the model includes a variable indicating the departmental policy on a required The dependent variable in our analysis is the number of officer- report when the officer displays a firearm. As can be seen from involved deaths for every 100,000 people in the community the summary statistics (found in appendix A in the online version served by the police department. This variable ranged from 0 to of this article), about half of the agencies in our data set require a nearly 13, with a mean of 1.75. While many large agencies had written report if an officer displays a firearm. For the purposes of particularly high death rates, other big cities and counties had testing hypotheses 3 and 4, the model also includes variables for the comparably low rates (see table 1). Meanwhile, several of the proportion of officers who receive in-service community policing agencies with the highest death rates were relatively small and training and whether the agency assigns officers to specific bears or medium-sized cities (see table 2). Figure I shows the highest areas of the community. death rates for the largest agencies in our data set, those serving populations over 500,000. We merged the LEMAS police agency data with local-level demographic data on the community served, including population The data used to construct our agency-level independent counts, race, education levels, and income from the U.S. Census variables come from the LEMAS study of 2013. That study, Bureau's American Community Survey using place and county data sets of five-year estimates for 2013. These data were obtained Table 1 Top 10 Agencies by Population of Community Served using the National Historical Geographic Information System (Minnesota Population Center 2016).' Control variables created State Population Death Rate (percent) from these data include the percentage of the population that Los Angeles County Sheriff California 9.893,481 New York City Police New York 8,268,999 is black, percentage of the population with a higher education Cook County Sheriff illinois 5,212,372 D.04 degree, a Gini index of income inequality, and an index of Harris County Sheriff Texas 4, 182,285 1,29 racial homogencity. Past work suggests there may be a positive Los Angeles Police California 3,827.261 5.77 relationship between the percentage of the community that is black San Diego County Sheriff California 3. 138,265 0.80 Orange County Sheriff California 3,051,771 0.16 and the incidence of police use of force because of a perceived Chicago Police 2,706, 101 4,62 hreat to the existing social order. We also expect poorer, more Miami-Dade County Police Florida 2.549.075 341 racially diverse, and more economically unequal localities to Dallas County Sheriff Texas 412,481 experience higher rates of fatalities by police. Given that police shootings of civilians occur more frequently in larger cities (Willits Table 2 Top 10 Agencies with Highest Police-Involved Gun Death Rates and Nowacki 2014) and that many studies have restricted analysis State Death Rate (percent) Population to large cities (e.g., Smith 2003; Smith and Holmes 2003), we New Orleans Police Louisiana 12.60 357.013 include a control variable for log population. We expect larger cities Redding Police California 12,17 30,370 by population to have higher rates of gun deaths per capita. We Inglewood Police Calllomia 11.75 110,585 also include a control variable for the percentage of the population Anderson Police Carolina 11 24 26,599 Midland Police Texas 11.22 that lives in an urban area. All else being equal, we expect less Sierra Vista Police Arizona 11.07 45, 187 urban communities to be associated with lower rates of gun deaths Jackson Police Mississippi 173,997 by police. Finally, the percentage of the community that is non- Tucson Police Arizona 10.89 523.278 Hispanic white also was used for creation of the agency-level Salt Lake City Police Utah 10.63 188, 14 Suisun City Police alifornia 10.58 28,350 variable that measures the racial composition match between the agency and the community. Preventing the Use of Deadly Force: The Relationship between Police Agency Policies and Rates of Officer-Involved Gun Deaths 221Home Tools Research paper.pdf x H ? T 6 /9 Table 3 OLS Model Results, with State Fixed Effects: Officer-Involved Death Rates In Communities with Pupuletions over 75,000 TUCSON POLICE DEPT Coefficient Standard Error Departmental Variables City department 2.619* 0.227 PHOENIX POLICE DEPT Firearm display report -0.327* 0,120 Community policing training -0.059 0.049 Officer beat assign 0,118 0.148 1% whites overrepresented -0.008 0.006 0.003 0.031 ALBUQUERQUE POLICE DEPT 56 of department female Community Variables Inequality Gini 0.087- 0.015 8.2 0THAT % of district urban -0.019* 0.005 JACKSONVILLE SHERIFF'S OFFICE Education -0.033. 0.00 Racial homogeneity -0.002 0.007 % of district black 0.007 0.006 7.9 TripTiri Population (lag) 0.524* D.087 FRESNO POLICE DEPT Constant 5.696* 1.404 =1,102 | Adjusted R = .351 | * p<.05 tatttet portland police bureau analyze the data. because state laws and rules are likely to affect outcome variable we include stare fixed effects it should also be noted that our dependent is constructed from data ranging houston dept while agency community independent variables therefore possible certain policies were enacted between these baltimore cases would obfuscate impact of policy in findings. this nonsignificant results cannot interpreted as a irrii1 clear finding ineffectiveness. angeles att79 shown table four expected associated with higher or lower gun death rates one turned seattle out significant predictor rates: figure highest among large agencies requiring written report for drawing pointing firearm over consistent hypothesis reporting requirement had negative relationship rates. officers provide documentation instances which control city departments. compared they display their was decrease other types departments tend cover all number fatalities per population. given territory within boundaries. county mean effect substantively typically layered may patrol only significant. do not currently have covered by force. if officer- expect more most involved instead findings suggest densely populated parts region interactions period at least fewer people died civilians incidence police-involved officer-involved shootings. shows potential lives saved enacting model includes percentage full-time who female. research has generally contrary expectations communities forces found female less accused excessive larger shares white than general population force harrington wood use deadly did systematically deaths. similarly kposowa required policing training nonrookie percentages those assign regular beats deaths summary statistics thus i appendix online version supported article cable describing each provided b online. binary positive logged rate fairly community. well situated continuous range ordinary squares regression boundaries police. public administration review . march april home tools paper.pdf x h rate. additionally racial homogeneity measures tucson degree dominated single race no effect. producing null phoenix inherent traits an attributes ittri11 inequality education deft easy implement significantly levels fatal officer discussion demonstrate controlling host important mandating reports firearms los place nearly half set. interviews leaders several randomly selected indicated some decades others put recently match changing standards best practices law enforcement. chicago circumstances recommended lexipol provider manuals updates individualized musselman private communication oklahoma pouce certification such offered commission on accreditation enforcement tyson created controversy denver ranks contacted opposed unions associations cities daley m. d. communications san francisco note: serving areas populations hypothesized rule discourage carr according lemas. frawing guns order avoid onerous paperwork follow recommendation leadership. estimates draws weapon reduction opportunities thereby decreasing analysis supports association although additional will establish causality. another reason optimism about possibility using educational attainment reduce civilian success could living urban build time habits caution displaying weapons. habit formation educated however breaking when established informational incentives runs expectations. suggests holding constant whether sys- effective economic furthermore department cematically police- time-restrained decisions much like ones frequently asked make particularly susceptible routines difficult break et al. prone characteristics served previous weapons even suggested increase necessary help them habit. black residents due perceptions threat existing social natural follow-up question benefit endanger preventing force: causing hesitate life-threatening result conservative estimate situations. address ran separate ol.5 same earlier but switching gy on-duty further fully confirm causal themselves. relationships perhaps u.s. justice follows through its promise collecting accurate extremely rare events comprehensive detailed time-serial c project can add evidence dacabase reliable information incident-level factors including victim. specificity assignment allow direct testing predictors beat system create personal study allowed. experiments connections conducted controlled setting examine making link. laboratory been showing familiar bias reactions members citizens repeat offenders wanted greenwald oakes hoffman example knows test someone neighborhood warrant potentially dangerous situations assigned demand quick reactions. recognize confront pursue person. case recent presented here relied raleigh north carolina combination sets never before used patterns uses confirmed bears sorts long-standing assumptions regarding importance counteract proven benefits comes increased evaluation simple yet powerful policy. neighborhoods. beyond scope able definitive conclusion note building population- provides strong adoption standardized aggregate total span creates limitations. does correlate point dying gunshot line duty. covers change saving claims changes without putting risk. demographics during frame meaningful compare snapshot notes combining methodology encounters see hup multiple years into agency-level w diminish randomness size variable. smaller two included very small year appearance pattern ppear high cut pain truly present. aggregating so greatly limit agencies. seized sensitivity any observed points car related substantial factors. held across iterations. problems inability town township poll ments excluded reverse census organi localities separately occurred began implementing municipalities. there lemas set cutoff inclusion mandatory display. case. references contributor decision require alpert geoffrey roger dunham. understanding farce. new find york: cambridge university press negative. bailey william c. less-than-lethal police-citizen killings extent select areas. crime delinquency>