Analyzing Fatal Police Shootings: The Roles of Social Vulnerability, Race, and Place in the US

Abstract

Introduction:

Social vulnerability, race, and place are three important predictors of fatal police shootings. This research offers the first assessment of these factors at the zip code level.

Methods:

The 2015-2022 Mapping Police Violence and Washington Post Fatal Force Data (2015-2022) were used and combined with the American Community Survey (2015-2022). The social vulnerability index (SVI) was computed for each zip code by using indicators suggested by CDC, then categorized into low-, medium-, and high-SVI. The analytical file included police officers who fatally shot 6,901 individuals within 32,736 zip codes between 2015-2022. Negative Binomial Regression (NBRG) models were run to estimate the association between number of police shootings and zip code SVI, racial composition, and access to guns using 2015-2022 data.

Results:

Moving from low-SVI to high-SVI revealed the number of fatal police shootings increased 8.3 times, with the highest increases in Blacks (20.4 times), and Hispanics (27.1 times). The NBRG showed that moderate-, and high-SVI zip codes experienced higher fatal police shootings by 1.97, and 3.26 times than low-SVI zip codes; zip code racial composition, working age population, number of violent crimes, number of police officers and access to a gun, were other predictors of fatal police shootings.

Conclusions:

Social vulnerability and racial composition of a zip code are associated with fatal police shooting, both independently and when considered together. What drives deadly police shootings in the United States is not one single factor, but rather complex interactions between social-vulnerability, race, and place that must be tackled synchronously. Action must be taken to address underlying determinants of disparities in policing.

Introduction

In 2022, an estimated 1,096 people were fatally shot by police in the United States.¹ Those who are fatally shot by police are disproportionately Black and Hispanic, males, or living in areas with high rates of unemployment, social vulnerability, and concentrated disadvantage.^2–4 96% of victims of fatal police shootings were males, with 71% aged between 18 and 44.¹ Black individuals faced nearly 3 times greater likelihood of fatal police shootings, compared to 1.3 times for Hispanics, 3.9 times for Native Hawaiians/Pacific Islanders, and 1.7 times for Native Americans, as opposed to White individuals.^1,2 This disparity amplifies adverse effects on mental health and physical well-being, and an increase in premature mortality in communities of color.⁵

Place-based factors and structural racism play essential roles in individual-level racial disparities.^6,7 Historical policies like redlining and structural racism have fueled residential segregation, leading to neighborhoods of concentrated disadvantage with combined social vulnerabilities. These compounding vulnerabilities exacerbate racial disparities beyond individual factors.^3,7 For instance, high Social Vulnerability Index (SVI) counties show a 5.8% increased risk of fatal police shootings compared to low SVI counties.⁴ This disparity is pronounced for minoritized groups, with fatal shootings rising by 2.3x for Whites, 9.6x for Blacks, and 15x for Hispanics in high SVI counties versus low SVI counties.⁴

Numerous studies support the community violence hypothesis, which attributes increased policing and police shootings to heightened violent crime rates, necessitating greater police presence and potentially leading to more fatal outcomes.^3,8–10 Other theories have sought to explain disproportionate levels of fatal police shootings across communities. The conflict theory of law proposes that policing serves as a tool for social control, power differences that characterize social interactions of those in power.^3,11–13 The minority threat hypothesis and group-threat theory suggest that racially minoritized and socially marginalized populations face more aggressive policing due to posing a perceived threat to the status quo.^3,13–15

The place hypothesis suggests that racially minoritized groups in economically disadvantaged, segregated areas are perceived as threatening by police, amplifying interactions and risks. Scholarly opinions differ on the drivers of these disparities, highlighting social vulnerability’s pivotal role in police violence exposure and risk.^{3,4,6,16–20}

Existing research efforts have found associations between fatal police shootings and various measures of social disorganization and community violence, including a few recent studies that examined multiple indicators of police violence over more extended time periods.^{3,4,14,21–29} However, the majority of studies have focused on singular outcomes within limited time frames and larger geographic units like cities, counties, and states. ^4,21–24 A literature gap persists regarding the link between SVI and police violence on a zip code-level. Zip code-level investigations have explored various outcomes in restricted geographic scopes and demographic groups.^20,25,26

This study analyzes nationwide zip code-level data on fatal police shootings from 2015 to 2022, emphasizing social vulnerabilities. From the public health perspective, the social determinants of health play a significant role in predicting overall well-being, where people live can significantly impact their life expectancy.³⁰ The research explores differences in fatal police shootings based on victim race, SVI, racial composition, and public safety indicators. The conceptual framework was adopted, which outlines factors on the individual, community, and societal levels that impact fatal shooting risk, for the development of this study and interpretation of its findings (Appendix 1). Police violence is multifaceted. This research describes the individual associations of specific indicators of police violence on fatal police shootings and how these associations change in the presence of other factors. The main study aim is to enhance understanding of social vulnerabilities’ impact on fatal police encounters and factors influencing shooting risks on individual, community, and societal levels.

Methods

This is a secondary analysis of two primary datasets: The 2015-2022 Mapping Police Violence Data (MPV)³¹ and Washington Post Fatal Force Data (WaPo).³² The MPV has been tracking fatal police shootings since 2013, while WaPo has been recording all on-duty police shootings since 2015. The precise address of the incident was provided by the MPV Individuals were included in the analysis if both datasets reported the incident, and data were merged with independent variables to compute Social Vulnerability Index (SVI), racial demographics, age distribution, and total population from the U.S. Census Bureau’s American Community Survey.³³ County-level public safety factors, such as number of violent crimes and number of police officers, were also obtained from the 2015-2022 County Health Rankings (CHR) ³⁴ and Law Enforcement Officers Killed and Assaulted (LEOKA)³⁵, respectively.

Study Population.

The final dataset included 6,901 individuals who had been fatally shot by police between 2015 and 2022 within 32,736 zip codes, zip codes with no fatal police shootings were imputed with zeros. The unit of the analysis was zip code-years.

Measures.

The dependent variable was the total number of fatal police shootings at the zip code level computed from MPV and WaPo data between 2015-2022. Fatal police shooting comprised fatalities in which gunshot was the cause of death. The combined race and ethnicity of the victim was first extracted from the WaPo, then from the MPV if missing from the WaPo.

The study utilized the Social Vulnerability Index (SVI) as the primary independent variable, calculated following the Centers for Disease Control’s method ³⁶ and using the [ACS-5Y; 2013-2017 has being used as a proxy for measures from 2015-2017 and ACS-5Y; 2017-2021 for 2018-2022] American Community Survey. The Social Vulnerability Index (SVI) with 15 variables was utilized, excluding the “% of minority population”, then SVI scores were categorized into low-SVI (SVI<0.34), medium SVI (0.34≤SVI<0.67), and high-SVI (SVI≥0.67) for interpretability of results.^4,37 To address overlapping ZIP codes across different counties, the population ratio of ZIP codes in different counties was calculated, and the ZIP codes to the counties with the majority population were assigned.

Zip codes were classified as majority Non-Hispanic White (≥65% NHW), racially integrated (35% ≤ NHW < 65%), and racially minoritized (<35% NHW).³⁸ A few studies have highlighted the impact of segregation, which detrimentally affects poor Black individuals while benefiting poor White individuals. ^27,39–42

Models controlled for population age groups (0-17, 18-24, 25-34, 35-49, 50-64, and 65+ years old) and density, county-level violent crimes and violent crimes per officer, and year of the incident. Previous studies (Miller, Grennan) explored the impact of female police representation in conflicts, and full model controlled for the percentage of female officers in a county.^43,44 For 2021 and 2022 RAND dataset on State-Level Estimates of Household Firearm Ownership was used ⁴⁵ and number of registered weapons U.S. by Statistica ⁴⁶ to create a categorical variables for having access to gun from low to high.

Statistical Analysis.

Three sets of zip code level analyses were run. First, a descriptive analyses to compare zip codes by SVI category using independent t-tests with unequal variances and chi-square tests. Second, Negative Binomial Regression (NBRG) models assessed the association between the number of police shootings and zip code SVI categories. Third, several analyses using SVI as a continuous variable were performed and controlling for population characteristics at the zip code and county-level and year of the incident. Finally, to examine how zip code racial composition moderates the relationship between SVI categories and fatal shootings, an interaction term between SVI categories and racial composition categories was included. The analysis utilized data sets from 2015 to 2022.

To produce more robust estimates, these steps were followed: 1) Evaluated variance inflation factors (cutoff of 10) to address multicollinearity with a less than 1.8 value indicating low multicollinearity.⁴⁷ Sequential variable addition confirmed results stability. 2) Employed the Bonferroni adjustment to mitigate multiple testing issues, setting the statistical p-value at 0.002.⁴⁸ (3) Zero observations in certain zip codes led to data skewness, and normal approximation was less likely to be satisfactory.⁴⁹ After assessing the overdispersion parameter alpha, a Negative Binomial Regression (NBRG) model clustered at the county level was chosen as appropriate for the research objectives to calculate the Incidence Rate Ratio (IRR) of fatal police shootings. The chi-squared value of 291.43 for the full model indicated alpha as non-zero, reinforcing the appropriateness of NBRG. Stata version 17 was the software platform utilized for all analyses.

This study was approved by the Institutional Review Board Office of the Johns Hopkins Bloomberg School of Public Health.

Results

Examining 6,901 individuals in 32,736 zip codes from 2015 to 2022, the study observed an average SVI of 0.51 (SD: 0.28), with scores of 0.17 (0.10) for low, 0.51 (0.11) for moderate, and 0.83 (0.1) for high-SVI categories. Fatal police shootings rose significantly, by 8.3 times, from 557 to 4,622 individuals as SVI levels increased from low to high, with the most notable spikes seen in Hispanic (27.1 times) and Black (20.4 times) populations. Unequal variance t-tests highlighted significant discrepancies between SVI states across various factors, excluding violent crimes per officer and access to guns. Refer to Table 1 and Appendix 2 for detailed population characteristics.

Table 1.

Number of fatal police shootings, population characteristics, and public safety factors at zip codes

Variables	Social vulnerability states
	Low	Moderate	High	All
	n = 10,740	n = 10,993	n = 11,003	n = 32,736
	Mean	Mean	Mean	Mean
Number of fatal police shootings by race (N)
White	417	1,155	1,990	3,562
Black	72	284	1,468	1,824
Hispanic	39	184	1,059	1,282
Other	29	99	105	233
Total	557	1,722	4,622	6,901
Population age categories (%)
Age 0-17	18.96	21.09	23.14	21.08
Age 18-24	8.73	7.97	8.97	8.56
Age 25-34	10.17	11.45	12.88	11.51
Age 35-49	17.28	17.92	17.88	17.7
Age 50-64	24.28	21.99	19.85	22.02
Age 65 over	20.58	19.57	17.28	19.13
Racial composition categories (%)
Majority White	0.9	0.85	0.46	0.74
Integrated	0.05	0.12	0.29	0.15
Racially minoritized	0.05	0.03	0.25	0.11
Racial composition (%)
White NH	89.4	82.9	58.2	76.7
Black NH	2.6	4.6	15.1	7.5
Hispanics	4	6.6	17.5	9.4
Other Race NH	3.9	5.9	9.2	6.4
Population density (Mean/SD)	4,338(8,897)	9,814(13,845)	15,562(17,632)	9,904(14,660)
Violent crime ratio (Mean/SD)	245(188)	267(193)	364(233)	292(213)
Violent crime ratio categories (%)
Low	0.33	0.27	0.14	0.25
Medium	0.27	0.28	0.20	0.25
High	0.22	0.25	0.28	0.25
Very high	0.17	0.2	0.38	0.25
Violent crimes per officer (Mean/SD)	149(276)	172(848)	443(4,169)	254(2,459)
Violent crimes per officer categories (%)
Low	0.33	0.26	0.16	0.25
Medium	0.26	0.27	0.22	0.25
High	0.21	0.24	0.29	0.25
Very high	0.2	0.23	0.32	0.25
Female officers (%)	0.09	0.09	0.1	0.09
Access to gun (%)
Number of registered weapons at state (1,000), (Mean/SD)	25(25)	23(19)	24(15)	24(20)
Number of registered weapons at state, categories (%)
Low	0.26	0.27	0.24	0.26
Medium	0.28	0.26	0.21	0.25
High	0.23	0.25	0.26	0.25
Very high	0.23	0.21	0.29	0.25
Number of zip codes per year	10,740	10,993	11,003	32,736
Number of observations	87,293	87,295	87,289	261,877

Notes:

¹⁾The Bonferroni adjustment was used to account for multiple significance tests. The statistical p-value (alpha) was set to 0.002 (0.05/26=0.0019), there were significant differences among SVI categories (p<0.001).

²⁾The analysis utilized data sets from 2015 to 2022.

Table 2 reports the IRR of NBRG. In unadjusted models, the Incidence Rate Ratio (IRR) of fatal police shootings increased significantly in moderate-SVI (2.12 times) and high-SVI (4.1 times) zip codes compared to low-SV areas. The IRR rose in racially integrated (1.78x) and racially-minoritized (1.86x) zip codes compared to White zip codes. After adjusting for racial composition, IRR decreased slightly for SVI zip codes. Inclusion of population age showed no change for low-SVI and high-SVI zip codes, and the significance for racially minoritized areas lessened. Association with fatal shootings remained for racially integrated zip codes.

Table 2.

Association between SVI, Race and population factors on number of fatal police shootings in zip codes

Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
	IRR [95% CI]	IRR [95% CI]	IRR [95% CI]	IRR [95% CI]	IRR [95% CI]	IRR [95% CI]
SVI (Ref. Low)
Moderate	2.120*** [1.913-2.350]		2.072*** [1.868-2.297]	2.032*** [1.831-2.255]	1.974*** [1.755-2.222]	1.935*** [1.585-2.363]
High	4.100*** [3.714-4.526]		3.781*** [3.440-4.157]	3.755*** [3.393-4.155]	3.256*** [2.883-3.676]	3.369*** [2.764-4.107]
Racial composition categories (%) (Ref. Majority White)
Integrated		1.776*** [1.607-1.963]	1.332*** [1.206-1.472]	1.192*** [1.080-1.316]	0.998 [0.787-1.266]	1.075 [0.563-1.502]
Racially minoritized		1.862*** [1.599-2.169]	1.190* [1.022-1.386]	1.079 [0.927-1.256]	0.345*** [0.196-0.608]	0.252* [0.063-0.999]
Population age categories (%) (Ref. 0-17 and 65+ years old)
18-24 years old				1.008*** [1.003-1.012]	1.006** [1.001-1.010]	-
25-34 years old				1.031*** [1.025-1.037]	1.027*** [1.021-1.033]	-
35-49 years old				1.015*** [1.008-1.021]	1.014*** [1.009-1.020]	-
50-64 years old				0.995 [0.989-1.002]	0.994 [0.988-1.000]	-
Violent crime ratio (%) (Ref. Low)
Medium					1.195* [1.014-1.409]	1.276 [0.957-1.702]
High					1.558*** [1.281-1.895]	1.457* [1.098-1.934]
Very high					1.788*** [1.439-2.223]	1.776*** [1.335-2.362]
Violent crimes per officer (%) (Ref. Low)
Medium					1.290* [1.046-1.590]	1.222 [0.943-1.582]
High					1.104 [0.859-1.419]	1.041 [0.800-1.542]
Very high					1.201 [0.936-1.542]	1.175 [0.896-1.542]
Female officers (%)					1.749 [0.440-6.954]	2.078 [0.485-8.903]
Interaction Race and SVI
Medium SVI and Racially Integrated					0.995 [0.783-1.263]	1.042 [0.584-1.585]
Medium SVI and Racially Minoritized					2.111* [1.086-4.104]	2.863 [0.642-12.773]
High SVI and Racially Integrated					1.160 [0.909-1.479]	1.027 [0.628-1.680]
High SVI and Racially Minoritized					3.023*** [1.645-5.556]	4.457*** [1.099-18.076]
Access to gun (Ref. Low)
Medium						1.989*** [1.674-2.365]
High						1.838*** [1.504-2.245]
Very high						2.122*** [1.717-2.624]
Total Population	Included	Included	Included	Included	Included	Included
Year	Included	Included	Included	Included	Included	Included

^*p<0.05,

^**p<0.01,

^***p<0.001

Note: The analysis utilized data sets from 2015 to 2022, for models 1-5. The population age categories were excluded from the model 6, due to the use of data from two years (2021 and 2022), There is an anticipation of minimal variation in age distribution at the zip code level within this two-year timeframe.

In the full model, residents in medium-SVI zip codes had an incident rate ratio of 1.98 (CI: 1.79-2.19), and those in high-SVI zip codes had an IRR of 3.53 (CI: 3.19-3.90), indicating a higher incidence of police fatal shootings. Additionally, the analysis revealed that the frequency of police fatal shootings increased in zip codes with a higher proportion of young adults aged 18-24 years old (IRR: 1.01, CI: 1.00-1.01), 25-34 years old (IRR: 1.03, CI: 1.02-1.03), and middle-aged individuals aged 35-49 years old (IRR: 1.01, CI: 1.01-1.02).

Moreover, the presence of violent crime in a zip code was found to be associated with an increase in fatal police shootings. For instance, zip codes with very high levels of violent crimes experienced 1.79 times more police fatal shootings (CI: 1.44-2.22). Additionally, residents in zip codes with medium levels of violent crimes per officer had a higher incidence of police fatal shootings (IRR: 1.29, CI: 1.05-1.59).

Results from the models with SVI as a continuous variable are provided in Appendix 3. The predicted average number of events for SVI stands at 0.066 (CI: 0.059-0.074). Comparing the predicted event counts at two different levels of SVI with a 10-point change (SVI = 0.5 and SVI = 0.6), the ratio increased by 1.25 (0.026/0.021).

Table 2, Model 5 presents results of interaction between SVI and racial composition categories. Compared to residents in zip codes characterized by low SVI and a majority-White composition, individuals residing in racially minoritized zip codes with moderate SVI had an IRR of 2.11 (CI: 1.08-4.10), while those in high-SVI zip codes had an IRR of 3.02 (CI: 1.64-2.56).

Using 2021 and 2022 data, the association between gun access and fatal police shootings was investigated (Table 2, Model 6). Compared to low-access areas, fatal shooting rates were higher in zones with medium (IRR: 1.98; CI: 1.67-2.36), high (IRR: 1.83; CI: 1.50-2.24), and very high (IRR: 2.12; CI: 1.71-2.62) access to firearms, even after adjusting for covariates.

Figure 1 presents the relationship between SVI, individual race and ethnicity, zip code racial composition, and fatal police shootings. Regardless of race, fatal police shootings increased as social vulnerability increased. White individuals experienced the highest rate of fatal police shootings in majority white zip codes, whereas Black and Hispanic individuals had the highest rate of fatal police shootings in racially integrated areas (See Panels A, B and C).

Figure 1.

Impact of SVI on number of fatal police shootings by racial composition of zip codes

Overall, fatal shootings increased when moving from low-SVI to high-SVI areas; however, with a sharp increase for Hispanic and Black individuals. For instance, with an average SVI of 0.5, the predicted number of fatal police shootings for White, Black, and Hispanic individuals were 0.015, 0.31, and 0.34, respectively. Moving to an SVI of 0.8, these numbers increased to 0.031, 0.064, and 0.071, respectively (Panel D). See Appendix 4.

In Table 3, the IRR of fatal police shootings is compared among White, Black, and Hispanic individuals. Social vulnerability shows a positive association with fatal police shootings for all racial groups, with the highest IRR in Blacks, followed by Hispanics and Whites. For instance, the shooting ratio increased by 4.41 times in high-SV zip codes compared to low-SV areas, with ratios of 3.6 for Hispanics and 3.4 for Whites. Hispanics had the highest IRR (3.89), followed by Blacks (2.82) and Whites (0.23). Population age, violent crime percentage, and female police officers were predictors of fatal shootings for all groups, with female officers increasing IRR among Blacks. A separate model was conducted by excluding the exposure variable (population size), finding showed a higher IRR of being fatally shot in racially minoritized areas for White, Black, and Hispanic populations. Additionally, the percentage of female officers was a predictor of fatal shootings in the white population. (Appendix 5)

Table 3.

Association of SVI and population factors on fatal police shootings across different racial groups

Variables	White	Black	Hispanics
	IRR [95% CI]	IRR [95% CI]	IRR [95% CI]
SVI (Ref. Low)
Moderate	1.963*** [1.733-2.224]	2.560*** [1.927-3.399]	2.434*** [1.819-3.258]
High	3.540*** [3.110-4.031]	4.828*** [3.724-6.260]	3.772 *** [2.833-5.023]
Racial composition categories (%) (Ref. Majority White)
Integrated	1.0310 [0.923-1.152]	2.320*** [1.955-2.754]	3.323*** [2.626-4.204]
Racially minoritized	0.937 [0.796-1.102]	2.325*** [1.920-2.817]	2.970*** [2.070-4.259]
Population age categories (%) (Ref. 0-17 and 65+ years old)
18-24 years old	1.004 [0.999-1.010]	1.011** [1.004-1.018]	0.996 [0.987-1.004]
25-34 years old	1.017*** [1.010-1.025]	1.045*** [1.037-1.052]	1.030*** [1.018-1.043]
35-49 years old	1.018*** [1.012-1.025]	1.031*** [1.023-1.038]	1.013 [0.997-1.030]
50-64 years old	1.001 [0.995-1.008]	0.991 [0.981-1.000]	0.969*** [0.951-0.987]
Violent crime ratio (%) (Ref. Low)
Medium	1.152 [0.975-1.360]	1.644** [1.209-2.238]	2.265*** [1.441-3.561]
High	1.440*** [1.189-1.745]	2.006*** [1.421-2.832]	3.175*** [1.841-5.474]
Very high	1.426** [1.145-1.775]	3.521*** [2.476-5.007]	3.476*** [2.033-5.941]
Violent crimes per officer (%) (Ref. Low)
Medium	1.222* [1.001-1.491]	1.324 [1.000-1.752]	0.911 [0.604-1.374]
High	1.035 [0.825-1.299]	1.060 [0.807-1.391]	1.420 [0.883-2.285]
Very high	1.212 [0.967-1.520]	1.066 [0.779-1.458]	1.298 [0.804-2.095]
Female officers (%)	0.671 [0.189-2.378]	9.232* [1.644-51.831]	1.527 [0.103-22.593]
Total Population by race	Included	Included	Included
Year	Included	Included	Included

^*p<0.05,

^**p<0.01,

^***p<0.001

Note: The analysis utilized data sets from 2015 to 2022.

Sensitivity analysis:

Considering overdispersion parameter alpha the NBRG was selected, but the sensitivity analysis using Zero-Inflated Binomial Regression (ZINB) revealed that residents living in areas with moderate to high-SVI, integrated and racially minoritized neighborhoods, areas with medium to high rates of violent crimes, and individuals over 18 years old have a higher relative risk of being fatally shot by police. (Appendix 6)

Discussion

This study investigates the association between the number of fatal police shootings by a zip code’s social vulnerability and whether its racial composition moderates this association. The findings suggest that zip codes’ SVI is a significant predictor of fatal police shooting, and that racial composition moderates the association between SVI and fatal police shooting. These findings are consistent with previous studies, which have found associations between SVI and police violence at the census tract, city, county, and state levels.^4,21–24 Other factors that have been positively associated with the risk of being shot by a police officer include community safety, violent crime rates, the number of police officers, and access to firearms in an area.⁵⁰

The results, in particular the higher rates of fatal police shootings in majority-minority communities and communities with higher SVI, are more in line with the ‘minority threat hypothesis’ and ‘place hypothesis’. In the racial/ethnic-specific model, there was no significant link between racial composition of a zip code and fatal shootings of White individuals, unlike for Black and Hispanic populations. In contrast to the findings of Grennan and Miller, female officers exhibited a higher IRR in fatal shootings involving Black individuals compared to their White and Hispanic counterparts.^43,44 However, after adjusting for population size, a higher IRR was observed in the White, and Hispanics populations by female officers.

Existing studies also support these theories, emphasizing the impact of race and location on fatal police shooting risk. ^21,51–54 The findings further endorse the place, and social disorganization hypotheses at the zip code, indicating increased fatal shootings in segregated, economically disadvantaged areas, as previously mentioned fatal outcomes in majority-minority communities facing concentrated disadvantage. ^3,24,51

The analysis revealed a positive correlation between gun access and fatal police shooting rates. This aligns with previous research highlighting gun access as a significant predictor of such incidents.^55–57 This association may be explained by increases in confrontations with armed civilians, as the presence of weapons may create more adversarial police-civilian interactions and high gun accessibility may increase the likelihood that officers assume a civilian is armed even when they are not, thus impacting police decision-making.⁵⁵

These findings underscore the critical need to address social vulnerability and racial disparities in police shooting rates, emphasizing the complex interplay of power dynamics and structural injustices in police interactions. A comprehensive approach is essential to foster a more just and equitable criminal justice system and society.²⁴ Policymakers should consider long-term strategies to reduce fatal police shootings and their disproportionate impact on minoritized communities through efforts to reduce social vulnerability, such as addressing income and employment disparities and other longstanding impacts of structural racism.^3,4,20,58 Short-term policies like enhancing police training in use of force, de-escalation, and social interaction, addressing racial biases, implementing body cameras, and promoting community representation within police departments may aid in curbing fatal police shootings.^59–63 However, further research is necessary to assess the effectiveness of these initiatives.⁶⁴

Given the connection between gun access and fatal police shootings, a study has shown that as firearm availability in neighboring counties increases, police-involved fatal shootings within a focal county also increase.²⁹ Policies that address gun access and gun violence may reduce fatal police shootings more specifically in socially disorganized communities, such as restrictions on public carrying, permission to purchase a firearm, more comprehensive background check policies, and community-based violence programs.^56,65

Limitations

The study uses fatal police shooting data from accessible sources (WaPo and MPVD), recognizing limitations in data precision, completeness, and consistency due to varied data collection practices across jurisdictions. The retrospective observational design prevents causal conclusions. A cross-sectional study may encounter issues like endogeneity if local violence and police presence are causally related. Untangling causation in areas with high Social Vulnerability Index is challenging. The study found crime in one area related to behaviors in neighboring units due to spatial dynamics²⁹; in this study the spatial spillover effects has not been controlled.

Conclusions

The results suggest that both social vulnerability and zip code racial composition independently influence fatal police shooting risk, with a combined impact when considered together. Fatal police shootings in the U.S. may result from complex interactions between social vulnerability, race, and geography, rather than one single factor. Effective strategies are crucial to study, describe, and tackle this pressing issue.

Mapping vulnerable areas and analyzing racial demographics is crucial to understanding and addressing disparities in fatal police encounters. A deeper exploration of these factors is essential for effectively addressing the complex dynamics contributing to fatal police shootings in the U.S.