Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Feb 1.
Published in final edited form as: Spat Spatiotemporal Epidemiol. 2023 Jan 13;44:100567. doi: 10.1016/j.sste.2023.100567

Unintended Reductions in Assaults near Sobriety Checkpoints: A Longitudinal Spatial Analysis

Jack Seifarth 1, Jason Ferris 2, Corinne Peek-Asa 3, Douglas J Wiebe 4, Charles C Branas 1, Ariana Gobaud 1, Christina Mehranbod 1, Brady Bushover 1, Christopher N Morrison 1,5,*
PMCID: PMC9896375  NIHMSID: NIHMS1866866  PMID: 36707194

Abstract

Background.

Sobriety checkpoints are a form of proactive policing in which law enforcement officers concentrate at a point on the roadway to systematically perform sobriety tests for all passing drivers. We investigated whether sobriety checkpoints unintentionally reduce assaults in surrounding areas.

Methods.

Exposures of interest were sobriety checkpoints conducted by the Los Angeles Police Department between 2012 and 2017. Comparison units were matched 1:2 to sobriety checkpoints, selected as the same point location temporally lagged by exactly ±168 hours. The outcome was the density of police-reported assaults around the checkpoint location.

Results.

In mixed effects regression analyses, assault incidence was lower when sobriety checkpoints were in operation compared to the same location ±168 hours [b= −0.0108, 95% CI: (−0.0203, −0.0012)].

Conclusions.

Sobriety checkpoints were associated with decreased assault incidence, but estimated effect sizes were small and effects did not endure long after checkpoints ended.

Keywords: assault, sobriety, checkpoint, proactive, policing

1. Introduction

Violent crime is a leading cause of injury and death and is a major contributor to health disparities in the United States (US). A total of 189,949 people were killed in violent incidents in the decade from 2011 to 2020 (CDC, 2022) and a further 38 million were treated in emergency departments (CDC, 2020). This colossal health burden disproportionately affects racial and ethnic minority populations and other disadvantaged people. For example, violent crime is the leading cause of death for Black men aged 15 to 34 (CDC, 2019). Identifying effective preventive interventions that reduce the overall burden and health inequities due to violent crime is a national public health imperative (The White House, 2013; Subbaraman, 2020).

Hot spot policing is a proactive policing strategy that involves deploying police officers to high-crime areas to conduct patrols, searches, and arrests. The strategy can be conceptualized as a preventive intervention that may reduce the overall burden due to violent crime, because empirical studies find evidence that it reduces violent crime in nearby locations (Braga, 2015; Braga et al., 2019); however hot spot policing is often criticized as contributing to—rather than mitigating—health inequities. In support of hot spot policing, multiple systematic reviews suggest the strategy is associated with reduced firearm violence within small geographic areas, such as blocks and block groups, without displacement to other areas (Braga et al., 2014; Braga et al., 2019). The approach is predicated on empirical observations that crime and violence concentrate within geographic locations (Kennedy, 1990; Blakely and Snyder, 1998; Weisburd et al., 2012; Nobles et al., 2013; Weisburd, 2015; Montolio and Planells-Struse, 2016; Cowen et al., 2019; Vildosola et al., 2019; Andresen and Ha, 2020; Baobeid et al., 2021; Thomas et al., 2022). Rationalist theories of crime opportunities (e.g., rational choice theory, routine activities theory) explain that prospective offenders weigh the costs and benefits of committing crimes (Cornish and Clarke, 1987), and that crime occurs with the convergence in space and time of a motivated offender, a suitable target, and the absence of suitable guardianship (Karstedt, 1993). General deterrence theory adds that high concentrations of police raise the risks of detection and arrest, thereby raising the potential costs of violence and reducing the likelihood that prospective offenders will choose to commit a crime (Nagin et al., 2015). Numerous studies suggest hot spot policing is associated with the diffusion of crime control benefits into surrounding areas, rather than displacement (Clarke and Weisburd, 2006; Weisburd et al., 2006; Bowers et al., 2011).

Despite the strong empirical evidence and clear theoretical foundation, hot spot policing has only modest backing in the general population (Metcalfe and Pickett, 2018). Aggressive strategies—such as New York City’s “stop-and-frisk” policy—disproportionately affect people from racial and ethnic minority populations and contribute to deteriorating police-community relations (Gelman et al., 2007; La Vigne et al., 2017; Weisburd, 2016). Contradictory to the intended effect, evidence suggests the vast majority of stops from this form of hot spot policing do not lead to arrest or detection of illegal activity (Dunn and Shames, 2019) and that these practices cause stress and disempowerment, as well as carrying the possibility of escalating violence (Geller et al., 2014; DeVylder et al., 2017; McLeod et al., 2020). In contrast, situational interventions and other “soft” hot spot policing strategies emphasize respectful interactions to build community trust and support while aiming to reduce violence (Rios et al., 2020). Recent evidence identifies that the frequent presence of law enforcement officers can achieve prosocial outcomes, even where arrests are few (Ariel et al., 2016).

Sobriety checkpoints are a proactive policing intervention that aims to reduce alcohol-involved motor vehicle crashes and could inadvertently reduce violent crime, specifically assault. Within this intervention, a team of law enforcement officers establishes a highly visible roadside checkpoint to stop drivers for a possible sobriety test. The strategy has been used in the US since the early 1980s (National Highway Traffic Safety Administration, 1983; Community Preventive Services Task, 2014). In 1990 the Supreme Court ruled that the approach did not impinge on drivers’ civil liberties, provided police conducted sobriety tests only among drivers whom they had reason to suspect may have been drinking and that communities were given advanced notice of planned checkpoints (Michigan State Police v. Sitz, 1990). Sobriety checkpoints are currently permitted in 38 states and used in 25 states (Alert Media Inc., 2022). Like hot spot policing, the intervention is grounded in deterrence theory, wherein general deterrent effects among the whole population of drivers can be separated from specific deterrent effects among people who pass through the checkpoints (Homel, 1993). Specific deterrent effects are thought to be small because arrest rates at sobriety checkpoints are generally very low (Voas, 2008). Instead, the primary mechanism is general deterrence, where drivers who become aware of checkpoint operations (e.g., by driving past, through social media, by word-of-mouth) change their future behavior in response to the increased perceived risks of detection. Fewer people choosing to drive after drinking leads to decreased incidence of alcohol-involved crashes (Homel, 1993; Ferris et al., 2015). Four systematic reviews conclude that sobriety checkpoints are an effective and cost-effective intervention to reduce alcohol-involved crashes (Peek-Asa, 1999; Shults et al., 2001; Erke et al., 2009; Bergen et al., 2014). Sobriety checkpoints reduce alcohol-involved crashes by 17% and all-cause crashes by 10–15% (Erke et al., 2009). Our own research identifies that each sobriety checkpoint is associated with fewer alcohol-involved motor vehicle crashes across the extent of cities, including Los Angeles, CA, and, consistent with general deterrence theory, these effects are detectable for approximately one week after the checkpoint date (Morrison et al., 2021; Morrison et al., 2022).

In sum, violent crime is a major contributor to the injury burden and health inequity in the US, soft proactive policing strategies are associated with relative reductions in violent crime and can contribute to prosocial outcomes and positive police-community relations, and sobriety checkpoints have many parallels with other soft policing strategies. The aim of this study was to examine the unintended effects of sobriety checkpoints on assaults. We hypothesized that sobriety checkpoints would be associated with decreased incidence of assaults in surrounding areas. We selected the city of Los Angeles because the Los Angeles Police Department conducts more checkpoints annually that any other US city (CDC, 2015).

2. Methods

2.1. Study Setting

This longitudinal study examined associations between the presence of sobriety checkpoints and the geographic incidence density of police-reported assaults within the City of Los Angeles, California, from 2012–2017. We excluded independent cities that were wholly contained within Los Angeles—such as Santa Monica and West Hollywood—yielding a study area that covered 503 square miles and had a 2017 population of 3.95 million.

2.2. Study Design

This study used a space-time panel design which we developed in prior analyses relating temporary spatially structured exposures to injury outcomes (Morrison et al., 2018). We conceptualized a study universe as all geographic points along the City of Los Angeles roadway network, partitioned into 3,156,480 minutes for 2012–2017. A sobriety checkpoint occurs at a single geographic point (expressed as latitude-longitude coordinates) over multiple consecutive minutes, and this space-time location represents the event unit. Given that the spatial counterfactual to this intervention (i.e., the same space-time location without the sobriety checkpoint) was unobservable, two viable possibilities remained for selecting comparison units—the same spatial location at a different time, and a different spatial location at the same time. Selecting the same location at different times was the best available solution because spatially varying confounders are held constant within geographic points (e.g., roadway design, proximity to alcohol outlets)(Schedler and Ensor, 2021). Selecting different locations at the same time was also possible, but contamination due to spillover of the intervention effects meant internal validity would be questionable (Benjamin-Chung et al., 2018). That is, sobriety checkpoint operations are advertised across the extent of the study area, and the advertised police activity could affect behavior elsewhere in the city.

2.3. Sobriety Checkpoints

The Los Angeles Police Department (LAPD) has a comprehensive program to prevent driving under the influence of alcohol (DUI). With funding from the California Office of Traffic Safety through the National Highway Traffic Safety Administration, and in accordance with Michigan Dept. of State Police v. Sitz, LAPD advertises the location and time of sobriety checkpoints in press releases posted to the LAPD website (LAPD, 2021b). Sobriety checkpoints are announced as occurring at intersections (e.g., “Hollywood Boulevard and Bronson Avenue”) during a specific date and time period (e.g., “Friday, January 20, 2012, 8 P.M. to Saturday, January 21, 2012, 2 A.M.”)(LAPD, 2021b). We extracted the start date-time, end date-time, and street address for all sobriety checkpoints advertised in press releases on the LAPD website from January 1, 2012, to December 31, 2017. We geocoded street addresses to obtain latitude and longitude values using ggmap in R (Kahle and Wickham, 2013). In total, LAPD advertised 642 sobriety checkpoints during the study period, of which we were able to geocode 627 (geocoding rate = 97.7%).

The units of analysis for this longitudinal study were space-time locations within Los Angeles during 2012–2017. The space-time location for the 627 geocoded sobriety checkpoints represented a group of “event units”, and we selected the same spatial locations at different times (when LAPD did not conduct sobriety checkpoints) as a group of “comparison units”. These comparison units were matched to event units at a ratio of 2:1 and were assigned the same geographic coordinates (latitude-longitude) as the corresponding event unit. The time-period for the comparison units was offset compared to the matched event units by precisely one week before (−168 hours) and one week after (+168 hours) the advertised checkpoint time. For example, LAPD conducted a checkpoint from 8:00pm on Saturday, January 21st, 2012, to 2:00am on Sunday, January 22nd at Hollywood Boulevard and Bronson Avenue (the event unit). The two comparison units were 8:00pm to 2:00am beginning Saturday, January 14th, 2012, and 8:00pm to 2:00am beginning Saturday, January 28th, 2012, at Hollywood Boulevard and Bronson Avenue. This approach yielded an analytic dataset composed of 1,881 space-time locations representing 627 event units and 1,254 comparison units. The main independent measure was a binary variable identifying event units vs. comparison units.

2.4. Dependent Measures

We accessed data for crimes reported by the Los Angeles Police Department (LAPD) for 2012–2017 through the through City of Los Angeles open data website (LAPD, 2021a). These data are digitized versions of crime reports submitted by officers of the LAPD, and available fields include the date that the crime occurred, time of day (hour-minute), geographic coordinates (latitude-longitude), and crime code. We defined assaults as all simple and aggravated assaults, excluding domestic violence and child abuse. Included crime codes were 230 [Assault with a Deadly Weapon], 231 [Assault with a Deadly Weapon against LAPD Police Officer], 250 [Shots Fired], 251 [Shots Fired Inhabited Dwelling], 761 [Brandishing], 435 [Lynching], 436 [Lynching – attempted], 622 [Battery on Firefighter], 623 [Battery on Police Officer], 624 [Battery – misdemeanor], and 625 [Other Miscellaneous Assault]. These data have been used extensively in prior studies of crime and violence (Miller et al., 2020; Barboza et al., 2021; Campedelli et al., 2021; Sheppard and Stowell, 2022).

The dependent variable was a continuous measure capturing the incidence density of assaults at the space-time locations for the event units and comparison units. We first selected all assaults that occurred in Los Angeles during the time-period corresponding with each unit. For example, 23 assaults occurred from 8:00pm on Saturday, January 21st, 2012, to 2:00am on Sunday, January 22nd. We displayed the selected assaults as points on a projected coordinate system (California State Plane V [US feet]; NAD83) using ArcGIS 10.7.1., then calculated a kernel density raster for the full extent of the city of Los Angeles, specifying a bandwidth of 5 miles and a cell size of 620 square feet. We extracted the kernel density value for the raster cell that contained the space-time location, which we interpreted as the density per square miles for police-reported assaults at the location and time for the event unit or comparison unit (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Density of reported assaults during a single sobriety checkpoint’s operations compared to the same location temporally lagged 7 days before and 7 days after within the limits of the city of Los Angeles.

We chose to use kernel density values as continuous measures of assaults per square mile rather than other available approaches (e.g., assault counts within radial buffers around space-time locations). Kernel densities are effectively a distance-based measure that allows intervention effects to decay over space, while still incorporating outcomes that occur away from the space-time location of interest. In contrast, container-based measures dichotomize crimes as inside vs. outside the area of influence for the interventions, possibly leading to measurement error. Moreover, the irregular shape of the City of Los Angeles means that outcome counts calculated using container-based measures would be biased due to boundary effects for space-time locations that occur near the city’s outer extent. Kernel density approaches have been used in similar prior space-time analyses of injury events, including crime and violence (Kajeepeta et al., 2020; Kondo et al., 2018; Theall et al., 2022).

In addition to calculating assault density for the time-period of the checkpoint operations, we created temporally lagged versions of the dependent measures to account for the possibility that checkpoints affect assaults for up to 48-hours after checkpoint operations cease. We created eight 6-hour groups defined according to the advertised end of the corresponding checkpoint for each event and comparison unit. For example, for the checkpoint scheduled from 8:00pm to 2:00am on Saturday, January 21st, 2012, at Hollywood Boulevard and Bronson Avenue, the 0–6 hour lag for the event unit was located spatially at Hollywood Boulevard and Bronson Avenue and temporally from 2:00am to 7:59am on Sunday, January 22nd, 2012. The 6–12 hour lag was at the same location from 8:00am to 1:59pm on Sunday, January 22nd, 2012, and so on up to 48 hours after checkpoint ended operations.

2.5. Statistical Analysis

To assess associations between sobriety checkpoint presence and assault density we fit a mixed effects regression model using the lme4 package in R. The dependent measure was assault density per square mile, square root transformed to account for positive skewing. The main independent measure was the binary variable denoting checkpoint event vs. comparison units. A categorical variable identifying the matched group of 1 event unit and 2 comparison units was included as a random effect after a Hausman test supported the use of random effects compared to a fixed effects model (p = 0.15). We report coefficients from the linear mixed-effects model. Separate regression models were fitted for the main analysis capturing associations for the duration of the checkpoints and for the eight groups that capture temporally lagged associations (i.e., 0–6 hours, 7–12 hours, etc.).

The associations of interest could be confounded by spatially and temporally varying conditions that are related to sobriety checkpoint operations and affect assault incidence. Our design controlled methodologically for known and unknown spatially varying confounders (e.g., alcohol outlet density (Morrison et al., 2016), social disadvantage (Sparks, 2011)) because these features were constant within matched groups. However, confounding by time-varying features remained a possibility, so we included statistical controls for daily maximum temperature, daily total precipitation, and the presence of a holiday (Lloyd et al., 2013). Data for temperature and precipitation were obtained from the National Oceanic and Atmospheric Administration’s historical weather station data for Los Angeles International Airport (NOAA National Center for Environmental Information, 2021). Daily maximum temperature and daily total precipitation were used because weather data throughout the day were not consistently reported at all times of day, especially late at night when sobriety checkpoints occurred. The presence of a holiday was operationalized as a binary variable coded as one if a checkpoint was within two days of a major US holiday (New Year’s Eve, New Year’s Day, Memorial Day, 4th of July, Labor Day, Thanksgiving, Christmas Eve, and Christmas Day), and zero otherwise.

3. Results

There were 183,672 assaults reported by the Los Angeles Police Department between 2012 and 2017 with a mean of 30,391 reported assaults per year (SD = 2,568). The 627 sobriety checkpoints conducted during that period had an average duration of 5.65 hours (SD = 0.66). Table 1 shows the bivariate distribution for assault density and the time-varying covariates (temperature, precipitation, holiday) for the events and comparison units. Mean assault density was 0.09 per square mile for event units (SD = 0.09) and 0.10 for comparison units (SD = 0.09). Precipitation and temperature were similar for the event and comparison units. Event units were less likely to be on holidays compared to comparison units.

Table 1.

Mean (Standard Deviation) and minimum-maximum for all sobriety checkpoints and controls included in the study. Significance is assessed via two-sample t-test.

Event Units (n=627) Comparison Units (n=1,254)
Mean (SD) Min-Max Mean (SD) Min-Max p value
Duration (hrs) 5.65 (0.66) 4–8 5.65 (0.66) 4–8 ---
Assault density (crimes/sq mile) 0.094 (0.088) 0–0.60 0.100 (0.092) 0–0.60 0.15
Precipitation (in) 0.025 (0.142) 0–1.58 0.038 (0.189) 0–1.58 0.10
Max temperature (F) 71.3 (7.1) 54–99 71.6 (7.5) 54–98 0.39
Proportion on holiday (%) 5.6% 0–1 8.0% 0–1 0.05
Open in a new tab

Table 2 shows the results of the linear mixed effects model for the association between sobriety checkpoints and assault incidence. In an initial crude model, event units that had sobriety checkpoints were associated with a 1.1% decrease in the square root of assault density compared with comparison units that did not have sobriety checkpoints (Table 2; b=−0.011, 95% CI: −0.021, −0.001). After adjusting for temperature, precipitation, and holidays, checkpoints were associated with a 1.1% decrease in the square root of assault density compared to control days (b= −0.011, 95% CI: −0.020, −0.001).

Table 2.

Mixed effects linear regression outputs assessing the association between checkpoint presence and the square root of assault density (assaults per square mile) during the checkpoint operations only. A separate regression model was fit for each subsequent time period (only during is shown here). Coefficients and their 95% confidence intervals are reported. Significant values are bolded.

Crude Model Adjusted Model
Coefficient (95% CI) Coefficient (95% CI)
Checkpoint presence −0.011 (−0.021, −0.001) −0.011 (−0.020, −0.001)
Daily precipitation --- −0.026 (−0.057, 0.005)
Max temperature --- 0.0012 (0.0003, 0.0020)
Near a holiday --- 0.012 (−0.008, 0.033)
Open in a new tab

Figure 2 displays graphically the parameter estimate for the main adjusted model and the eight models specified using temporally lagged versions of the independent measure. As noted above, assault density was 1.08% lower during the sobriety checkpoint event units vs. the comparison units while the checkpoints were in operation. This association remained for 0–6 hours after the checkpoints ended (b= −0.010, 95% CI: −0.019, −0.001). Assault density was not different for event units vs. comparison units for any other lagged time period, although 18–24 hours after the end of the checkpoint, assaults marginally but insignificantly decreased again (b=−0.008, 95% CI: −0.018, 0.002).

Figure 2.

Figure 2.

Open in a new tab

Coefficient estimates and 95% confidence intervals from the adjusted, mixed effects linear regression for the nine time periods examined. Significant estimates for the association between checkpoint presence and the square root of assault density are marked with an asterisk (*).

4. Discussion

Sobriety checkpoints are an effective intervention to reduce alcohol-involved motor vehicle crashes. They are operationally similar to soft policing interventions that reduce violence (Braga et al., 2014; National Academies of Sciences, Engineering, Medicine, 2018)—such as community patrols—and could theoretically have similar preventive effects. This longitudinal spatial analysis conducted in Los Angeles, California, found checkpoints were unintentionally associated with reductions in assault incidence. Checkpoint-induced assault reductions were observed for 6 hours after the checkpoint had ended. Checkpoints were no longer associated with assault reductions 6–48 hours after the checkpoint had ended, suggesting that the observed benefits dissipate quickly.

This analysis is the first to observe associations between sobriety checkpoints and changes in neighborhood violence. Previous reviews have similarly observed small geographic reductions in firearm violence after hot spot policing interventions (Braga et al., 2014; Braga et al., 2019). A wide literature has also found that sobriety checkpoints reduce DUIs and alcohol-involved crashes and injuries (Elder et al., 2002; Erke et al., 2009; Bergen et al., 2014). The concentrated and highly visible police presence at a sobriety checkpoint raises the risk of detection and arrest for prospective offenders of all crimes, leading to fewer violent assaults. Reductions were only observed during the checkpoint operation and for 6 hours afterward, suggesting that deterrence dissipates quickly consistent with other studies observing law enforcement presence as a fleeting benefit (Draca et al., 2011). Checkpoints primarily occurred in the evenings, so six hours after checkpoint operations normally occurred during the late night and early morning hours (e.g., 1:00am – 7:00am). Assault reductions could therefore last for the remainder of the night of a checkpoint or last for six hours after the end of the checkpoint. Studies examining reductions in DUIs and alcohol-involved crashes observed reductions over a longer period of up to a week after a checkpoint occurred (Morrison et al., 2019; Morrison et al., 2021; Morrison et al., 2022).

We examined if there was any evidence of temporal displacement. We did not observe any significant increases in assault after the initial reduction during the checkpoint operations, in line with a previous study examining alcohol-involved crashes (Morrison et al., 2019). However, we note there is a slight decrease in assaults the night after a checkpoint occurred (18–24 hours after). The perceived risk of detection could remain elevated in the same location the following night, which would decrease the incidence of assaults.

Our study is subject to important limitations. Our chosen space-time panel study design can be conceptualized as a single-difference design examining change in the outcome within locations. Concerns due to contamination led us to reject a single-difference design that examined change in the outcome between locations, or a difference-in-difference design that examined change in the outcome within and between locations (Wing et al., 2018). Given that we detected very modest intervention effect sizes and that the effects dissipated almost immediately after the intervention was removed, it is unlikely that our findings are artifacts of our chosen study design. It is possible that accounting for global temporal trends could explain the detected associations; alternate designs may shed light on this finding, but we could not test this possibility due to concerns related to contamination. Additionally, results may not generalize outside the study area of Los Angeles. Bias due to endogeneity is also a possibility, in that police-reported outcomes may be affected by the presence of additional police in a location to conduct the sobriety checkpoint. However, this form of bias would lead to stronger positive associations, rather than the negative associations we detected.

This research provides preliminary evidence that sobriety checkpoints have similar deterrent effects on assaults compared to other soft policing interventions. However, effect sizes were small, and associations did not endure long after checkpoints were removed. Future studies could consider how this specific form of proactive policing compares to other forms (e.g., walking or saturation patrols) and identify ways to extend and leverage the modest and fleeting effect that we detected.

Funding

Research reported in this publication was supported by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health under Award Number R21AA025749. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declaration of Competing Interest

None.

Conflicts of Interest Statement

We have no conflicts of interest to report.

References

  1. Alert Media Inc., DUI Checkpoint Alerts. http://www.duiblock.com, 2022. (accessed September 20, 2021).
  2. Andresen MA, Ha OK. Spatially varying relationships between immigration measures and property crime types in Vancouver Census Tracts, 2016. Br J Criminol 2020;60(5):1342–67. 10.1093/bjc/azaa029. [DOI] [Google Scholar]
  3. Ariel B, Weinborn C, Sherman LW. “Soft” policing at hot spots—do police community support officers work? A randomized controlled trial. J Exp Criminol 2016;12(3):277–317. 10.1007/s11292-016-9260-4. [DOI] [Google Scholar]
  4. Baobeid A, Koç M, Al-Ghamdi SG. Walkability and its relationships with health, sustainability, and livability: Elements of physical environment and evaluation frameworks. Front Built Environ 2021;7. 10.3389/fbuil.2021.721218. [DOI] [Google Scholar]
  5. Barboza GE, Schiamberg LB, Pachl L. A spatiotemporal analysis of the impact of COVID-19 on child abuse and neglect in the city of Los Angeles, California. Child Abuse Negl 2021;116:104740. 10.1016/j.chiabu.2020.104740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Benjamin-Chung J, Arnold BF, Berger D, Luby SP, Miguel E, Colford JM Jr., Hubbard AE. Spillover effects in epidemiology: Parameters, study designs and methodological considerations. Int J Epidemiol 2018;47(1):332–47. 10.1093/ije/dyx201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bergen G, Pitan A, Qu S, Shults RA, Chattopadhyay SK, Elder RW, Sleet DA, Coleman HL, Compton RP, Nichols JL, Clymer JM, Calvert WB. Publicized sobriety checkpoint programs: A community guide systematic review. Am J Prev Med 2014;46(5):529–39. 10.1016/j.amepre.2014.01.018. [DOI] [PubMed] [Google Scholar]
  8. Blakely EJ, Snyder MG. Separate places: Crime and security in gated communities. In: Reducing crime through real estate development and management. Felson M, Peiser RB, editors. Washington, D.C.: Urban Land Institute; 1998. p. 53–70. [Google Scholar]
  9. Bowers K, Johnson S, Guerette RT, Summers L, Poynton S. Spatial displacement and diffusion of benefits among geographically focused policing initiatives. Campbell Syst Rev 2011;7(1):1–144. 10.4073/csr.2011.3. [DOI] [Google Scholar]
  10. Braga AA. Crime and policing revisited. US Dept Justice; 2015. [Google Scholar]
  11. Braga AA, Papachristos AV, Hureau DM. The effects of hot spots policing on crime: An updated systematic review and meta-analysis. Justice Q 2014;31(4):633–663. 10.1080/07418825.2012.673632. [DOI] [Google Scholar]
  12. Braga AA, Turchan B, Papachristos AV, Hureau DM. Hot spots policing of small geographic areas effects on crime. Campbell Syst Rev. 2019;15(3):e1046. 10.1002/cl2.1046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Campedelli GM, Aziani A, Favarin S. Exploring the immediate effects of COVID-19 containment policies on crime: An empirical analysis of the short-term aftermath in Los Angeles. Am J Crim Justice 2021;46(5):704–27. 10.1007/s12103-020-09578-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. CDC, Sobriety checkpoints. https://www.cdc.gov/motorvehiclesafety/calculator/factsheet/checkpoints.html, 2015. (accessed September 20, 2021).
  15. CDC, Leading causes of death - males - non-Hispanic black - United States, 2016. https://www.cdc.gov/healthequity/lcod/men/2016/nonhispanic-black/index.htm, 2019. (accessed September 20, 2021).
  16. CDC, Leading causes of nonfatal injury reports, 2000 – 2020. https://wisqars.cdc.gov/nonfatal-leading, 2020. (accessed September 20, 2021).
  17. CDC, WISQARS™ - Web-based Injury Statistics Query and Reporting System. https://www.cdc.gov/injury/wisqars/index.html, 2022. (accessed September 20, 2021).
  18. Clarke RV, Weisburd D. Diffusion of crime control benefits: Observations on the reverse of displacement; 2006.
  19. Community Preventive Services Task Force. Publicized sobriety checkpoint programs to reduce alcohol-impaired driving: Recommendation of the community preventive services task force. Am J Prev Med 2014;46(5):540–41. 10.1016/j.amepre.2014.01.019. [DOI] [PubMed] [Google Scholar]
  20. Cornish DB, Clarke RV. Understanding crime displacement: An application of rational choice theory. Criminol 1987;25(4):933–48. 10.1111/j.1745-9125.1987.tb00826.x. [DOI] [Google Scholar]
  21. Cowen C, Louderback ER, Shouraseni Sen R. The role of land use and walkability in predicting crime patterns: A spatiotemporal analysis of Miami-Dade county neighborhoods, 2007–2015. Secur J 2019;32(3):264–86. 10.1057/s41284-018-00161-7. [DOI] [Google Scholar]
  22. DeVylder JE, Oh HY, Nam B, Sharpe TL, Lehmann M, Link BG. Prevalence, demographic variation and psychological correlates of exposure to police victimisation in four US cities. Epidemiol Psychiatr Sci 2017;26(5):466–77. 10.1017/s2045796016000810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Draca M, Machin S, Witt R. Panic on the streets of London: Police, crime, and the July 2005 terror attacks. Am Econ Rev 2011;101(5):2157–81. 10.1257/aer.101.5.2157. [DOI] [Google Scholar]
  24. Dunn C, Shames M. Stop-and-frisk in the De Blasio Era. Lee D, editor. New York: Civil Liberties Union; 2019. https://www.nyclu.org/sites/default/files/field_documents/20190314_nyclu_stopfrisk_singles.pdf (accessed November 23, 2022). [Google Scholar]
  25. Elder RW, Shults RA, Sleet DA, Nichols JL, Zaza S, Thompson RS. Effectiveness of sobriety checkpoints for reducing alcohol-involved crashes. Traffic Inj Prev 2002;3(4):266–74. 10.1080/15389580214623. [DOI] [Google Scholar]
  26. Erke A, Goldenbeld C, Vaa T. The effects of drink-driving checkpoints on crashes—a meta-analysis. Accid Anal Prev 2009;41(5):914–23. 10.1016/j.aap.2009.05.005. [DOI] [PubMed] [Google Scholar]
  27. Ferris J, Devaney M, Sparkes-Carroll M, Davis G. A national examination of random breath testing and alcohol-related traffic crash rates (2000–2012). Canberra, ACT, Australia; 2015. https://fare.org.au/a-national-examination-of-random-breath-testing-and-alcohol-related-traffic-crash-rates-2000-2012/ (accessed November 23, 2022). [Google Scholar]
  28. Geller A, Fagan J, Tyler T, Link BG. Aggressive policing and the mental health of young urban men. Am J Public Health 2014;104(12):2321–27. 10.2105/ajph.2014.302046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Homel R. Drivers who drink and rational choice: Random breath testing and the process of deterrence. In: Clarke RV and Felson M, editors. Routine activity and rational choice. Advances in Criminological Theory; 1993. p. 59–84. [Google Scholar]
  30. Kajeepeta S, Theall KP, Kondo MC, et al. The association between blighted property remediation and domestic crime by alcohol availability. Health Place. 2020;64:102364. doi: 10.1016/j.healthplace.2020.102364 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kahle D, Wickham H. Ggmap: Spatial visualization with ggplot2. R J 2013;5(1):144–161. [Google Scholar]
  32. Karstedt S. Routine activity and rational choice. New Brunkswick, NJ, American Sociological Association, Sage Publications, Inc,; 1993. [Google Scholar]
  33. Kennedy DB. Facility site selection and analysis through environmental criminology. J Crim Justice 1990;18(3):239–52. 10.1016/0047-2352(90)90004-U. [DOI] [Google Scholar]
  34. Kondo MC, Morrison C, Jacoby SF, et al. Blight abatement of vacant land and crime in New Orleans. Public Health Rep. 2018;133(6):650–657. doi: 10.1177/0033354918798811 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. LAPD, Crime data from 2010 to 2019. https://data.lacity.org/Public-Safety/Crime-Data-from-2010-to-2019/63jg-8b9z, 2021a. (accessed September 20, 2021).
  36. LAPD, Los Angeles Police Department to conduct DUI checkpoints and DUI saturation patrols. https://www.lapdonline.org/home/news_view/67738, 2021b. (accessed September 20, 2021).
  37. Lloyd B, Matthews S, Livingston M, Jayasekara H, Smith K. Alcohol intoxication in the context of major public holidays, sporting and social events: A time–series analysis in Melbourne, Australia, 2000–2009. Addiction 2013;108(4):701–9. 10.1111/add.12041. [DOI] [PubMed] [Google Scholar]
  38. McLeod MN, Heller D, Manze MG, Echeverria SE. Police interactions and the mental health of black Americans: A systematic review. J Racial Ethn Health Disparities 2020;7(1):10–27. 10.1007/s40615-019-00629-1. [DOI] [PubMed] [Google Scholar]
  39. Michigan Department of State Police, et al. , Petitioners v. Rick Sitz et al., 496 U.S. 444 (1990). https://www.law.cornell.edu/supremecourt/text/496/444.
  40. Miller AR, Segal C, Spencer MK. Effects of the COVID-19 pandemic on domestic violence in Los Angeles. NBER Work Pap Ser. 2020;28068. 10.3386/w28068. [DOI] [Google Scholar]
  41. Montolio D, Planells-Struse S. How time shapes crime: The temporal impacts of football matches on crime. Reg Sci Urban Econ 2016;61:99–113. 10.1016/j.regsciurbeco.2016.10.001. [DOI] [Google Scholar]
  42. Morrison C, Ponicki WR, Gruenewald PJ, Wiebe DJ, Smith K. Spatial relationships between alcohol-related road crashes and retail alcohol availability. Drug Alcohol Depend 2016;162:241–44. 10.1016/j.drugalcdep.2016.02.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Morrison CN, Ferris J, Wiebe DJ, Peek-Asa C, Branas CC. Sobriety checkpoints and alcohol-involved motor vehicle crashes at different temporal scales. Am J Prev Med 2019;56(6):795–802. 10.1016/j.amepre.2019.01.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Morrison CN, Kwizera M, Chen Q, Puljevic C, Branas CC, Wiebe DJ, Peek-Asa C, McGavin KM, Franssen SJ, Le VK, Keating M, Ferris J. The geography of sobriety checkpoints and alcohol-impaired driving. Addiction 2022;117(5):1450–1457. 10.1111/add.15766. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Morrison CN, Kwizera M, Chen Q, Puljevic C, Branas CC, Wiebe DJ, Peek-Asa C, McGavin KM, Franssen SJ, Le VK, Keating M, Williams FM, Ferris J. Alcohol-involved motor vehicle crashes and the size and duration of random breath testing checkpoints. Alcohol Clin Exp Res 2021;45(4):784–92. 10.1111/acer.14583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Morrison CN, Ukert B, Palumbo A, Dong B, Jacoby SF, Wiebe DJ. Assaults on days of campaign rallies during the 2016 US presidential election. Epidemiology 2018;29(4):490–93. 10.1097/EDE.0000000000000821. [DOI] [PubMed] [Google Scholar]
  47. Nagin DS, Solow RM, Lum C. Deterrence, criminal opportunities, and police. Criminology 2015;53(1):74–100. 10.1111/1745-9125.12057. [DOI] [Google Scholar]
  48. National Academies of Sciences, Engineering, Medicine. Proactive policing: Effects on crime and communities. Washington, DC, The National Academies Press; 2018. 10.17226/24928. [DOI] [Google Scholar]
  49. National Highway Traffic Safety Administration. The use of safety checkpoints for DWI enforcement. U.S. Dept of Transportation; 1983. [Google Scholar]
  50. NOAA National Center for Environmental Information, Climate data online search. https://www.ncdc.noaa.gov/cdo-web/search, 2021. (accessed September 20, 2021).
  51. Nobles MR, Fox KA, Khey DN, Lizotte AJ. Community and campus crime: A geospatial examination of the Clery Act. Crime Delinq 2013;59(8):1131–56. 10.1177/0011128710372188. [DOI] [Google Scholar]
  52. Peek-Asa C The effect of random alcohol screening in reducing motor vehicle crash injuries. Am J Prev Med 1999;16(1, Suppl 1):57–67. 10.1016/S0749-3797(98)00116-0. [DOI] [PubMed] [Google Scholar]
  53. Rios VM, Prieto G, Ibarra JM. Mano Suave–Mano Dura: Legitimacy policing and Latino stop-and-frisk. Am Sociol Rev 2020;85(1):58–75. 10.1177/0003122419897348. [DOI] [Google Scholar]
  54. Schedler JC, Ensor KB. A spatiotemporal case-crossover model of asthma exacerbation in the city of Houston. Stat (Int Stat Inst) 2021;10(1):e357. 10.1002/sta4.357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sheppard KG, Stowell JI. Police fatal force and crime reporting: A test of community responses to fatal police-civilian encounters. J Interpers Violence 2022;37(21–22):Np19730–np19758. 10.1177/08862605211043579. [DOI] [PubMed] [Google Scholar]
  56. Shults RA, Elder RW, Sleet DA, Nichols JL, Alao MO, Carande-Kulis VG, Zaza S, Sosin DM, Thompson RS. Reviews of evidence regarding interventions to reduce alcohol-impaired driving. Am J Prev Med 2001;21(4, Suppl 1):66–88. 10.1016/S0749-3797(01)00381-6. [DOI] [PubMed] [Google Scholar]
  57. Sparks CS. Violent crime in San Antonio, Texas: An application of spatial epidemiological methods. Spat Spatiotemporal Epidemiol 2011;2(4):301–9. 10.1016/j.sste.2011.10.001. [DOI] [PubMed] [Google Scholar]
  58. Subbaraman N United States to fund gun-violence research after 20-year freeze. Nature 2020;577(7788):12. 10.1038/d41586-019-03882-w. [DOI] [PubMed] [Google Scholar]
  59. The White House, Presidential memorandum -- engaging in public health research on the causes and prevention of gun violence. https://obamawhitehouse.archives.gov/the-press-office/2013/01/16/presidential-memorandum-engaging-public-health-research-causes-and-preve, 2013.
  60. Theall KP, Morrison CN, Jacoby SF, et al. Neighborhood blighted property removal and 311 calls for non-emergency services: A test of a marker of social control. Geogr Anal. 2022;54(2):261–273. doi: 10.1111/gean.12286 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Thomas SA, Harris CT, Drawve G. Exploring the influence of elements of the social and physical environment on neighborhood gun crime. Am J Crim Just 2022;47(3):370–98. 10.1007/s12103-020-09599-1. [DOI] [Google Scholar]
  62. Vildosola D, Carter J, Louderback ER, Roy SS. Crime in an affluent city: Applications of risk terrain modeling for residential and vehicle burglary in Coral Gables, Florida, 2004–2016. Appl Spat Anal Policy 2019;13:441–59. [Google Scholar]
  63. Voas RB. A new look at NHTSA’s evaluation of the 1984 Charlottesville sobriety checkpoint program: Implications for current checkpoint issues. Traffic Inj Prev 2008;9(1):22–30. 10.1080/15389580701682114. [DOI] [PubMed] [Google Scholar]
  64. Weisburd D The law of crime concentration and the criminology of place. Criminology 2015;53(2):133–57. 10.1111/1745-9125.12070. [DOI] [Google Scholar]
  65. Weisburd D, Groff ER, Yang S-M. The criminology of place: Street segments and our understanding of the crime problem. Oxford University Press; 2012. [Google Scholar]
  66. Weisburd D, Wyckoff LA, Ready J, Eck JE, Hinkle JC, Gajewski F. Does crime just move around the corner? A controlled study of spatial displacement and diffusion of crime control benefits. Criminology 2006;44:549–92. 10.1111/j.1745-9125.2006.00057.x. [DOI] [Google Scholar]
  67. Wing C, Simon K, Bello-Gomez RA. Designing difference in difference studies: Best practices for public health policy research. Annu Rev Public Health 2018;39(1):453–469. 10.1146/annurev-publhealth-040617-013507. [DOI] [PubMed] [Google Scholar]

RESOURCES