Skip to main content
Log in

Measuring the Spatial and Temporal Patterns of Police Proactivity

  • Original Paper
  • Published:
Journal of Quantitative Criminology Aims and scope Submit manuscript

Abstract

Objectives

To measure where officers engage in proactive, self-initiated activities, how much time they spend being proactive, and whether their proactive activities coincide with crime patterns.

Methods

This study uses Andresen’s Spatial Point Pattern Test to compare the spatial similarity between police proactivity and crime, as well as regression modeling to explore the relationship between proactivity and crime and the time spent on proactivity and crime.

Results

In the jurisdiction examined, high levels of proactivity are noted. This proactive activity is more likely to occur in places where crime is most concentrated. Additionally, the number of proactive calls and the proactive time spent per crime-and-disorder call remain high and stable across spatial scales. For each crime call received at a street block, police initiated 0.7 proactive activities and spent approximately 28 min carrying out proactive works.

Conclusions

This study develops a way of measuring proactive activity by patrol officers using calls for service data. We find that not only do officers in this jurisdiction exhibit higher levels of proactivity to prevent crime (compared to reacting to crime), but they also do so in targeted, micro-place ways. Agencies may consider using similar techniques to gauge the levels of proactivity in their agencies if proactive activity is a goal.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or Ebook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

Notes

  1. The city’s population estimate is 853,328 based on 2014 census information, and is predominantly white, with blacks representing about a third of the population, and other racial and ethnic groups accounting for only small shares. The census shows that 17.3 % of the city’s population live below the poverty line, and the city has high rates of crime compared to the average of other cities with similar populations. Visit http://quickfacts.census.gov/qfd/states/12/1235000.html for more information.

  2. The average time spent on each police-initiated case with arrest is 112 min, while the average time for police-initiated cases in our final sample is 38 min.

  3. After excluding events from our analysis, we analyzed the excluded events. Traffic-related events accounted for over 30 % of the data excluded from the analysis, followed by events labeled as “administrative issues” (17 %). Both of these categories were overrepresented in the excluded data compared to the original data. This is likely due to traffic-related cases lacking address information (60 % of excluded cases were traffic-related). Further, nearly all administrative cases (with the exception of those marked “special assignment”) were excluded as they could not be considered either crime or proactivity. Finally, the non-crime category was also overrepresented in the excluded data for the same reason. All of the other categories of events were underrepresented in the excluded data.

  4. In 2013, the city of Jacksonville had 174 census tracts, 490 block groups, 14,904 census blocks, and 54,832 street segments.

  5. The “dissemination area” is a small area composed of one or more neighboring dissemination blocks, with a population of 400–700 persons. All of Canada is divided into dissemination areas, which might be compared with census-block groups in the United States.

  6. Since the data were originally recorded at the hundred-block level, no spatial join process and thus no shapefile at the hundred-block level are required for the temporal OLS regression analysis. On the other hand, Weisburd et al. (2004) created a hundred-block shapefile based on the street database in Seattle. We did not adopt the same strategy for the Jacksonville street database, because house numbers associated with the odd and even sides of street segments tend to be inconsistent. For example, if we reorganize the street segment based on the house number on the even side in Jacksonville, we tend not to get consistent numbers on the odd side.

  7. It is important to note that, however, the absolute difference between crime and proactive calls at places may not be large. For example, at the hundred block level, in 90 % of places, the absolute difference between proactivity and crime is within 17. Only in 0.5 % of places, the absolute difference between two measures exceed 100. A handful of places experienced over a thousand more proactive calls, whereas places with more crime calls experienced more modest contrast with the maximum difference being 477. The average difference is 5.9 when crime outnumbers proactivity and 5.1 in the opposite direction.

  8. A linear function is assumed between crime and the factual proactive work.

  9. The local relationship is graphed by performing the lowess function in STATA. Lowess carries out a locally weighted regression of the dependent variable on the independent variable. It creates a new variable that, for each y i , contains the corresponding smoothed value. The smoothed values are obtained by running a regression of y on x using only the data (x i , y i ) and a few of the data near this point. The regression is also weighted so that the central point gets the highest weight and points that are farther away receive less weight. The subset used in calculating y is indices i = max(1, i − k) through i+ = min(i + k, N), where k = [(N* bwidth −0.5)/2]. The weights for each observation are tricube, \({\text{w}}_{\text{j}} = \left\{ {1 - \left( {\frac{{\left| {x_{j} - x_{i} } \right|}}{\Delta }} \right)^{3} } \right\}^{3}\) where ∆ = 1.0001 max(x i+ − x i , x i  − x i). The smoothed value is then the weighted regression prediction of x i . For more information, see http://www.stata.com/manuals13/rlowess.pdf.

  10. Regression models were also performed using the percentage of crime and the percentage of proactive events at each location. Results are similar to the ones presented in that both the monomial and binomial terms of the percentage of crime are significant across spatial units.

  11. A concern related to polynomial models may be the potential for large error resulting from the high covariance or dependency between parameters. To minimize the math error, we centered the model by subtracting the mean from the independent variable and performed the regression analyses again (a technique used by Bradley and Srivastava 1979). Results of the centered model remained unchanged for p value and R square, while increases were seen for coefficient values. Each crime is now related to slightly higher proactivity or proactive time police allocated at places in the centered model, making it safe and reliable to conclude that JSO is highly proactive and place-based in allocating its patrol resources. We also test spatial autocorrelation, a general issue in spatial-oriented analysis that describes the degree of dependency between the spatial location and the variable measured at that location (Chainey and Ratcliffe 2005; Ratcliffe 2010). In plain terms, spatial autocorrelation suggests that places near each other are more alike (positive correlation) or less alike (negative correlation). Failing to account for spatial autocorrelation in the dependent variable will tend to underestimate the real variance in the data, increasing the likelihood of a Type I statistical error (Ward and Gleditsch 2008, as cited in Ratcliffe 2010). There are many ways to control for spatial dependency. In this article, we test it with Lagrange Multiplier test and control for it by running a spatial error model. Even though significant spatial dependency indicators are observed in some of the models, results for our variables of interest do not change much.

  12. We report the median value since the mean is distorted by crime calls—like homicide investigations—which may take up a great deal of allocated time.

References

  • Andresen MA (2009) Testing for similarity in area-based spatial patterns: a nonparametric Monte Carlo approach. Appl Geogr 29(3):333–345

    Article  Google Scholar 

  • Andresen MA, Malleson N (2011) Testing the stability of crime patterns: implications for theory and policy. J Res Crime Delinq 48(1):58–82

    Article  Google Scholar 

  • Bradley RA, Srivastava SS (1979) Correlation in polynomial regression. Am Stat 33(1):11–14

    Google Scholar 

  • Braga AA, Bond BJ (2008) Policing crime and disorder hot spots: a randomized controlled trial. Criminology 46(3):577–607

    Article  Google Scholar 

  • Braga A, Weisburd D (2012) The effects of “pulling levers” focused deterrence strategies on crime. Campbell Syst Rev 8:6

    Google Scholar 

  • Braga A, Papachristos A, Hureau D (2012) Hot spots policing effects on crime. Campbell Syst Rev 8(8):1–96

    Google Scholar 

  • Chainey S, Ratcliffe JH (2005) GIS and crime mapping. Wiley, London

    Book  Google Scholar 

  • Famega CN (2005) Variation in officer downtime: a review of the research. Polic Int J Police Strateg Manag 28(3):388–414

    Article  Google Scholar 

  • Groff ER, Ratcliffe JH, Haberman CP, Sorg ET, Joyce NM, Taylor RB (2015) Does what police do at hot spots matter? The Philadelphia policing tactics experiment. Criminology 53(1):23–53

    Article  Google Scholar 

  • Harris Corporation and the Jacksonville, FL, Sheriff’s Office (2002) Using 3D urban models and tools to assist in public safety and provide law enforcement solutions. Crime Mapp News 4(3):1–4

    Google Scholar 

  • Hunt P, Saunders J, Hollywood JS (2014) Evaluation of the shreveport predictive policing experiment. Rand Corporation, Santa Monica

    Google Scholar 

  • Kelling GL, Pate T, Dieckman D, Brown CE (1974) The Kansas city prevention patrol experiment. Police Foundation, Washington, DC

    Google Scholar 

  • Klinger DA, Bridges GS (1997) Measurement error in calls-for-service as an indicator of crime. Criminology 35(4):705–726

    Article  Google Scholar 

  • Koper CS (1995) Just enough police presence: reducing crime and disorder behavior by optimizing patrol time in crime hot spots. Justice Q 12(4):649–672

    Article  Google Scholar 

  • Koper CS (2014) Assessing the practice of hot spots policing: survey results from a national convenience sample of local police agencies. J Contemp Crim Justice 30(2):123–146

    Article  Google Scholar 

  • Koper C, Lum C, Willis J, Woods D, Hibdon J (2015) Realizing the potential of technology in policing: a multisite study of the social, organizational, and behavioral aspects of implementing police technologies. Final report to the national institute of justice. http://cebcp.org/wp-content/evidence-based-policing/ImpactTechnologyFinalReport

  • Lum C (2008) The geography of drug activity and violence: analyzing spatial relationships of non-homogenous crime event types. Subst Use Misuse 43:195–218

    Article  Google Scholar 

  • Lum C, Nagin D (2016) Reinventing American policing: a seven-point blueprint for the 21st century. Crime Justice Rev Res (forthcoming)

  • Lum C, Koper CS, Telep WT (2011) The evidence-based policing matrix. J Exp Criminol 7:3–26

    Article  Google Scholar 

  • Lum C, Telep CW, Koper CS, Grieco J (2012) Receptivity to research in policing. Justice Res Policy 14(1):61–95

    Article  Google Scholar 

  • Nagin DS, Solow RM, Lum C (2015) Deterrence, police and criminal opportunities. Criminology 53(1):74–100

    Article  Google Scholar 

  • National Research Council (2004) Fairness and effectiveness in policing: the evidence. In: Skogan W, Frydl K (eds) Committee to review research on police policy and practices. The National Academies Press, Washington, DC

    Google Scholar 

  • Ratcliffe JH (2004) Geocoding crime and a first estimate of a minimum acceptable hit rate. Int J Geogr Inf Sci 18(1):61–72

    Article  Google Scholar 

  • Ratcliffe JH (2008) Knowledge management challenges in the development of intelligence-led policing. In: Williamson T (ed) The handbook of knowledge-based policing: current conceptions and future directions. Wiley, Chichester, pp 205–220

    Chapter  Google Scholar 

  • Ratcliffe J (2010) Crime mapping: spatial and temporal challenges. In: Piquero AR, Weisburd D (eds) Handbook of quantitative criminology. Springer, New York, pp 5–24

    Chapter  Google Scholar 

  • Ratcliffe JH, McCullagh MJ (2001) Chasing ghosts? Police perception of high crime areas. Br J Criminol 41(2):330–341

    Article  Google Scholar 

  • Ratcliffe JH, Taniguchi T, Groff ER, Wood JD (2011) The Philadelphia foot patrol experiment: a randomized controlled trial of police patrol effectiveness in violent crime hotspots. Criminology 49(3):795–831

    Article  Google Scholar 

  • Sherman LW, Eck JE (2002) Policing for crime prevention. In: Sherman LW, Farrington DP, Welsh BC, MacKenzie DL (eds) Evidence-based crime prevention. Routledge, New York

    Chapter  Google Scholar 

  • Sherman LW, Rogan DP (1995) Effects of gun seizures on gun violence: “Hot spots” patrol in Kansas City. Justice Q 12(4):673–693

    Article  Google Scholar 

  • Sherman LW, Weisburd D (1995) General deterrent effects of police patrol in crime “hot spots”: a randomized, controlled trial. Justice Q 12(4):625–648

    Article  Google Scholar 

  • Sherman LW, Williams S, Ariel B, Strang LR, Wain N, Slothower M, Norton A (2014) An integrated theory of hot spots patrol strategy implementing prevention by scaling up and feeding back. J Contemp Crim Justice 30(2):95–122

    Article  Google Scholar 

  • Taylor B, Koper CS, Woods DJ (2011) A randomized controlled trial of different policing strategies at hot spots of violent crime. J Exp Criminol 7:149–181

    Article  Google Scholar 

  • Ward MD, Gleditsch KS (2008) Spatial regression models. Quantitative applications in the social sciences series. Sage, Thousand Oaks, CA

    Book  Google Scholar 

  • Weisburd D, Braga AA (eds) (2006) Police innovation: contrasting perspectives. Cambridge University Press, New York

    Google Scholar 

  • Weisburd D, Bushway S, Lum C, Yang SM (2004) Trajectories of crime at places: a longitudinal study of street segments in the city of Seattle. Criminology 42(2):283–322

    Article  Google Scholar 

  • Weisburd D, Eck JE (2004) What can police do to reduce crime, disorder, and fear? Am Acad Polit Soc Sci 593:42–65

    Article  Google Scholar 

  • Weisburd D, Green L (1995) Policing drug hot spots: the Jersey City drug market analysis experiment. Justice Q 12(4):711–736

    Article  Google Scholar 

  • Weisburd D, Groff ER, Jones G, Cave B, Amendola KL, Yang SM, Emison RF (2015) The Dallas patrol management experiment: can AVL technologies be used to harness unallocated patrol time for crime prevention? J Exp Criminol 11(3):367–391

    Article  Google Scholar 

  • Weisburd D, Groff ER, Yang SM (2012) The criminology of place: street segment and our understanding of the crime problem. Oxford University Press, New York

    Book  Google Scholar 

  • Weisburd D, Telep CW, Hinkle JC, Eck JE (2010) Is problem-oriented policing effective in reducing crime and disorder? Criminol Public Policy 9(1):139–172

    Article  Google Scholar 

  • Weisburd D, Wyckoff LA, Ready J, Eck JE, Hinkle JC, Gajewski F (2006) Does crime just move around the corner? A controlled study of spatial displacement and diffusion of crime control benefits. Criminology 44(3):549–591

    Article  Google Scholar 

  • Yang SM (2010) Assessing the spatial–temporal relationship between disorder and violence. J Quant Criminol 26(1):139–163

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Cynthia Lum.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wu, X., Lum, C. Measuring the Spatial and Temporal Patterns of Police Proactivity. J Quant Criminol 33, 915–934 (2017). https://doi.org/10.1007/s10940-016-9318-5

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10940-016-9318-5

Keywords

Navigation