Skip to main content
SearchLogin or Signup

The Financial Implications of Merging Proactive CCTV Monitoring and Directed Police Patrol: A Cost-Benefit Analysis

Journal of Experimental Criminology (2016). https://doi.org/10.1007/s11292-016-9267-x

Published onJun 23, 2016
The Financial Implications of Merging Proactive CCTV Monitoring and Directed Police Patrol: A Cost-Benefit Analysis

Abstract:

Objectives: This study presents a cost benefit analysis of an intervention pairing proactive CCTV monitoring with directed police patrol in Newark, NJ. A recent randomized controlled trial found that the strategy generated significant crime reductions in treatment areas relative to control areas. The current study focuses on the financial implications of the experimental strategy through a cost-benefit analysis.

Methods: The study begins by measuring the costs and benefits associated with the CCTV Patrol Experiment, the findings of which can inform agencies with existing CCTV infrastructure. Follow-up analyses measure the costs and benefits of the intervention for jurisdictions absent existing CCTV infrastructure, meaning agencies would need to invest in a CCTV system in addition to the specific intervention outputs. Alongside benefits provided to the criminal justice system as a whole, this study presents the cost savings afforded to each of the separate CJ system components to determine whether the financial implications differently affect the Policing, Courts, and Corrections systems.

Results: Findings suggest that the CCTV Patrol Experiment is highly cost effective for agencies with existing CCTV infrastructure. Conversely, while cost effective for the criminal system as a whole, the strategy is cost prohibitive for each of the individual CJ system components when CCTV system costs are included. Sensitivity analyses suggest that benefits would have to increase between 25% and 100% in certain instances for the intervention to generate a cost benefit.

Conclusions: The CCTV Patrol Experiment is a cost-beneficial strategy for agencies with existing CCTV infrastructure. Agencies absent CCTV may want to consider whether funds would be better allocated towards strategies besides CCTV.

Keywords:

cost-benefit analysis; CCTV; situational crime prevention; directed patrol; policing

Citation:

Piza, E., Gilchrist, A., Caplan, J., Kennedy, L. and O’Hara, B. (2016). The Financial Implications of Merging Proactive CCTV Monitoring and Directed Police Patrol: A Cost-Benefit Analysis. Journal of Experimental Criminology, 12(3): 403-429. https://doi.org/10.1007/s11292-016-9267-x

Introduction

While the crime prevention capacity of CCTV has received increased empirical attention recently, its cost effectiveness remains largely unknown. With the recent exception of La Vigne et al. (2011), research has not systematically explored the costs of CCTV and whether they are mitigated by the benefits generated from crime reductions. This is important in the current fiscal climate, marked by dwindling public budgets (Wiseman, 2011) and claims that CCTV can be used as a force multiplier that simultaneously boosts efficiency and cuts spending (COPS, 2011; Cordero, 2011). The current study addresses this gap in the literature through a cost-benefit analysis of an intervention that paired proactive CCTV monitoring with directed police patrol in Newark, NJ (hereafter referred to as the CCTV Directed Patrol Strategy, Piza et al., 2015). We begin by measuring the costs and benefits associated with the direct intervention outputs, the findings of which can inform agencies with existing CCTV infrastructure. Further analyses measure the cost effectiveness of the intervention for jurisdictions absent existing CCTV infrastructure, meaning a CCTV system would have to be funded in addition to the intervention outputs. We first calculate the overall costs and benefits, including tangible societal and criminal justice costs. We then isolate findings for the criminal justice system as well as the disaggregate system components: Policing, Courts, and Corrections. This allows us to measure the overall cost effectiveness of CCTV while determining whether the financial implications differently affect Policing, Courts, and Corrections systems.

Our findings suggest that the CCTV Directed Patrol Strategy is highly cost effective for agencies with existing CCTV infrastructure. When CCTV system costs are accounted for, very modest cost benefits are found via the Overall and CJ System calculations. However, the strategy is cost prohibitive for each of the individual CJ System components when CCTV costs are considered. This implies that the introduction of human agents, deployed in a proactive manner, in CCTV operations may be a cost effective way to generate crime reductions. Agencies that need first to invest in a CCTV system, or are unwilling to completely absorb start-up CCTV system costs, are likely to experience a cost deficit with the CCTV Directed Patrol strategy.

Review of Relevant Literature

CCTV, Crime Prevention, and Cost-Benefit Analysis

The rise of CCTV as a mainstream crime prevention tool can be traced to the Home Office's CCTV Challenge, which provided government funding for the implementation and expansion of CCTV systems throughout the United Kingdom in the mid-1990s. At the time, many scholars lamented the lack of empirical evidence available to determine whether such a large financial commitment was warranted (see, for example, Pease, 1999: p. 53). The time since has seen an increase in the empirical evidence on CCTV. Welsh and Farrington conducted two systematic reviews and meta analyses of CCTV (2002, 2009). In addition, a number of program evaluations have been published since (e.g. Caplan et al., 2011; Cameron et al., 2008; King et al., 2008; LaVigne et al., 2011; McLean et al., 2013; Ratcliffe et al., 2009; Reid and Andresen, 2014). This body of literature has produced general insights into CCTV, namely that the technology works best in small, well defined areas (particularly car parks), better deters property crime than violent crime, and is most effective when integrated within proactive crime prevention activities than when deployed as a stand-alone tactic.

The installation of CCTV cameras typically costs millions of dollars (Babwin, 2007; Goldstein and Eiserer, 2012). Maintenance expenses require similar financial commitment after installation. Ratcliffe and Groff (2011), for example, found that Philadelphia spent $200,000 a month in maintenance costs, while La Vigne et al. (2011) found that the cost associated with maintaining Chicago’s system exceeded the startup costs by the fifth year of the program. Finally, the installation of a CCTV system requires investment in personnel to conduct surveillance-related functions. Chicago reported that their personnel costs ($3,341,000) totaled more than the initial start-up ($1,431,000) and maintenance ($1,713,000) costs combined (La Vigne et al., 2011: p. 69). Taken together, the expenses associated with CCTV represent a substantial financial commitment. Norris (2003: p. 256) estimated that CCTV systems have cost over £3 billion in Great Britain over the 10-year period of 1992-2002, not including personnel. While comparable figures are not available for the United States, they are arguably similar given the widespread adoption of CCTV by American police agencies. In the U.S., 49% of local Police Departments report using public CCTV, with usage increasing to 87% for agencies serving jurisdictions with populations of 250,000 or more (Reaves, 2015).

Because money spent on CCTV could instead be devoted to other strategies, CCTV must demonstrate that it is worth funding over less expensive options. Welsh and Farrington's (2004) systematic review and meta-analysis found that improved street lighting was equally effective as CCTV in reducing crime. On the one hand, given the lower price tag of improved street lighting, this finding suggests that installing high-performing streetlights may be the wiser of the two investments. On the other hand, benefits realized through CCTV-generated crime reductions could conceivably offset the cost of the system. Because CCTV can provide numerous crime control benefits in addition to crime deterrence (see Ratcliffe, 2006), such a finding may support the use of CCTV over improved street lighting despite its higher price tag. Cost-benefit analysis can help policy makers decide between such competing strategies, or determine whether it would be fiscally reasonable to implement both. Unfortunately, situational crime prevention (SCP) evaluations have incorporated cost-benefit analysis far less frequently than other disciplines (Horowitz and Zedlewski, 2006; Sherman, 2010). Two noteworthy exceptions are particularly relevant to the current discussion. Painter and Farrington (1999, 2001) conducted a cost-benefit analysis of improved street lighting in Dudley and Stroke-on-Trent, finding that the financial savings exceed the financial costs. La Vigne et al. (2011) conducted a cost-benefit analysis of CCTV in Baltimore and Chicago, finding that in both cities crime reductions generated monetary benefits that exceeded the costs of installing, maintaining, and operating the CCTV system. While these studies separately found both improved street lighting and CCTV as cost beneficial, savings were more pronounced for street lighting, with Painter and Farrington (2001) finding a maximum benefit: cost ratio of 10:1 compared to a maximum of 4.3:1 found by La Vigne et al. (2011). However, because these studies focused on interventions in different countries, we caution against interpreting the findings as more supportive of street lighting. More so, we feel that the novelty of these studies further highlights the need for more cost-benefit analysis of common SCP interventions, as policy makers would benefit from a greater body of research evidence.

Tangible and Intangible Crime Costs: What Should be Included in a Cost-Benefit Analysis of Situational Crime Prevention?

Prior cost-benefit research has predominately focused on two types of crime costs: tangible costs of crime (e.g., government expenditures on the criminal justice system, lost wages, victims’ medical costs, etc.) and intangible societal costs of crime (e.g., pain and suffering, fear of crime, etc.) (Cohen and Bowles, 2010). Traditionally, cost-benefit analysis has stressed the inclusion of all costs associated with a problem (Rice, 1966) to ensure that the full spectrum of victimization is reflected in the analysis (Kleiman et al., 2014). However, scholars have recently raised concerns regarding the predominate methods of estimating intangible benefits, specifically willingness-to-pay (WTP) surveys and jury awards for civil damages. The WTP method takes a top-down costing approach by conducting representative surveys of U.S. residents, asking their willingness to vote for a proposal requiring each household in their community to pay between $25 and $225 to prevent 1 in 10 of a variety of crime types (Cohen et al., 2004). As noted by Dominguez and Raphael (2015: p. 616), such an approach assumes that respondents know the actual crime rate and, during the interview, cognitively process a given percentage decline and its implications. Since public perceptions of crime rates are largely inaccurate, with respondents reporting rising crime rates even during years of persistent decreases (McCarthy, 2014), WTP estimates are likely biased upwards. A second common source of intangible costs is jury awards in civil damage cases (Cohen, 1988; Roman, 2009). As noted by Tonry (2015), such lawsuits are typically not against offenders, but against wealthy corporations of commercial properties where crimes occurred. Because such cases are not representative of typical victimizations, with most victims lacking the resources to pay attorney fees to pursue civil damages and most crimes not harmful enough to command large awards, they are likely inaccurate and biased upwards (Tonry, 2015).

The inclusion of intangible estimates greatly increases the associated cost of a given crime incident. Miller al. (1996), for example, estimated the cost of crime victimization in 1993 dollars at $105 billion in tangible costs and $450 billion (i.e. over 4 times larger) when intangible costs were included. This directly influences study findings and their policy implications, as “a hypothetical percentage point decline in crime will be perceived as having an impact on crime that is exaggerated relative to true effect” (Dominguez and Raphael, 2015: p. 626). Using as an example commonly reported costs of a single homicide (between $4.8 and $12.4 million) and rape (between $136,000 and $312,000), Tonry (2015: p. 658) argues that “the benefits of almost any individual prevention program plausibly estimated to have prevented homicides or rapes will exceed costs.” Given these concerns, Dominguez and Raphael (2015: p. 626) recommend caution in pricing intangible costs, which they identify as “a difficult and poorly defined task,” while Tonry (2015) strongly advocates for eliminating intangible costs all together from cost-benefit analysis. This approach has been adopted in prior SCP evaluations, with Painter and Farrington (1999: p. 111) overtly excluding intangible costs due to their “controversial nature.”

Considering the nuances surrounding cost of crime estimates, researchers should first decide upon the specific questions they need answered by the analysis (Kleiman et al., 2014:). By framing the analysis in a manner most appropriate for the problem at hand, researchers can help maximize the utility of the study findings. An example of this issue can be drawn from the aforementioned study of La Vigne et al. (2011), to our knowledge the only CCTV evaluation to include a cost-benefit analysis, and recent interpretations of the study’s findings. La Vigne et al. (2011) conducted two versions of their cost-benefit analysis. First, the analysis accounted for both tangible criminal justice costs and intangible societal costs. Second, the cost-benefit analysis excluded societal costs, focusing only on tangible criminal justice costs. As explained by La Vigne et al. (2011), this was done in order to provide a "more relevant ratio from a local financing perspective, as any victimization cost savings that might be attributed to the camera system are not transferred to governments' budgets (p. 22)." La Vigne et al. (2011) found that in Chicago, for every $1 spent the system generated $4.30 in savings when both crime and victim costs were considered and $2.81 in savings when only crime costs were considered. Savings were less substantial in Baltimore, where CCTV yielded $1.49 in benefits for every $1 spent when both crime and victim costs were considered. When victimization costs were excluded, the cost of CCTV was roughly equal to the benefit: $1.06 for every $1 spent.

Soon after the La Vigne et al. (2011) study, the Office of Community Oriented Policing Services (COPS) (2011) released a report titled The Impact of the Economic Downturn on American Police Agencies. The report argued that police agencies could use CCTV as a force multiplier while also cutting expenditures. This report directly cited the findings of La Vigne et al. (2011), stating "one neighborhood in Chicago alone saw a cost savings of $4.30 for every dollar that was used for the system" (COPS, 2011: p. 26). However, using these figures to advocate for CCTV use by police is questionable for two specific reasons. First, the cost savings of $4.30 includes intangible societal costs, such as diminished quality of life and suffering caused by the victimization. As argued by Ratcliffe (2015: p. 166) "monetary costs to society mean little to the police as they do not recoup the costs of any crime reduction directly," a main reason La Vigne et al. (2011) provided two separate cost-benefit estimates. Second, even the reduced figure of $2.81 exaggerates the benefits police gained from CCTV. In accordance with the general cost of crime literature (Cohen, 1998; Cohen and Piquero, 2009; Cohen et al. 1994), La Vigne et al. (2011: p. 20-21) included three distinct expenditures in their criminal justice costs: the cost of arrest (Aos et al., 2001), the cost of pre-sentencing (Roman et al., 2008), and the cost of incarceration (Durose and Langan, 2004; Roman and Chalfin, 2006; Stephan, 2004). While the sum of these costs provides a valid measure of criminal justice expenditures, they are not cumulatively absorbed by a single entity. Cost of arrests directly impacts Policing, cost of pre-sentence and adjudication impacts the Courts, and cost of incarceration directly impacts Corrections. Therefore, the cumulative criminal justice cost savings may not demonstrate CCTV to be a cost effective solution for Policing, simply because Police do not directly benefit from each of these cost categories. It is with these issues in mind that we designed the current study.

Scope of the Current Study

The current study builds upon the randomized controlled trial (RCT) conducted by Piza et al. (2015) through a cost-benefit analysis. The CCTV Directed Patrol Strategy is well suited for cost-benefit analysis for two main reasons. For one, SCP interventions more readily lend themselves to cost-benefit analysis than other tactics because of the relative ease by which cost estimates can be derived, the crime specific target of such programs, and the easy identification of pre- and post- intervention time periods (Chisholm, 2000: p. 4). The use of an RCT additionally lends confidence to the results, as "the validity of a cost-benefit analysis obviously depends on the quality of the evaluation research design on which it is based" (Painter and Farrington, 2001: p. 3). This also answers recent calls for researchers to more readily translate findings of randomized experiments into financial terms to quantify the costs of particular policy choices (Kuklinski et al., 2015; Sherman, 2010).

In deciding upon a specific costing strategy (Kleiman et al., 2014) we considered prior cost-benefit studies as well as the recent commentary by Dominguez and Raphael (2015) and Tonry (2015). First, we limit our measures to tangible costs, following the approach of Painter and Farrington (1999, 2001). Excluding intangible costs avoids the pitfalls described by Dominguez and Raphael (2015) and Tonry (2015) while also producing a more conservative estimate of the achieved benefits. From this perspective, an analysis limited to tangible costs is a more rigorous test of cost effectiveness than one that also includes intangible costs. We follow the approach of La Vigne et al. (2011) by presenting both the overall findings and findings relative to the criminal justice costs. We further account for the fact that while disparate activities of separate criminal justice system components relate to one another (e.g. an offender arrested by the Police is later processed by the Courts) expenses are not spread evenly across all involved entities. To reflect this reality, we categorized the criminal justice cost-savings according to the entity that directly benefits from the averted expenditure. We considered cost of arrests as Policing Costs, cost of pre-sentence and adjudication as Court Costs, and cost of incarceration as Corrections Costs. We are particularly interested in the Policing Costs, as this is the entity most likely to invest in CCTV.

Study Setting

CCTV in Newark, NJ

Newark is the largest city in New Jersey, with a population of just over 277,000 residents, spanning over 26 square miles (U.S. Census Bureau, 2010). Newark is continuously regarded as a dangerous urban center and, most notably, has historically struggled with issues of serious street-level violence (Tuttle, 2009). The NPD recently adopted a number of technological advancements, including the installation of a CCTV system beginning in 2007. CCTV operators monitor camera feeds from the NPD’s Video Surveillance Unit (VSU). The operators report detected incidents of concern via the department's Computer-Aided Dispatch (CAD) system. All incidents awaiting dispatch are stored in CAD's Calls Pending Queue (along with incidents reported via 9-1-1) with officers dispatched to incidents with higher priority levels before lower priority incidents. This differential response policy is in accordance with accepted standards of police dispatch (LEITSC, 2008) and reflects CCTV practices reported elsewhere (e.g. Gill et al., 2005; Lomell, 2004; Ratcliffe et al., 2009).

In their evaluation of the first 73 cameras installed in Newark, Caplan et al. (2011) found that auto theft was the only crime type included in the analysis that experienced an overall reduction. However, an analysis of the individual camera locations, rather than the entire CCTV system, found auto theft levels did not change at over half (39 of 73) of the individual camera sites. A replication of the micro-level analysis conducted after the full 146-camera system was in place produced similar results, with auto theft reducing at less than half (54 of 146) of the individual camera sites (Piza et al., 2014a). These evaluations led to a series of follow-up studies meant to better understand the procedural aspects of Newark's CCTV system. These studies found that the presence of surveillance barriers prevented the integration of CCTV with proactive police activities. Piza et al. (2014b) found that while crime incidents detected and reported by CCTV operators were closed by on-scene enforcement actions more often than incidents reported via 9-1-1, proactive surveillance activity drastically decreased as the CCTV system expanded. By the time the system expanded to 146 cameras, weekly averages of only 2.11 detections and 1.22 enforcement actions were observed (Piza et al., 2014b). While this suggests that the increasing system size was largely responsible for low levels of activity, the differential response policy of police dispatch likely also played a role. During interviews with researchers (Piza et al., 2014c: p. 12), Newark’s CCTV operators “overwhelmingly reported that aspects of police dispatch —specifically large queue times— discouraged them from reporting” many of the criminal infractions they observed. Piza et al. (2015: p. 49) observed that, "If proactive enforcement is truly related to CCTV’s deterrent effects, this prior research may explain the lack of a significant, wide spread effect in Newark."

The Newark Police Department's CCTV Directed Patrol Strategy (Piza et al., 2015)

The aforementioned analyses were followed by an RCT that sought to overcome the high camera-to-operator ratio and the differential response deployment of officers via the calls-pending queue. During all experimental tours of duty, one additional CCTV operator was deployed to the control room and exclusively dedicated to monitoring the target areas. In addition, two unmarked patrol units were deployed to the target areas to respond to incidents detected by the experimental operators. Instead of the standard practice of reporting criminal activity through the department's CAD system, the experimental operator reported crime detections directly to the field units via two-way radio. The experiment incorporated a randomized block design, with 38 CCTV schemes (encompassing 64 individual cameras) assigned to either the treatment or control group.1 These 38 schemes were matched into pairs based on their level of three types of calls-for-service: violent crime, social disorder, and narcotics activity.2

The CCTV Directed Patrol strategy took place during the 11-week period from 7/20/2011 to 10/01/2011. The experiment was designed to run four days a week, Wednesday through Saturday, from 8pm to 12am to reflect the days and times at which the crimes of interest were at their height during the previous year. In recognition of prior research demonstrating that temporally focused police interventions may produce deterrent effects lasting a longer time frame (Sherman, 1990; Telep et al., 2014; Wyant et al., 2012) outcomes were measured across three distinct time periods: tours of duty (Wednesday through Saturday, 8pm to 12am, 7/20/2011-10/01/2011), days (24-hour period or each day the experiment ran), and the entire 11-week period (7/20/2011-10/01/2011). Using three time periods also represented a pseudo sensitivity analysis, ensuring that the results were not merely the by-product of a single temporal parameter. In addition to the during-experiment period, experiment effect was measured for the post-experiment period (10/05/2011-12/17/2011) to determine whether program effects extended beyond the experiment period.

Piza et al. (2015) found that the CCTV Directed Patrol strategy generated significant crime prevention benefits in the treatment areas relative to the control areas (see Table 1). In the during-experiment period, statistically significant crime reductions (noted by IRR<1) were observed in four instances. In addition, narcotics activity experienced a statistically significant reduction during the post-experiment period. Weighted Displacement Quotients (WDQs) measure the presence of any displacement (negative values) or diffusion of benefits (positive values). The Total Net Effect (TNE) combines the target area reduction and spatial displacement/diffusion effects into an estimate of the number of crimes prevented (or generated) by the intervention. For each time period, the cumulative values suggest a crime reduction.

Table 1. CCTV Patrol Experiment Effect (Piza et al., 2015)

Abbreviations: IRR=Incident Rate Ratio; WDQ=Weighted Displacement Quotient; TNE=Total Net Effect; n.s.=not significant; n.a.=not applicable

Methodology

In conducting this cost-benefit analysis, we began by creating confidence intervals for the TNE values reported by Piza et al. (2015). La Vigne et al. (2011) measured the statistical significance of cost-benefits by monetizing the 95% confidence bounds of their statistical models. Both the lower and upper bounds were multiplied by the estimated cost of crime to allow for assessment of statistical significance. La Vigne et al. (2011: p. 123-130) repeated this process twice: once for the difference-in-difference models measuring crime changes in the target areas and once for models measuring the buffer (displacement) areas.

Creating confidence bounds for the effect estimates of Piza et al. (2015) necessitated an alternate process due to the different statistical approach. Piza et al. (2015) first measured intervention effect in the target area via negative binomial regression models. Intervention effect was reported via the Incident Rate Ratio (IRR), interpreted as the rate at which the dependent variable is observed, with a value of 1 as the baseline (e.g. an IRR of 0.90 suggests a 10% decrease in the dependent variable: Braga and Bond, 2008: p. 590). For all crime types exhibiting statistically significant IRR values suggestive of a crime reduction (i.e. below 1), pre-, during-, and post-intervention crime counts were used to calculate the TNE. The TNE estimates the cumulative (combined target and displacement area) number of crimes prevented or generated by the intervention by simultaneously accounting for observed crime changes in the target, displacement, and control zones (Ratcliffe and Breen, 2008). TNE is calculated via the formula:

TNE = [Rb(Ca/Cb)-Ra] + [Db(Ca/Cb)-Da]

where R, C, and D represent the target, control, and displacement zones, respectively, b represents the time period before the intervention, and a represents the period during the intervention (Guerette, 2009: p. 41). The first part of the formula shows the crime level changes in the target area as compared to the control area, which is analogous to the IRR of the negative binomial regression models. To account for the uncertainty of the point estimates, we applied the observed confidence intervals of IRRs to create lower and upper bound estimates for the TNE. For each observed crime reduction, the upper and lower bounds were first subtracted from the IRR to measure the range of effect. Because the IRR reports percent change in incident counts, subtracting the confidence bounds shows the percentage by which the estimates should be adjusted for uncertainty. For example, violent crime during the tours period exhibited an IRR of 0.52 with a 95% CI of 0.27 – 0.98. The lower bound of the CI is 25% smaller than the IRR (0.27 - 0.52 = -0.25) while the upper bound is 46% larger (0.98 – 0.52= 0.46). Hence, the first part of the TNE formula was first decreased by 25% and then increased by 46% in order to create the lower and upper bound estimates for the total crime incidents generated or prevented by the intervention. This process was repeated for each statistically significant crime reduction. TNE confidence bounds are presented in the final column of Table 1.

Measuring Costs of Crime and Benefits of Crimes Prevented

Table 2 displays the cost of all crime types included in this study. In calculating crime costs, we adapted figures presented by La Vigne et al. (2011: p. 22, Table 3.10) and McCollister et al. (2010: p. 104, Table 3). Figures provided in 2009 dollars by La Vigne et al. (2011) and 2008 dollars by McCollister et al. (2010) were converted into 2011 dollars via the Bureau of Labor Statistics’ Consumer Price Index Inflation calculator.3 This conversion method follows the approach of McCollister et al. (2010).

We used cost estimates of La Vigne et al. (2011) to measure criminal justice costs. These estimates were compiled from a series of studies that updated the cost estimates of Cohen (1994) (i.e. Aos et al., 2001; Durose and Langan, 2004; Roman and Chalfin, 2006; Roman et al., 1998; Stephan, 2004). La Vigne et al. (2011) provided a breakdown of the CJ costs according to the precise type of expenditure: cost of arrests (Policing Costs), cost of pre-sentence and adjudication (Court Costs), and cost of incarceration (Corrections Costs). Since La Vigne et al.’s (2011) estimates of Victim Costs included intangible expenses, we used the estimates of McCollister et al. (2010), the only study to disaggregate tangible and intangible victim costs according to a systematic review of cost-benefit research (Wickramaskera et al., 2015).4

Costs of social disorder and certain violent crime categories used in the current study were not estimated in the extant literature. Since the crimes included in the social disorder category were the least serious in the current study, the lowest estimated cost was selected for these incidents. None of the referenced studies provided cost estimates for shootings, stabbings, or fights, which are unique categories of assault. Rather, costs were distinguished between aggravated assault and simple assault. Fights were assigned the costs of simple assault to reflect their less serious nature. Shootings and stabbings were assigned the costs of aggravated assault.5

The extant literature assigns cost values to individual crime types that comprise the violent crime category in the current study. For the Policing and Court categories, the costs were identical for each of the crime types. The Corrections and CJ System costs differed across crime types. Therefore, the median value of the violent crime types was used as the metric by which crime reduction savings were determined for Corrections ($6,830.25), as well as the CJ System costs ($34,627.95). This approach avoids the extreme (and likely unrealistic) estimates that would result from using either the minimum or maximum violent crime cost and is a more conservative estimate than the mean.

Table 2. Cost of Crime Incidents

1 This table was adapted from Table 3.10 of La Vigne et al. (2011: p. 22). All costs were converted into 2011 dollars. Policing costs are considered the cost of arrest. Court costs are considered the cost of pre-sentencing and adjudication. Corrections costs are considered costs of incarceration. The total CJ costs are a summation of these three values.

2 None of the referenced costs-of-crime studies provided cost estimates for social disorder. Therefore, the costs of the social disorder category were estimated using costs of the other categories. Since the crimes included in the social disorder category were the least serious in the study (i.e. drinking in public, panhandling, etc.) the lowest cost of each cost category estimated by La Vigne et al. (2011) was selected.

3 None of the referenced costs-of-crime studies provided cost estimates for shootings, stabbings, or fights, unique categories of assault. Rather, costs are distinguished between "aggravated assault" and "simple assault." Therefore, fights were assigned the costs of “simple assault” to reflect their less serious nature. Shootings and stabbings were assigned the costs of "aggravated assault."

The TNE figures discussed in the previous section were used to calculate the monetary benefits of the observed crime reductions. Since TNE values were reported for three separate time periods (tours, days, and 11-weeks), benefits were calculated for each distinct period as well. Each TNE was multiplied by the cost associated with each crime category to determine cost savings(see Table 3).To account for uncertainty in the estimates, costs were also multiplied for the lower and upper bound TNE estimates. This generated confidence intervals used to measure the statistical significance of the monetary benefits.

Table 3. Total Net Effects & Associated Cost Savings

Measuring Costs of the Intervention

Cost-benefit analysis is concerned with two basic components: inputs, the resources expended in the implementation of an intervention, and outcomes, the consequences of the intervention. Inputs are the additional human, physical or financial resources used in a project, separate from elements that would have occurred anyway in the absence of the intervention (Dhiri and Brand, 1999: p. 12). When strictly adhering to this definition, the inputs of the CCTV Directed Patrol strategy can be separated into two categories: personnel costs and vehicle costs (see Table 4A.).

During each tour-of-duty, five additional officers were dedicated to the CCTV operation: one video surveillance operator, three patrol officers, and one patrol supervisor. The experiment was designed for each tour to last 4 hours, from 8:00pm through 12:00am. However, enforcement activities led to personnel working more than four hours often. When an arrest was made towards the end of a tour, officers and CCTV operators stayed past the scheduled end time to process the arrest and export/submit video footage as evidence, respectively. The precise number of hours each officer worked was measured from his or her log sheet, submitted at the end of each tour. Over the experiment, CCTV operators worked 167 hours, patrol supervisors worked 168 hours, and patrol officers worked 500.5 hours. The NPD conducted the experiment on an overtime basis, with personnel participating outside of their everyday assignments. Therefore, overtime hourly wages were used to calculate the personnel expenditures. The three patrol officers and the one CCTV operator were paid the police officer wage of $85.26 an hour, which includes fringe benefits.6 Patrol Supervisor costs varied because officers of various ranks, each with different hourly wages, were assigned as Patrol Supervisors throughout the experiment.7 The hourly wages, including fringe benefits, for each supervisor rank are as follows: Captain, $118.31; Lieutenant, $103.86; Sergeant, $96.41. Personnel costs totaled $73,647.56.8

Vehicle costs included two separate calculations. First, we estimated the expenses associated with the general operation of the two patrol units. While a vehicle’s total usage is positively correlated with maintenance expenses (Lauria, 2007), the cost associated with a single tour-of-duty is elusive. After consulting with NPD officials, we determined that the cost of vehicle deployment should match the department's vehicle fee schedule, as determined by the Outside Employment Unit (OEU). OEU works with outside entities requesting police presence for non-NPD related events, such as directing traffic at construction sites or providing crowd control at outdoor events. As part of this function, OEU charges a fee of $15 per police vehicle for each event, which is allocated towards vehicle maintenance and repairs. We estimate the cost of vehicle operation as $15 per patrol car, totaling $30 per tour and $1,200 for the experiment.

Second, fuel costs were calculated from two separate measures: per-gallon cost of fuel and gallons of fuel consumed. Fuel costs were estimated at $3.616 per gallon, the average weekly price of fuel in the East Coast (including New Jersey) over the 11-week study period, as reported by the U.S. Energy Information Administration.9, 10 Estimating gallons of fuel consumed required a multi-step process. We first estimated the number of miles driven by consulting the field notes of researchers present within the CCTV control room during each experiment tour. Researchers monitored officer response via the CCTV cameras and noted the number of minutes officers were stationary at the scene. Researchers also measured the amount of time that officers spent processing arrests.11 The sum of the on-scene and arrest process times represented the total minutes that police were out-of-service and not driving. This time was subtracted from the total length of the tour to estimate the number of minutes officers drove. We estimated that the officers drove a total of 13,631 minutes totaling 4,537.60 miles driven.12 Fuel consumption was measured by dividing this mileage by 16 to reflect the miles per gallon of the specific vehicle used by the officers (Ford Crown Victoria) as reported by the U.S. Department of Energy.13 We estimated the consumption of 283.604 gallons of fuel during the experiment. At $3.616 per gallon, this translates to a fuel cost of $1,025.51.

Measuring Costs of a CCTV System

As previously discussed, inputs include additional resources introduced by the intervention (Dhiri and Brand, 1999). Given that the CCTV system was in place at the time of the experiment, it did not represent an input in the strict sense of the term. However, because research has demonstrated the effect of stand-alone CCTV deployment as limited, agencies considering the deployment of CCTV may not do so without also assigning personnel to proactively address incidents of concern, in a manner similar to the CCTV Directed Patrol strategy. Indeed, successful CCTV programs have reported police units working closely with CCTV operators to proactively detect and directly address street-level incidents of concern (see the chapters on the Baltimore Police Department in La Vigne et al., 2011). Furthermore, agencies that allocated funds to the installation and maintenance of cameras may be unwilling to ignore these expenses when evaluating CCTV-related interventions, even those enacted well after the installation of cameras.

Table 4B. displays the cost of the CCTV system. Donations and grant funding provided to the Newark Police Foundation, the fundraising arm of the NPD, paid for the installation of the CCTV system. The bulk of the funding came from a $3.2 million grant from the Newark Community Foundation in 2007. This grant directly funded the hardware, camera mounting, network connectivity, and software configuration for the first 109 camera sites. Therefore, we estimated the cost associated with installing a single camera in 2007 as $29,357.80 ($3.2 million/109), which translates to $31,849.38 in 2011 dollars. With 32 cameras in the experiment target area, we estimated the installation cost as $1,019,180.16

In addition to installation costs, CCTV systems have recurring maintenance costs, such as software upgrades and hardware repairs. The NPD did not purchase a maintenance contract for their system and did not keep readily accessible records of their maintenance expenses. Therefore, we consulted La Vigne et al. (2011) in estimating maintenance costs. The systems studied by La Vigne et al. (2011), and their associated recurring costs, were very divergent. Chicago's system included approximately 8,000 cameras, about 6,000 of which were owned and operated by outside entities (i.e. private businesses and citizens). Chicago’s system streams live footage from all cameras to computers connected to its network, enabling all officers to view CCTV footage from their desktop monitors, a configuration unique to Chicago (La Vigne et al. 2011: p. 54-55). Conversely, Baltimore operates approximately 500 cameras, all owned and operated by the Baltimore PD, with video feeds visible from a single control room (La Vigne et al., 2011: p. 23). Since this configuration is more representative of typical police-led CCTV systems, including NPD’s, we used Baltimore's maintenance costs as guidance. Baltimore's maintenance costs ($775,000) were 14.4% of their initial start-up costs ($5,479,000) (La Vigne et al., 2011: p. 47). We estimated NPD’s maintenance costs at 14.4% of the start-up costs, totaling $146,761.94.

In addition to the staff dedicated to the CCTV Directed Patrol strategy, additional officers are needed to fulfill day-to-day functions. CCTV operations involve a host of activities, including reviewing past footage, monitoring cameras in response to citizen calls-for-service (as opposed to the proactive monitoring activities conducted in the experiment), burning disks at the request of outside parties, completing administrative reports, fielding telephone calls, and operating complimentary technologies, such as Gun Shot Detection Systems (Gill et al., 2005; Keval and Sasse, 2010; Leman-Langlois, 2002; Piza et al., 2014b). It is important to distinguish between such day-to-day tasks associated with the general CCTV system and the proactive monitoring practices of the CCTV Directed Patrol strategy, divergent activities dependent on the efforts of separate personnel. Such day-to-day personnel expenses were estimated to reflect VSU’s staffing configuration at the time of the e, which consisted of two operators and one sergeant. The operators cover two different 8-hour shifts, one from 11am through 7pm and another from 7pm through 3am. The yearly salary of the officers is $81,783.00, plus a fringe of $29,032.97. The sergeant works a shift from 9am through 5pm at a yearly salary of $92,488.50 plus a fringe of $32,833.42. To reflect the study period, we calculated the amount of salary paid to the personnel over an 11-week period. We first divided the salaries by 52 to determine the weekly salary amount. This amount was then multiplied by 11 to determine the experiment period salary. The VSU personnel salary totaled $73,394.08.14

Table 4. Costs of Intervention Inputs & CCTV System

1 The hourly wage for all personnel includes a fringe benefit of 35.5%.

2 The overtime hourly wages for each Patrol Supervisor rank are as follows: Captain, $118.31; Lieutenant, $103.86; Sergeant, $96.41.

3 This cost reflects the Newark Police Department's Outside Employment Unit's daily vehicle fee of $15.

4 Vehicle Driving Minutes estimated from observational data collected within the CCTV control room.

5 Assumes that officers drove at a speed of 20 miles per hour throughout the tours.

6 U.S. Department of Energy (www.fueleconomy.gov).

7 U.S. Energy Information Administration (www.eia.gov).

8 All CCTV-related expenses were converted to 2011 dollars.

9 The per-camera installation costs were based upon a $3.2 million grant from the Newark Community Foundation in 2007, which directly funded the hardware, camera mounting, network connectivity, and software configuration for 109 individual camera sites.

10 Estimated at 14.4% of initial start-up costs, as observed in the comparable system of Baltimore, MD (La Vigne et al., 2011).

11 The personnel costs relate only to the portion of the salaries that would be paid over an 11-week period, to reflect the experiment period. The yearly salaries and 35.5% fringe rate for the Unit Commander ($125,321.92) and each CCTV Operator ($81,783) were divided by 52 to determine the weekly salary amount. This weekly amount was then multiplied by 11 to determine the experiment period salary.

Calculating Program Benefits: Costs Benefit Ratios (CBR)

A Cost Benefit Ratio (CBR) is the monetary benefit of the intervention divided by the cost of the intervention (Bowers et al., 2004: p. 297). We calculated the CBR via the formula:

CBR = B/C

where B is the monetary benefit of the crime reduction and C is the cost of the intervention inputs. Interpretation of the CBR is straightforward, with values above one indicating that the intervention was cost beneficial and values below one suggesting that it was cost prohibitive (Bowers et al., 2004). Five CBRs were calculated, one for each component of interest: Overall Savings, CJ System, Policing, Courts, and Corrections. To inform agencies absent CCTV systems, an alternate CBR was calculated where the cost of installing, maintaining, and operating a CCTV system was included in the formula.

To account for the delay in program benefits, we applied a discounting rate (Dn) to the numerator of the CBR formula. Discounting has commonly been used in cost-benefit analysis, including evaluations of SCP efforts (Painter and Farrington, 1999, 2001). Discounting was unnecessary in the existing system analysis, because the personnel and vehicle costs were all allocated during the same year as the achieved benefits (2011). We discounted benefits in the non-existing system calculations since the installation of Newark's CCTV system occurred in 2007, 4 years before the RCT. We used the discounting formula commonly used in criminal justice (Dhiri and Brand, 1999) as well as other public policy research (Howard, 2013; Kolb and Scheraga, 1990; Weimer, 2008):

Dn=1/(1+r)n

where r is the annual discount rate and n represent the number of years between the program start date (in this case, the installation of CCTV) and the realization of cost benefits. We used a time period of 4 years to reflect the time between the CCTV system installation and deployment of the CCTV Directed Patrol strategy. We used a yearly discount rate of 3%, which Cohen and Bowles (2010: p. 146) identify as the standard rate used in U.S.-based economics research.

We concluded with a sensitivity analysis meant to "indicate the range of values within which assumptions can be safely ignored or the specific conditions that must be found or produced if a policy or program is to yield the desired result" (Barnett and Escobar, 1987: p. 391). Sensitivity analysis is an important step in any cost-benefit analysis since it ensures the rigor of the findings against a range of alternate assumptions (Chisholm, 2000; Dhiri and Brand, 1999; Welsh et al., 2013). In the current study, we presented alternate assumptions for the officer wages and CCTV System, which jointly accounted for the vast majority of the intervention expenses. Data from the Bureau of Labor Statistics (BLS) (2015) were used to operationalize alternate officer wages. According to BLS (2015), New Jersey has the highest mean wage of Police and Sherriff’s officers in the U.S. with Newark officer wages ranking in the 90th percentile. We recalculated program benefits using officer wages from the 75th and 50th percentiles to provide estimates for police departments in states more representative of national salaries.15 We were unable to find national estimates for CCTV camera costs, so we recalculated benefits with CCTV system expenses both decreased and increased by 25%. These adjustments were used for both the existing system and the non-existing system CBRs.

Findings

Table 5 displays the CBR values. In each of the existing system calculations, the intervention was cost beneficial. Statistically significant CBRs ranged from 19.36 to 31.62 in the Overall calculations, from 18.81 to 26.13 for the CJ System, from 6.99 to 14.13 for Policing, from 5.30 to 9.51 for Courts, and from 4.18 to 5.18 for Corrections. This provides strong support for the cost benefit of the CCTV Directed Patrol strategy, as the direct costs of the intervention inputs were completely offset by the benefits generated by the crime reduction. The benefits were not restricted to any particular component, with all CBRs suggesting significant cost savings. CBRs associated with a non-existing system, meaning the costs of implementing CCTV are considered along with the direct intervention inputs, provide much less support for the CCTV Directed Patrol strategy. Modest evidence in support of the intervention was found in the Overall and CJ System models. For the Overall calculations, CBRs suggest significant cost benefits for the both the tours (CBR=2.04) and days (CBR=1.63) periods. Only the tours period achieved statistical significance (CBR=1.68) for the CJ System. For Policing, the 11-week period exhibited a significant cost deficit (CBR=0.45), while findings of the tours and days period did not achieve statistical significance. For both Courts and Corrections, CBRs were well below 1 and statistically significant across all time periods. CBRs ranged from 0.34 to 0.61 for Courts, and from 0.15 to 0.33 for Corrections. This suggests that for any single CJ system component the intervention would be cost prohibitive when CCTV system costs are considered alongside the intervention inputs. Many of CBRs suggest monetary losses of over 50% for every $1 spent.

Table 5. Cost-Benefit Ratios (CBR)

Table 6 displays the results of the sensitivity analysis. All significant existing system CBRs in the primary analysis maintained statistical significance in the sensitivity analysis, showing that the cost benefits of the intervention are maintained across both variations of the officer wage calculations. The non-existing system wage variations also produced results similar to the main findings. When officer wages are adjusted, Overall and CJ System calculations exhibited significant CBRs suggestive of a cost benefit for the tours period. For Policing, Courts, and Corrections all CBRs were statistically significant and suggestive of large cost deficits. While the significance of specific CBRs differed, results were generally the same between the main findings and sensitivity analysis: outside of modest support in the Overall and CJ System calculations, the experimental strategy was largely cost prohibitive in the non-existing system calculations. When system costs were reduced by 25%, both the tours and days period were suggestive of a cost benefit for the Overall and CJ System calculations. Findings for the individual components largely echoed the main results. When system costs were increased by 25%, CBRs were statistically significant and suggestive of a cost deficit across each time period for Policing, Courts, and Corrections, which concurs with the main study findings. However, both the Overall and CJ System findings were influenced by the increased CCTV system cost estimate. For the CJ System, the lone significant CBR (0.80) suggested a cost deficit over the 11-week period. Conflicting evidence was found in the Overall calculations, with the tours period experiencing a significant cost benefit (CBR=1.35) while the 11-week period showed evidence of a significant cost deficit (CBR=0.83). When considered together the sensitivity analyses suggest that findings are largely not influenced by different officer wages or the less expensive CCTV system estimate. However, when CCTV System costs are increased by 25%, the findings are predominately unsupportive of the experimental strategy.16

Table 6. Sensitivity Analysis Results

Discussion

Findings suggest that agencies with pre-existing CCTV infrastructure stand to receive a significant cost benefit from employing the CCTV Directed Patrol strategy. This has implications for Policing, as research has consistently shown the deterrent effect of stand-alone CCTV deployment as limited. Additional investments may be necessary to make pre-existing CCTV systems effective crime prevention tools. Additional investments have begun to take the form of complementary technologies, such as gunshot-flash recognition and video analytics (La Vigne et al., 2011: p. 23). Such technologies require extensive funding, and may not assist proactive CCTV activities as expected (for example, see Piza et al., 2014b: p. 1038). Introducing additional human agents, deployed in a proactive manner, into CCTV operations may be a more cost effective tactic than complimentary technologies. In terms of Policing, the entity most likely to invest in CCTV, the findings suggest that agencies with existing CCTV infrastructure can yield cost benefits between $6.99and $14.13 for every $1 invested in the CCTV Directed Patrol strategy.

However, the existing system analysis inherently assumes that the agency considers the cost of pre-existing CCTV as somewhat negligible. For certain agencies this may be the case. In Newark, the NPD paid for the entire CCTV system with funds raised through grants and philanthropy. The allocation of resources towards proactive monitoring and patrol tactics represents a new disbursement from the agency's budget, rather than an additional disbursement. However, other agencies that allocated internal funds to the installation and maintenance of cameras may not be as willing to ignore their start-up expenses. In this sense, the CCTV Directed Patrol strategy is cost prohibitive for agencies needing to also invest in a CCTV system. Again using Policing as an example, each $1 investment was associated with a loss of $0.55 (CBR=0.45). While increasing benefits in future implementations of the CCTV Directed Patrol strategy may be possible, expecting the crime prevention capacity to increase so drastically may be unrealistic. It should be noted that minimizing costs might have a similar effect as producing additional benefits. In the extreme, live monitoring functions could be abandoned all together outside of the experimental strategy, meaning no salaries would have to be paid for VSU personnel (see King et al., 2008 for a description of such a passive CCTV operation in San Francisco). Police can also staff VSU entirely with volunteer police officers, given the increased popularity of using volunteer officers in special operations units (Hilal and Olsen, 2010). Police may also increase benefits by conducting the CCTV Directed Patrol strategy for a longer period, or strategically targeting crime types not included in the original experiment (e.g. property crime). However, expanding the scope of an intervention does not guarantee increased success, and can actually reduce program effect by creating a treatment group too large to adequately address (Sherman et al., 1989) or by requiring the participation of additional personnel who may not be as capable as the originators (Zedlewski, 2009: p. 358). Nonetheless, the sensitivity analysis demonstrated that reducing CCTV system costs, the most expensive aspect of the experimental strategy, by 25% failed to generate significant cost benefits for any individual CJ System component. Conversely, increasing CCTV System costs by 25% generated overwhelming support against the use of the CCTV Directed Patrol strategy. Therefore, we conclude that agencies with existing CCTV infrastructure stand to gain much more from adopting the CCTV Directed Patrol strategy than other agencies. Agencies absent CCTV may want to consider whether funds would be better allocated towards alternate strategies.

This study also suggests the possibility for creating a cost-sharing model across the criminal justice system. Crime reductions generated by the CCTV Directed Patrol strategy, a police-led operation, provided cost benefits for the CJ System. This suggests that governments can consider re-allocating funds to police agencies to offset expenses of crime control strategies when said strategies have the potential to benefit the entire CJ System. Indeed, prior research has demonstrated that shifting resources from imprisonment to police-led crime prevention strategies would greatly enhance crime prevention efforts (Durlauf and Nagin, 2011). We recognize that the reduction of one agency's budget to directly increase another’s may be a controversial proposition in the current fiscal climate. Thus, we repeat our call for caution from the prior paragraph if a cost sharing model is not feasible.

Conclusion

The current study generated what we consider to be important policy implications for the crime prevention field. Nonetheless this study, like most others, suffers from specific limitations that should be mentioned. First, we were not able to directly measure certain aspects of the intervention inputs and had to generate estimates. This was especially the case with the vehicle fuel costs. This was the result of the post-hoc nature of the cost-benefit analysis. Had the cost-benefit analysis been planned as a component of the RCT we could have generated more direct measures. Researchers could have recorded the odometer mileage prior to and following each tour of duty or the NPD could have maintained records of when each vehicle was fueled (and how much fuel was added in each instance). Future research should design cost-benefit analyses in a more ad-hoc manner to allow for direct measurement of all relevant expenditures. In addition, we should note that the use of officer wages in intervention cost calculations has been criticized previously. Wayson and Funke (1989) demonstrated that the cost of an officer hour is more than the direct wages when fringe benefits, vacation, sick leave, and supervision are accounted for. However, we feel that our wage calculations are appropriate in this instance. For one, they include fringe benefits, a key concern of Wayson and Funke (1989). In addition, since officers worked on an overtime basis with a direct supervisor, costs associated with vacation, sick leave, and supervision are not as relevant here as in prior research. Furthermore, a recent study used hourly officer wages as a primary measure of police costs (Burgdorf and Kilmer, 2015). Nonetheless, we acknowledge criticisms of the officer wage approach.

Furthermore, the concise time frame of the intervention did not allow for a longitudinal analysis for the purpose of determining whether cost effect changed over time. La Vigne et al. (2011) identified that cost benefits in Chicago and Baltimore did not manifest until well after the CCTV systems were installed. In that sense, we may have generated different findings were we able to use a similar analytical method. Furthermore, the CCTV Directed Patrol strategy took place towards the end of the summer months, when crime rates in Newark were at their highest during the pre-experiment period (Piza et al., 2015). The results may have exaggerated the cost benefits generated by the intervention. The results of the cost-benefit analysis employed here could have differed had the CCTV Directed patrol strategy occurred during a different part of the calendar year. The estimates presented likely reflect the maximum cost benefit criminal justice agencies could expect when incorporating directed patrol into their CCTV operations.

Lastly, the crime costs used in this study came from a subset of the extant cost of crime literature. A recent systematic review (Wickramasekera et al., 2015) found 21 studies that estimated crime costs, with considerable variance in the estimates. Incorporating different studies for our cost estimates may have altered our findings. In addition, any limitations present in the prior studies directly influence our findings. A key concern is the use, or lack thereof, of sanction probabilities in cost of crime estimates. The Courts (pre-sentencing) and Corrections (incarceration) costs accounted for the conditional probability of sanction (see La Vigne et al., 2011: p. 20). For Courts, costs were calculated on a per-minute basis for trials and mean trial lengths for each offense. For Corrections costs, the probability of sentencing, percentage of time served, and facility type (jail vs. prison) were accounted for. Such conditional probabilities were not included in the Policing (arrest) costs (Aos et al., 2001). We acknowledge that the lack of uniform cost methods may have inflated the costs of crime and, therefore, the financial benefits of the experimental strategy. Benefits upwards of $30 per $1 of costs have typically not been found in prior cost-benefit analyses (though, Painter and Farrington (2001: p. 7) claim a benefit of £74 for every £1 spent if the lighting system were paid off over a 20-year period). However, we should note that the intervention inputs (overtime personnel funding and vehicle fueling/maintenance) in this program were much less costly than intervention inputs associated with many of the re-entry and family-focused projects previously subjected to cost-benefit analysis. For example, Multidimensional Treatment Foster Care (MTFC) costs about $7,400 per family in Washington State (Welsh et al., 2015: p. 269). The cost of the CCTV Directed Patrol strategy ($75,861.73) represents the approximate cost of including only 10 families in the MTFC. The high CBRs observed in the current study may be related to the less expensive intervention inputs as well as the costing methods employed. Indeed, CBR values were much lower when accounting for the cost of the CCTV system, which brings intervention input expenses more in line with prior research.

While acknowledging these limitations, we believe that this study positively contributes to the crime prevention literature. This study adds a cost-benefit analysis to the CCTV literature while introducing the approach of disaggregating cost savings according to the CJ system component that would directly benefit. Conducting cost-benefit analysis in such a manner has direct implications for policy makers considering the adoption of new strategies, and can be used in the evaluation of crime prevention interventions outside of CCTV.

References

Aos, S., Phipps, P., Barnoski, R., and Lieb, R. (2001). The comparative costs and benefits of Programs to Reduce Crime (Version 4.0). Olympia, Washington: Washington State Institute for Public Policy.

Babwin, D. (2007). Chicago video surveillance gets smarter. USA Today. Originally published September 27, 2007. Retrieved 5/25/15 at: http://www.usatoday.com/tech/products/2007-09-27-4171345706_x.htm

Barnett, W. and Escobar, C. (1987). The economics of early educational intervention: A review. Review of Educational Research, 57(4): 387-414.

Bowers, K., Johnson, S., and Hirschield, A. (2004). Closing off opportunities for crime: An evaluation of alley-gating. European Journal on Criminal Policy and Research, 10: 285-308.

Braga, A. and Bond, B. (2008). Policing crime and disorder hot spots: A randomized controlled trial. Criminology, 46(3): 577-607.

Brown, P. (1988). Policy analysis, welfare economics, and the greenhouse effect. Journal of Policy Analysis and Management, 7(3): 4571-475.

Bureau of Labor Statistics [BLS]. (2015). Occupational employment statistics. Occupational Employment and Wages, May 2014. 33-3051 Police and Sheriff’s Patrol Officers. United States Department of Labor: Washington, DC. Retrieved 2/15/16 at http://www.bls.gov/oes/current/oes333051.htm.

Burgdorf, J. and Kilmer, B. (2015). Police costs of the drug market intervention: Insights from two cities. Policing: A Journal of Policy and Practice, 9(2): 151-163.

Cameron, A., Kolodinski, E., May, H., and Williams, N. (2008). Measuring the effects of video surveillance on crime in Los Angeles. Report prepared for the California Research Bureau. USC School of Policy, Planning, and Development.

Caplan, J., Kennedy, L. and Petrossian, G. (2011). Police-monitored cameras in Newark, NJ: A quasi-experimental test of crime deterrence. Journal of Experimental Criminology, 7(3): 255-274.

Chisholm, J. (2000). Benefit-cost analysis and crime prevention. Trends & Issues in Crime and Criminal Justice. Australian Institute of Criminology: Canberra, Australia.

Cohen, M. (1994). The costs and consequences of violent behavior in the United States. In Reiss, A. and Roth, J. (eds.), Consequences and Control of Understanding and Preventing Violence: Volume 4: 67-166.National Research Council, National Academy Press: Washington, DC.

Cohen, M. (1988). Pain, suffering, and jury awards: A study of the cost of crimes to victims. Law and Society Review, 22(3): 537-555.

Cohen, M. and Bowles, R. (2010). Estimating costs of crime. In Piquero, A. and Weisburd, D. (eds.). Handbook of Quantitative Criminology: 143-162. Springer: New York.

Cohen, M. and Piquero, A. (2009). New evidence on the monetary value of saving a high risk youth. Journal of Quantitative Criminology, 25(1): 25-49.

Cohen, M., Rust, T., Steen, S., and Tidd, S. (2004). Willingness-to-pay for crime control programs? Criminology, 42(1): 89-109.

Community Oriented Policing Services, Office of. (2011). The impact of the economic downturn on American police agencies. U.S. Department of Justice: Washington, DC.

Cordero, J. (2011). Reducing the cost of quality policing: Making community safety cost effective and sustainable. NJLM Educational Foundation, Friends of Local Government Services. The Cordero Group: Trenton, NJ.

Dhiri, S. and Brand, S. (1999). Analysis of costs and benefits: Guidance for evaluators. Home Office. Research, Development and Statistics Directorate. London.

Dominguez, P. and Raphael, S. (2015). The role of cost-of-crime literature in bridging the gap between social science research and policy making. Criminology and Public Policy, 14(4): 589-632.

Durose, P. and Langan, P. (2004). Felony sentences in state courts, 2002. Bureau of Justice Statistics Bulletin NCJ 1206916. Washington DC: U.S. Department of Justice.

Durlauf, S. and Nagin, S. (2011). Imprisonment and crime. Can both be reduced? Criminology and Public Policy, 10(1): 13-54.

Gill, M., Spriggs, A., Allen, J., Hemming, M., Jessiman, P. and Kara, D. (2005). Control room operation: Findings from control room observations. London: Home Office.

Goldstein, S. and Eiserer, T. (2012). Crime hot spots may get high-tech help. The Dallas Morning News. Originally published April 3, 2012.

Guerette, R. (2009). Analyzing crime displacement and diffusion. Problem-Oriented guides for police. Problem-solving tools series. No. 10. U.S. Department of Justice, Office of Community Oriented Policing Services. Center for Problem-Oriented Policing.

Hilal, S. and Olsen, D. (2010). Police reserve officers: Essential in today's economy and an opportunity to increase diversity in the law enforcement profession. Police Chief, 77 (10): 92-94.

Horowitz, J. and Zedlewski, E. (2006). Applying cost-benefit analysis to policing innovations. Justice Research and Policy, 8(1): 52-65.

Howard, G. (2013). Discounting for personal and social payments. Patience for others, impatience for ourselves. Journal of Environmental Economics and Management, 66 (3): 583-597.

Keval, H. and Sasse, M. (2010). “Not the usual suspects”: A study of factors reducing the effectiveness of CCTV. Security Journal, 23(2): 134-154.

King, J., Mulligan, D., and Raphael, S. (2008). CITRIS Report: The San Francisco community safety camera program. An evaluation of the effectiveness of San Francisco’s community safety cameras. Research in the Interest of Society. Center for Information Technology Research in the Interest of Society, University of California, Berkeley.

Kleiman, M., Caulkins, J., and Gehred, P. (2014). Measuring the costs of crime. Final reported submitted to the National Institute of Justice in partial fulfillment of award number 2011-IJ-CX-K059.

Kolb, J. and Scheraga, J. (1990). Discounting the benefits and costs of environmental regulations. Journal of Policy Analysis and Management, 9(3): 381-390.

Kuklinski, M., Fagan, A., Hawkins, J., Briney, J., Catalano, R. (2015). Benefit-cost analysis of a randomized evaluation of communities that care: Monetizing intervention effects on the initiation of delinquency and substance use through grade 12. Journal of Experimental Criminology, 11(2): 165-192.

La Vigne, N., Lowry, S., Markman, J., and Dwyer, A. (2011). Evaluating the use of public surveillance cameras for crime control and prevention. US Department of Justice, Office of Community Oriented Policing Services. Urban Institute, Justice Policy Center: Washington, DC.

Lauria, D. (2007). Cost-benefit analysis of Tacoma's assigned vehicle program. Police Quarterly, 10(2): 192-217.

Law Enforcement Information Technology Standards Council [LEITSC]. (2008). Standard functional specifications for law enforcement computer aided dispatch (CAD) systems. U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Assistance, and the National Institute of Justice: Washington, DC.

Leman-Langlois, S. (2002). The myopic panopticon: The social consequences of policing through the lens. Policing and Society, 13(1): 43-58.

Lomell, H. (2004). Targeting the unwanted: Video surveillance and categorical exclusion in Oslo, Norway. Surveillance& Society, 2: p. 346-360.

McCarthy, J. (2014). Most Americans still see crime up over last year. Gallup. Retrieved 2/22/16 from http://www.gallup.com/poll/179546/americans-crime-last-year.aspx?g_source=Most%20Americans%20still%20see%20crime%20up%20over%20last%20year&g_medium=search&g_campaign=tiles

McCollister, K., French, M. and Feng, H. (2010). The cost of crime to society: New crime-specific estimates for policy and program evaluation. Drug and Alcohol Dependence, 108(1-2): 98-109.

McLean, S., Worden, R., and Kim, M. (2013). Here's looking at you: An evaluation of public CCTV cameras and their effects on crime and disorder. Criminal Justice Review, 38 (3): 303-334.

Miller, T., Cohen, M., and Wiserma, B. (1996). Victim costs and consequences: A new look. US Department of Justice, National Institute of Justice; Washington, DC.

Norris, C. (2003). From personal to digital: CCTV, the panopticon, and the technological mediation of suspicion and social control. In Lyon, D. (ed.) Surveillance as social sorting: Privacy, risk and digital discrimination. Routledge: London and New York.

Painter, K. and Farrington, D. (2001). The financial benefits of improved street lighting, based on crime reduction. Lighting Research and Technology, 33(1): 3-12.

Painter, K. and Farrington, D. (1999). Street lighting and crime: Diffusion of benefits in the Stroke-on-Trent project. In Painter, K. and Tilley, N. (eds.) Surveillance of Public Space: CCTV, Street Lighting and Crime Prevention: 77-122. Crime Prevention Studies, Volume 10. Criminal Justice Press: Monsey, NY.

Pease, K. (1999). A review of street lighting evaluations: Crime reduction effects. In Tilley, N. & Painter, K. (eds.), Surveillance of public space: CCTV, street lighting and crime prevention. Crime Prevention Studies, (Vol. 10). Monsey, NY: Criminal Justice Press.

Phillips, C. (1999). A review of CCTV evaluations: Crime reduction effects and attitudes towards its use. In Tilley, N. and Painter, K. (eds.) Surveillance of Public Space: CCTV, Street Lighting and Crime Prevention. Crime Prevention Studies Vol. 10. Criminal Justice Press: Monsey, NY.

Piza, E., Caplan, J. and Kennedy, L. (2014a). Analyzing the influence of micro-level factors on CCTV camera effect. Journal of Quantitative Criminology, 30(2):237-264.

Piza, E., Caplan, J., and Kennedy, L. (2014b). Is the punishment more certain? An Analysis of CCTV Detections and Enforcement. Justice Quarterly, 31(6): 1015-1043.

Piza, E., Caplan, J., and Kennedy, L. (2014c). CCTV as a tool for early police intervention: Preliminary lessons from nine case studies. Security Journal, DOI: 10.1057/sj.2014.17.

Piza, E., Caplan, J., Kennedy, L., and Gilchrist, A. (2015). The effects of merging proactive CCTV monitoring with directed police patrol: A randomized controlled trial. Journal of Experimental Criminology, 11(1): 43-69.

Ratcliffe, J. (2006). Video surveillance of public places. Problem-Oriented Guides for Police. Response Guide Series. Guide No. 4.U.S. Department of Justice Office of Community Oriented Policing Services. Center for Problem-Oriented Policing: Washington, DC.

Ratcliffe, J. (2015). Towards an index for harm-focused policing. Policing: A Journal of Policy and Practice, 9(2): 164-182.

Ratcliffe, J. and Breen, C. (2008). Spatial evaluation of police tactics in context (SEPTIC) spreadsheet, version 3. Downloaded from www.jratcliffe.net.

Ratcliffe, J. and Groff, E. (2011). Preliminary findings from the Philadelphia CCTV study. Presentation at the American Society of Criminology Annual Meeting. Washington DC.

Ratcliffe, J., Taniguchi, T., and Taylor, R. (2009). The crime reduction effects of public CCTV cameras: A multi-method spatial approach. Justice Quarterly, 26(4): 746-770.

Reaves, B. (2015). Local police departments, 2013: Equipment and Technology. U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics: Washington, DC.

Reid, A. and Andersen, M. (2014). An evaluation of CCTV in a car park using police and insurance data. Security Journal, 27: 57-79.

Rice, D. (1966). Estimating the cost of illness. Volume 1. US Department of Health, Education, and Welfare: Washington, D.C.

Roman, J. (2009). What is the price of crime? New estimates of the cost of criminal victimization. Doctoral Dissertation, University of Maryland.

Roman, J., Woodard, J., Harrell, A., and Riggs, S. (1998). Relative costs and benefits of the superior court drug intervention program. Washington DC: The Urban Institute. www.urban.org/publications/407753.html.

Roman, J. and Chalfin, A. (2006). Does it pay to invest in reentry programs for jail inmates? Washington DC: Urban Institute. www.urban.org/ReentryRoundtable/roman_chalfin.pdf.

Roman, J., Sundquist, A., and Knight, C. (2008). Cost-benefit analysis of reclaiming futures. Washington DC: The Urban Institute.

Sherman, L. (1990). Police crackdowns: Initial and residual deterrence. In Tonry, M. and Morris, N. (eds.), Crime and Justice: A Review of Research, Vol. 12: 1-48. University of Chicago Press: Chicago.

Sherman, L. (2010). An introduction to experimental criminology. In Handbook of Quantitative Criminology: 399-436, Piquero, A. and Weisburd, D. (eds.). Springer: New York.

Sherman, L., Buerger, M., and Gartin, P. (1989). Beyond dial-a-cop. A randomized test of repeat call policing (RECAP). Crime Control Institute: Washington, D.C.

Star Ledger. (2010, November 30). Newark finalizes 167 police layoffs after union refuses Booker's plea to return to negotiating table. Retrieved 5/30/15 at: http://www.nj.com/news/index.ssf/2010/11/union_head_expects_167_newark.html.

Stephan, J. (2004). State prison expenditures, 2001 (NCJ Publication No. 202949). Washington DC: U.S. Department of Justice, Bureau of Justice Statistics.

Telep, C., Mitchell, R., and Weisburd, D. (2014). How much time should the police spend at crime hot spots? Answers from a police agency directed randomized field trial in Sacramento, California. Justice Quarterly, 31(5): 905-933.

Tonry, M. (2015). The fog around cost-of-crime studies may finally be clearing. Prisoners and their kids suffer too. Criminology and Public Policy, 14(4): 653-671.

Tuttle, B (2009). How Newark became Newark. Rutgers University Press: Piscataway, NJ.

U.S. Census Bureau (2010). State and county quick facts. Washington DC: United States Census Bureau. http://quickfacts.census.gov. Accessed 4 April 2015.

Wayson, B. and Funke, G. (1989). What price justice? A handbook for the analysis of criminal justice costs. National Institute of Justice: Washington, D.C.

Wiseman, J. (2011). Strategic cutback management: Law enforcement leadership for lean times. National Institute of Justice: Washington, D.C.

Welsh, B. and Farrington, D. (2002). Crime prevention effects of closed circuit television: A systematic review. London: Home Office (Research Study No. 25).

Welsh, B. and Farrington, D. (2004). Surveillance for crime prevention in public space: Results and policy choices in Britain and America. Criminology and Public Policy, 3(3): 497-526.

Welsh, B. and Farrington, D. (2009). Public area CCTV and crime prevention: An updated systematic review and meta-analysis. Justice Quarterly, 26 (4): 716-745.

Welsh, B., van der Laan, P., and Hollis, M. (2013). Systematic reviews and cost-benefit analysis: Toward evidence-based crime policy. In Welsh, B., Braga, A., and Bruinsma, G. (eds.). Experimental criminology. Prospects for advancing science and public policy: 253-276. Cambridge University Press: New York, NY.

Weimer, D. (2008). Introduction: cost-benefit analysis and public policy. In Weimer, D. (ed.) Cost-benefit analysis and public policy. Journal of Policy Analysis and Management. Wiley Publishing: Somerset, NJ.

Wickramasekera, N., Wright, J., Elsey, H., Murray, J. and Tubeuf, S. (2015). Cost of crime: A systematic review. Journal of Criminal Justice, 43(3): 218-228.

Wiseman, J. (2011). Strategic cutback management: Law enforcement leadership for lean times. U.S. Department of Justice, Office of Justice Programs, National Institute of Justice.

Wyant, B., Taylor, R., Ratcliffe, J, and Wood, J. (2012). Deterrence, firearm arrests, and subsequent shootings: A micro-level spatio-temporal analysis. Justice Quarterly, 29(4): 524-545.

Zedlewski, E. (2009). Conducting cost benefit analyses in criminal justice evaluations: Do we dare? European Journal on Criminal Policy and Research, 15(4): 355-364.

Comments
0
comment

No comments here