Intelligence-Led Policing in Practice: Reflections From Intelligence Analysts

Abstract

Intelligence-led policing (ILP) is a managerial law enforcement model that seeks to place crime intelligence at the forefront of decision-making. This model has been widely adopted, at least notionally, in the United States, United Kingdom, Canada, and Australia. Drawing on interviews with intelligence analysts from two Australian state law enforcement agencies, this article contributes to the relatively small body of literature that has examined ILP in practice. The article identifies three relational themes that inhibit the successful implementation of ILP: analysts and data, analysts and tools, and analysts and decision makers. Furthermore, it calls attention to the need to better understand the structure and operations within law enforcement agencies, including the similarities and differences among organizational units, in order to better understand how these nuances shape the practice of ILP.

Keywords

intelligence-led policing crime intelligence intelligence analysts law enforcement

Introduction

Intelligence-led policing (ILP) seeks to move from a “reactive” or “prosecution-directed mode” of policing to a more “proactive” style of crime prevention (Innes, Fielding, & Cope, 2005, p. 41; Innes & Sheptycki, 2004, p. 1). This involves a shift away from responding to individual incidents and “threats” in isolation to one that is future orientated and strategic in nature (J. G. Carter, 2013; Ratcliffe, 2016). ILP seeks to achieve this by placing what is often referred to as “crime intelligence” at the forefront of decisions concerning the prevention and control of crime (Guidette & Martinelli, 2009, p. 132; Ratcliffe, 2016, p. 89).¹ Under this model, crime intelligence should guide operations “rather than operations dictating intelligence-gathering priorities” (Ratcliffe, 2016, p. 6). While gaining momentum for some years prior to this point, ILP was widely adopted throughout the United Kingdom following the adoption of the National Intelligence Model (NIM) by the U.K.’s Association of Chief Police Officers in 2000. The NIM, in short, is a “business model” that takes an “intelligence-led approach to policing” (Belur & Johnson, 2016; James, 2013; National Centre for Police Excellence, 2005, p. 8) and in 2004, the U.K. Government legislated that all police forces in England and Wales were to have fully implemented the NIM by the following year. ILP has since been widely adopted in many countries including the United States, Canada, Australia, and New Zealand (J. G. Carter, 2013; Gul, 2009; Ratcliffe, 2016).

Crime intelligence or the intelligence process, more broadly, is often conceptualized in terms of ideal models. The most well-known of such models is the “intelligence cycle.” There are numerous versions of the intelligence cycle; however, they broadly follow the same six stages: planning, collection, collation and evaluation, analysis, dissemination, and feedback (Gill & Phythian, 2013; Warner, 2013). The most common criticism of the intelligence cycle is that the practice of intelligence is much more fluid than as it is depicted in the model. A further criticism is how it views intelligence in isolation, paying no consideration to the broader law enforcement environment in which it sits, including a failure to articulate how intelligence analysis is linked to decision-making. In contrast, Ratcliffe’s (2016) 3-I model seeks to clearly illustrate the role of intelligence within policing. The 3-I model is not necessarily a representation of how intelligence functions within law enforcement currently, but a depiction of how law enforcement organizations should function if they are to be truly intelligence-led. The 3-I model consists of three components (crime intelligence analysis, decision makers, and criminal environment) and three processes connecting these components (interpret, influence, and impact), creating a triangle (Gul, 2009). There are two main differences between the intelligence cycle and the 3-I model, and they primarily relate to the latter two “Is.” The first is the significance placed on crime intelligence analysts having an influence over decision makers. The second, in turn, is for decision makers to actually implement various initiates who do have an impact on crime. This is perhaps the most contentious component of this model and is a key difference from other models that often see the creation of an intelligence product and its dissemination as the conclusion of this process (Coyne & Bell, 2011; Ratcliffe, 2016).

Following the adoption of ILP, at least in principle, by numerous law enforcement agencies in the late 1990s and early 2000s, several studies have examined the use of intelligence analysis within policing (e.g., Cope, 2004; Darroch & Mazerolle, 2013; Ratcliffe, 2005; Sanders, Weston, & Schott, 2015). For example, Cope (2004) conducted a total of 38 interviews with intelligence analysts, research officers, informant handlers, and intelligence managers from both county and urban police forces in the United Kingdom. While it was clear that analysis as a tool for guiding police activity was generally accepted, a poor understanding of what intelligence analysis is among police officers and a limited understanding of policing among analysts had a negative impact on the usefulness of intelligence products. Cope (2004, p. 201) concluded that there is a “chasm between the theory and practice of intelligence-led policing and the current role of analysis.” In a similar study, Ratcliffe (2005) interviewed 50 decision makers and intelligence staff from New Zealand Police, finding that there was a perceived lack of understanding of ILP at senior levels of the organization. Concerns were also expressed about the appropriateness of the training intelligence staff receive and the quality and availability of data, which can significantly impact the practice of ILP.

A number of studies have also highlighted the numerous challenges with implementing ILP (Darroch & Mazerolle, 2013; Fyfe, Gundhus, & Ronn, 2017; Sanders et al., 2015; Sheptycki, 2004). Sheptycki (2004, p. 307) identified 11 “organisational pathologies” that “bedevil” law enforcement information systems, including the lack of interoperability, nonreporting and nonrecording of relevant intelligence, intelligence overload, information silos, and duplication. In a more recent study, Darroch and Mazerolle (2013) conducted a survey of 286 officers and 32 interviews with the members of New Zealand Police to identify the organizational factors, including changes in policy, procedure, training and leadership, that influence the adoption of ILP. Darroch and Mazerolle suggested that leadership is particularly critical to the adoption of ILP, noting that “where leadership lacked a critical mass or was deficient in the command structure, ILP [intelligence-led policing] innovation struggled” (p. 24). This is supported by a detailed examination of the implementation of the NIM in the United Kingdom, which found that opposition from senior commanders had contributed significantly to the NIM’s overall failure to have a “meaningful impact on operational practice” (James, 2013, 2014, p. 79). In a Canadian context, Sanders et al. (2015, p. 711) conducted 86 in-depth interviews and participant observation at six different police services, suggesting that these agencies had rhetorically adopted ILP but have translated “it to mean accountability in a time of austerity.” As Sanders et al. (2015, p. 722, emphasis in original) explained, “analysts are tasked with providing data to both justify current human and financial resources as well as illustrate the need for additional investment.” In other words, they have not actually implemented ILP but instead have translated it into an accountability tool for justifying their expenditures. Finally, it has been suggested by Fyfe et al. (2017) that implementing ILP will challenge the hierarchical chain of command in most law enforcement agencies due to its push for decision-making to be informed by crime intelligence. As such, for Fyfe et al., significant cultural change among police managers is essential for ILP to be successful.

While it is clear that ILP is widely adopted in theory, the extant body of literature illustrates that implementing it in practice is far from straightforward. Each empirical study highlights new challenges that seemingly limit the extent to which police organizations do in fact follow the principles of ILP. It is, furthermore, also evident that empirical research on ILP has not kept up with developments in policy and practice. For example, J. G. Carter (2016) observed that ILP has still received a relatively small amount of scholarly attention compared with others models of policing, such as community-oriented policing. Furthermore, Lewandowski, Carter, and Campbell (2017) suggested that the “nuances” within law enforcement agencies (e.g., cultural differences) are critical to our understanding of intelligence and require further study in a variety of contexts. Without continued examination, the risk, according to Ratcliffe (2016, p. 176), is that “it is easily possible for intelligence-led policing to fall by the wayside and become lost in the history of failed attempts of law enforcement to move away from the traditional focus on reactive, investigative policing.”

This study contributes to the extant literature by providing a detailed qualitative analysis of key impediments to the successful implementation of ILP from the perspectives of intelligence analysts. It draws on in-depth interviews with analysts from two Australian state law enforcement agencies, Victoria Police and New South Wales Police Force (NSWPF). We argue that although important changes have occurred, including greater recognition of intelligence and its value, it would be premature to state that they are truly intelligence-led organizations. In particular, it was evident that a great deal of diversity exists within law enforcement agencies that impact the practice of ILP in different ways. While the focus of this study is from an Australian law enforcement perspective, the findings have implications for all agencies that have adopted ILP.

The remainder of this article is divided into three sections. The first section outlines the methodology adopted for this study. The second section presents the core findings of this study that emerged from the interview data, focusing on three relational aspects that appear to inhibit ILP: analysts and data, analysts and tools, and analysts and decision makers. The third section discusses the main findings of the study, the implications for our understanding of ILP and areas requiring further research.

Data and Method

This article is a part of a larger project that examined intelligence and specific analytical tools within law enforcement. The project involves in-depth interviews with intelligence analysts from the two largest Australian state law enforcement agencies, Victoria Police and NSWPF. Victoria Police (2016) services a population of just under 6 million people across 237,639 km², has a staff of approximately 17,000, and an annual budget of $2.51 billion (AUD). NSWPF (2016) services just over 7.6 million people across 800,642 km², has a staff of approximately 20,500, and an annual budget of $3.47 billion (AUD). It should also be noted that law enforcement in Australia is state-based, meaning Victoria Police and NSWPF service not only large municipalities but also small regional towns.

A research proposal was submitted to the relevant committees from Victoria Police and NSWPF, outlining the objectives of the project, requirements from each organization, research conditions, and expected outcomes. Both committees approved the proposal and assigned a liaison officer to facilitate the initial contact with analysts. Purposive sampling was used whereby an e-mail containing the original research proposal was provided to each liaison officer, who then forwarded it to the intelligence analysts from their respective organizations. Initial response rates were low from both organizations (<10). A snowball sampling methodology was subsequently adopted whereby those who had responded were asked to identify any of their colleagues that they thought may wish to take part in the study (see Maxfield & Babbie, 2011). The result was the participation of 27 intelligence analysts: 17 from Victoria Police and 10 from NSWPF. The number of participants is comparable with similar studies in related fields (Chan, Brereton, Legosz, & Doran, 2001; Cope, 2004; Ratcliffe, 2005; Whelan, 2012).

Analysts came from a variety of units, with many having experience in both regional intelligence divisions that are primarily concerned with local crime, and specialist intelligence units/taskforces that are focused on particular groups (such as outlaw motorcycle gangs) or specific crime problems (such as drugs). The level of analytical experience varied among research participants, ranging from just 2 years up to 25 years. However, many had experience in other positions at their agency prior to becoming analysts, including as frontline officers (n = 19). There were also eight analysts in total who were “unsworn” (civilian) members of their organization.² As part of becoming an intelligence analyst, each participant had completed at least one internal analyst training program, with the more experienced analysts having completed multiple courses that had been updated over time (n = 19). Most interviewees had also undertaken a number of short courses (1–2 days) often focused on one software program or system. Twelve analysts had also obtained tertiary qualifications both outside their field (in disciplines such as political science) and directly related to intelligence analysis. It should be noted interviewees did not include those involved in non-intelligence roles, such as the collation of crime statistics, but are sometimes referred to as “crime analysts” or similar terms.

Interviews with Victoria Police analysts took place between September 2014 and January 2016 at general regional and metropolitan police stations. Interviews with NSWPF analysts took place over 2 days in December 2015 at their police headquarters and a metropolitan station. Interviews were semistructured and generally ran between 30 and 60 minutes. Each interview was audio recorded and transcribed, with a copy provided to each participant for their approval to ensure the accuracy of the transcript and to have the opportunity to redact any sensitive information that may have been inadvertently shared during the interview. In accordance with the research agreement with each organization involved in the study, participants remain anonymous.

Once the interviews were completed, the transcript was entered into QSR NVivo 11, where a thematic analysis was undertaken. A thematic analysis “is used to classify and organise data according to key themes, concepts and emergent categories” (Ritchie & Lewis, 2003, p. 220) and is “useful in capturing the complexities of meaning within a textual data set” (Guest, MacQueen, & Namey, 2014, p. 10). Each transcript was initially examined with a number of emerging broad themes or categories (e.g., “data”). Interview transcripts were subsequently re-examined and several subcategories identified (e.g., “data sharing”). While saturation occurred for select themes, particularly those associated with data and tools, this was not the case in relation to the need to better appreciate the difference between specialist units and the broader police intelligence division of labor (Sheptycki, 2017b).

There are limitations with the data and methodology applied in this study. Because participants come from only two state law enforcement agencies and the sample size is relatively small, the research is not necessarily representative. As such, generalizing from the findings should be undertaken with care. Although the focus of this article is on ILP at a state level, law enforcement agencies work extremely closely with security and intelligence agencies, particularly when it involves counterterrorism (J. G. Carter, 2013; Whelan & Dupont, 2017). In this study, federal law enforcement agencies are not represented, and nor are specialist units working with federal partners on these cases. As with all interviews, there is also a concern that there may be a difference in what participants say in any interview compared with what they truly think. Nonetheless, the data presented in this article provides valuable insight into the practice of ILP.

Findings

Three relational themes emerged from this study: analysts and data, analysts and tools, and analysts and decision makers. These themes represent core components of the practice of intelligence analysis and ILP.

Analysts and Data

Access to accurate data is an implicit first stage of the intelligence analysis process (Chan & Moses, 2017; NCPE, 2005). However, in line with the previous research (Ratcliffe, 2005; Sheptycki, 2004), interviewees identified a number of issues with their data, including incompleteness (e.g., missing data) and incorrectness (e.g., information entered incorrectly into their internal criminal databases). Several interviewees suggested that they will be inevitably working with incomplete data:

We often work with very incomplete information, that’s what intelligence is.

It’s not the CSI [Crime Scene Investigation—a television show] button, press one button and find everything out about everyone, that just doesn’t work. So limitations are what we’re able to see, what we are able to get, the information we don’t know, and that’s why it’s just an ongoing process . . . [W]hen you’re dealing with people that don’t want to get caught, it’s not as easy as picking up the breadcrumbs, often you can’t find stuff out, so that’s where its limited.

When I talk intel [intelligence] or when we talk intel through what we do as analysts, it’s a grey area, it’s not black and white. . . . The issue with intel is to determine not only what you know but also what you don’t know.

One of the key impacts of incomplete data, according to interviewees, is that there are times when analysts have insufficient information to base their assessments on. As one interviewee explains,

We really don’t have too much to base those assessments on. Using just the [name of internal criminal database] system which is only going to show people seen together at the same time, spoken to at the same time, somebody, some cop somewhere happened to think they were connected in some way. Sometimes the information that we have got to assess is not great, . . . just by looking at what we have, the information we have, it’s not really an accurate way of doing it.

This analyst suggests that by just relying on their own internal criminal databases they are likely to have a very incomplete picture of the individual, group, or crime type they are investigating, introducing potential concerns about the accuracy of intelligence assessments. However, according to interviewees, gaining access to further information is also inhibited by limited information sharing both between and within agencies. As one analyst suggests, “the information sharing in particular between different government bodies has got to change, . . . you’ll see badness [offending] occur where that information is siloed.” There are many legitimate reasons for siloing information as an essential means of information protection; however, it has also been noted that such “information silos” or police “empires” are the result of a number of factors, including over reliance on security clearance restrictions, technological interoperability issues, and interagency competition/rivalry (D. L. Carter & Carter, 2009; Sanders & Henderson, 2013; White, 2004). While research has also highlighted the importance of intra-organizational sharing of information and knowledge, including between different units such as specialist squads (e.g., Giacomantonio, 2014), internal networks have been much less researched than interagency networks (Whelan & Dupont, 2017). Interestingly, interviewees in this study suggested that information sharing is just as problematic within their own internal boundaries, as it is between organizations. For example, one analyst suggests:

That’s [information sharing] something we are bringing up constantly. We don’t have established . . . networks within the organisation and also externally. There’s definitely silos [of information] which we are trying to break, . . . just having more open communication across all the different squads because there’s always overlaps, people try to hold onto their work. I think we need to work together a bit more.

Clearly, a lack of information sharing, both internally and externally, goes against the principles of ILP, which at its core is “envisioned as a tool for information sharing” (D. L. Carter & Carter, 2009, p. 315). What this does highlight, however, is that even if police organizations seek to be truly intelligence led, they are reliant on the reciprocity of partner agencies and collaboration between internal work units. A police organization may themselves have adopted the principles of ILP but may fall short of being truly intelligence led without partner organizations adopting a similar approach.

A further issue is data or information overload. While there is usually a desire to have access to more information, this can also create a great deal of distraction:

I think you can maybe bring in too much information, do you know what I mean? It would be great to bring in all this stuff [information] to do target profiles and know everything about him [offenders]. [But] on the other hand, it will actually bring in a lot of stuff that’s just not relevant to the investigator.³

Data overload have long been recognized within law enforcement (Chan et al., 2001; Hauck, Atabakhsh, Ongvasith, Gupta, & Chen, 2002; Sheptycki, 2004; Whelan, 2017). Despite this, several interviewees suggested that it is becoming more problematic. As one participant explains,

there is so much out there now and actually the sheer volume of having to go through it, because yes, you can go through it with your queries [database searches] but you still have to have a human element quite often to just sort of . . . put two and two together.

As this analyst suggests, the problem of data overload had been compounded by the rise in the volume and veracity of the information now collected by police (Joh, 2014; Sanders & Sheptycki, 2017), which now includes e-mails, text messages, computer hard drives, and instant messaging (Décary-Hétu & Dupont, 2012, p. 162). New sources of information is perhaps best illustrated by the rise of social media (e.g., Facebook), where a recent study found that 81% of “law enforcement professionals” in the United States use social media as one source of information during investigations and one quarter use it daily (LexisNexis, 2014). This is supported by this study, with one interviewee suggesting that “one of the first checks we do is Facebook.” Interviewees also express concerns about the reliability of social media as an intelligence source. This is explained in detail by one analyst:

One of the big new things is social media and you can look through peoples’ interactions on social media, depending on their privacy settings, but the problem with that is you might call it the Facebook friend. You might be able to see that someone has 200 friends on Facebook and you might be able to see that their Facebook friends with so and so, and you go, wow. But even from personal experience, you’ve got people that you interact with on social media, there are some people’s friend invites that you just accept because you don’t want to hurt their feelings, but other than that you sort of forget that they are even there, and that’s the problem with social media analysis, it’s very tricky to identify who is a true [friend]. . . . [I]t’s easy to overstate connections between people because we just don’t have the visibility into the quality of that connection.

This analyst suggests that it is difficult to determine the value of the information contained on such social media sites, particularly in regard to determining the quality of the relationships between people. This raises questions around the circumstances with which analysis should and should not be incorporating this into their analysis as the apparent indiscriminate use of social media analysis is highly likely to exacerbate the problem of data overload.

Analysts and Tools

The “tools” or software and systems required to collect and analyze data were also perceived to be lesser than ideal by interviewees. For example, at the core of the data overload, issue discussed earlier is the capacity of law enforcement software and systems to assist analysts in analyzing growing quantities of data. Almost all interviewees, however, expressed a general concern about the systems available to them:

We’ve got horrendous IT systems in [name of law enforcement organisation].

Technology changes daily and we’re still struggling to get information systems to work. They didn’t work when we got them, and we’re still struggling, and the pressure is put on the members.

Suitable software that works, that doesn’t crash would be nice.

While systems have developed considerably over the past two decades, making data processing far more “user-friendly” (Duijn & Klerks, 2014, p. 150), the rate of data collection and retention has also rapidly increased during this time (Brodeur & Dupont, 2006; Décary-Hétu & Dupont, 2012; Hutchins & Benham-Hutchins, 2010; Sheptycki, 2000). The impact of this massive push to continually collect more data means that they have exceeded their ability to fully analyze this information.

While software and systems might be regarded as separate from the training analysts receive, in the context of this study, it is evident that they are intrinsically linked. In a similar way to the extant literature (Cope, 2004; Ratcliffe, 2005), interviewees suggested that the training they receive for the software and systems available to them does not meet their needs. In particular, several analysts highlighted their frustration with the inconsistency with which training is available. As one interviewee explains,

They [training programs] came in and out of vogue, and weren’t funded at different times. So sometimes we had the [software] licence but no funding for training, and other times there was no licence and no funding, and then at one stage there was. We eventually got to the point where we had licences and training funding, but the training funding was limited, so people had to get in quickly to get what they could. . . . I know there was a lot of angst years ago, when funding came for [name of software program], came and went, the system was available then it wasn’t available again. Then it was available with lots of training which was fantastic, then the training wasn’t funded so we had to run our own internal course and try and show people how we did it.

This analyst went on to say that it was their belief that things have improved in this regard within their organization, but it illustrates how inconsistencies in the availability of training can significantly impact on analysts’ ability to utilize the tools available to them. One interviewee even suggested that law enforcement agencies may be better served by providing the right tools and by upskilling their current analytical work force compared with additional human resources:

I even think that we’ve got enough, dare I say, it’s always handy to have more human resources but I think we do have a large contingent of analysts out there that would be able to accomplish a lot more on a percentage basis if you gave them the right tools rather than just throwing more bodies at it. Throwing more human resources into things isn’t necessarily the answer and I think that seems to be a default quite often. So skilling them up [training] and giving them the tools that they actually require.

To further emphasize this point, this analyst suggested that “a number of the tools that we have access to have capabilities far beyond what the user skill levels are” and that those who can utilize these software programs/systems to their full potential is “still limited to 5% to 10% of the intelligence analysts that are out there.” While the validity of this assertion cannot be tested, it does highlight the extent to which law enforcement agencies may be underutilizing the resources at their disposal by not providing adequate training for their software and systems.

Analysts and Decision Makers

According to Ratcliffe’s (2016) 3-I model, there are many possible clients that analysts need to consider when looking to influence decision makers, including frontline officers (such as patrol officers and detectives), police leadership, regulatory agencies, the general public, and security networks. Interviewees, however, only identified two primary sets of decision makers: police leadership and detectives.⁴

It has been suggested that positive working relationships—that is, those characterized by attributes such as mutual respect and reciprocity—lead to improved information sharing (e.g., D. L. Carter, 2009) and increased levels of trust (e.g., Schafer, 2010). In this study, analysts similarly suggested that their working relationships with managers and detectives are critical to achieving their objectives. Several analysts, however, suggested that their working relationship with intelligence managers is not always ideal. For example, one interviewee detailed the importance of what they perceive to be “good intelligence managers”:

Having a good intel manager who can say look, that’s really not intel work, that’s busy work, why are you using trained intel’s [intelligence analysts] to do this when you could do that yourself or we could just get that information from anywhere. These guys have skills that should be used in a particular way. Your managers have to be the sorts of managers that can use intel effectively. . . . There are people who are no longer with the organisation who I think intels [intelligence analysts] will be spitting on their graves for the next 20 years to come.

Research participants suggested that this tension is in part caused by senior managers not fully understanding what intelligence is and how it is produced, despite regarding intelligence as a valuable resource informing their decision-making. This is explained by one senior analyst:

I think they [managers] understand its [intelligence] function, functionally where it fits in the process, they know they need the intelligence to advise them so they are better informed to make their decisions. . . . I think our managers have known it for a long time. But their ability to really understand how it works, they still think it’s a little bit of black magic and voodoo, still getting their head around it. I think they understand its [intelligence] importance, but no, they don’t understand it to the degree, again, largely because there probably hasn’t been that exposure to it from the time they start all the way through. . . . So I think there is a real disconnect in their own understanding of how intel works. So that’s something that I think needs to change over a period of time.

This finding is broadly in line with previous research suggesting that senior managers have been ill-prepared for the shift toward ILP, where intelligence should drive decision-making (Phillips, 2012; Ratcliffe, 2016; Sanders et al., 2015). For example, Ratcliffe (2016, p. 134) suggests that “we simply have not prepared the command levels of policing for the new role they are being asked to perform.” This has been recognized as an issue within law enforcement for some time (Cope, 2004; Ratcliffe, 2005), but interviewees in this study would lead us to conclude that little has changed. Although unlikely to have a short-term impact, one analyst believes the solution to this issue is greater training and awareness for new recruits about intelligence as both a process and a product:

We need to shift our thinking to that ILP and training from grass roots, pretty much all through our organisation. [Be]cause by the time they are at Senior Sergeant, Inspector level, they start to understand the tasking and coordination process and they know they need the intel [intelligence]. But until then, . . . I think there is a huge knowledge gap in our core element of policing, about their role and [the] function of intelligence.

Interviewees also considered detectives to be a primary decision maker due to the fact that they will be usually in charge of investigations. As one analyst explains, “in the sense of guiding or directing the path of an investigation it has to be said first of all that the officer in charge of an investigation is almost always a detective.” The majority of interviewees also considered their working relationship with detectives to be critical to undertaking their role. As one analyst explains,

Your collaboration with an investigator is key to making the role of intelligence as successful as it can be. . . . So its providing that analytical support to basically, you know, the key objective of putting those people who’ve committed those crimes in jail through whatever means possible. . . . [I]t’s [intelligence analysis] to me, using everything that is at my disposal to give the investigator the ability to put that person inside.

It was also clear from the research participants that during some investigations their working relationships with detectives would be characterized as positive and subsequently more intelligence led. In this instance, interviewees appeared to regard their relationships with detectives as positive when they were meaningfully included in the investigative process. The positive nature of some analyst and detective relationships is summarized by one interviewee:

I know that in different squads there is different dynamics between teams. Some are more intelligence-led than others. The operational team that I previously worked with, it was very collaborative in terms of we [analysts] had a lot of financial data that they [investigators] didn’t have but they had a lot of surveillance and human source data. So it was combining those that allowed us to move forward and sort of show us what areas we needed to collect against and the same for them, what areas they needed to target. . . . So it needs to be collaborative.

In instances where analysts feel they have a positive working relationship with detectives, it appears they feel confirmed in their role within what Sheptycki (2017b, p. 620, emphasis in original) refers to as the “police division-of-labour”; the “intra-organisational information and intelligence flows bounded within the archetypal form of the multi-functional urban police service.” The majority of interviewees, however, also had experiences where the working relationship with detectives could be characterized as negative either in terms of being underappreciated or entirely excluded. This is discussed by several analysts:

[At] the divisional CIU’s [Crime Investigation Unit] the relationships aren’t that good. It’s not that good with the detectives because they think what the hell do you know, get out. So it’s actually up to us to push our way into investigations. They will have debriefings and not even call us in, and then the boss says, what are you doing [analyst’s name]? Nothing because they’ve not come to me, they haven’t invited me in there, they can shove it up their ass, they’re not getting anything.

Sounds silly but some squads you either get totally ignored by the detectives or you’re running around like the coffee lady, the tea lady.

There were differing views as to what contributed to such negative working relationships. For example, one analyst suggested that this may be a legacy of traditional policing practices, where intelligence analysts had little to no part in the investigation process: “people who have come up through more traditional policing backgrounds may not necessarily have as close a relationship with the intelligence team.” Another analyst suggested that it may in part be the result of a former practice of placing soon-to-retire officers (“dinosaurs”) into intelligence positions: “I know they’re trying to get rid of the dinosaur out there. Well that’s what they say, get rid of the guy that’s just come to the DIU [Divisional Intelligence Unit] to retire and doesn’t want to do anything.” It is well-documented that intelligence analysis has historically been marginalized by frontline officers, detectives, and senior management (Cope, 2004; O’Shea & Nicholls, 2002; Sanders et al., 2015; Taylor, Kowalyk, & Boba, 2007), suggesting that placing disinterested officers into intelligence positions undermines the value of intelligence analysis. Many interviewees, however, felt that one of the key factors contributing to poor working relationships between some analysts and detectives is that many investigators do not fully understand what intelligence is and, in particular, what analysts can actually do for them. These views are expressed by several analysts:

Sometimes they [detectives] don’t know what they want or they ask for things that aren’t possible to get. You’ve got a system that does X, Y and Z. No, it does Y, it doesn’t do Z, it doesn’t do X.

I’m only generalising here . . . [but] detectives don’t know what we can do for them and that’s critical. And sometimes jobs come in and I go what the hell, you haven’t even come to see us, and yet we could have helped you solve this. Seriously, they don’t know what they can do with phones half of them because phones are a very complex thing. They don’t know about social networking, what we can get from there.

There’s a distinction often lost, particularly in policing, . . . is the difference between intelligence and evidence. Intelligence is gathered by a number of means, may or may not be admissible as evidence in a court of law, but is intended to inform decision-making and not boost prosecution. Evidence has very specific laws under which it’s gathered so it can be admitted in court. A difficulty sometimes seen in law enforcement organisations is investigators confuse the two and are dissatisfied with its [intelligence analysis] outputs because they can’t use it in courts of law and they don’t see the value in them as a direction-finding mechanism.

This last analyst alludes to a difference between detectives and police managers. While there appears to be a consensus that both detectives and police managers do not fully understand what intelligence is and how it is produced, police managers consider it to be a valuable resource to support their decision-making. In contrast, according to this analyst, some detectives do not regard intelligence to be a valuable resource at all because it is not always prosecution oriented and may not meet evidentiary standards. There appears to be a consistent finding in which the primary clients of intelligence reports do not appear to properly appreciate intelligence (Cope, 2004; Phillips, 2012; Ratcliffe, 2005; Sanders et al., 2015). Not noted in the extant literature, however, is that this lack of knowledge of intelligence is perhaps less of an issue with what are referred to here as “more experienced detectives,” particularly those who have been involved in large-scale and long-term investigations that tend to have greater levels of intelligence analysis support. This view was suggested by several analysts, although it is explained in detail by one in particular:

In my observation, it depends a lot on how many prior investigations the investigator has worked on where they were supported by an intelligence officer. Many [investigators] will have worked at local area commands on many investigations where they had no input from an intelligence officer or the input was limited to being more or less a data acquisition officer. . . . When an investigator in our area has had only that experience, their expectations, their understanding of what they can get from intelligence staff is different, it’s much more limited, they may commence by only consulting the intelligence officer when they have a data need that they don’t know how to fulfil themselves. Once or twice they might have seen charts of particular types of data . . . and they may see the value in that in presenting what they have learnt as part of a prosecution, but they are only consulting the intelligence officer when the investigation is more or less complete. . . . Then investigators who have been working in major crime and large-scale investigations for a long period of time have been exposed to more intel officers, often incorporate them in the internal planning of the investigation in seeking judgments about what the intelligence officer thinks about the . . . potential for targeting different members of the criminal group subject to the investigation; asking them to estimate the likelihood that they can develop from an intelligence process about certain facts being or certain situations existing or certain eventualities occurring.

This interviewee highlights a common theme that has emerged from the interview data regarding the highly diverse nature of police organizations, echoing an earlier point that some units within law enforcement agencies are “more intelligence-led than others.” This suggests we need much greater understanding of how law enforcement agencies are structured and operate internally to truly appreciate how ILP is practiced.

Discussion and Conclusion

ILP has existed in theory ever since—if not, before—the introduction of the NIM (National Criminal Intelligence Service, 2000). Nonetheless, research has continued to demonstrate the many challenges involved in implementing ILP in practice, suggesting it continues to be more rhetoric than reality in policing organizations. Many explanations have been put forward as to the current state of ILP, including a poor understanding of what intelligence analysis is among police officers (Cope, 2004), various organizational pathologies (Sheptycki, 2004), and a lack of leadership (Darroch & Mazerolle, 2013; James, 2013). Research has also questioned the potential impact of “police cultures” (Chan, 1997; Whelan, 2016) in terms of resisting change and thereby a reluctance to move away from reactive or intuitive modes of policing (Belur & Johnson, 2016). In this study, we have focused on the impediments to the implementation of ILP, exclusively from the perspectives of analysts themselves. The main impediments to the effective implementation of ILP, as identified by analysts, fall under three categories: data, tools, and decision makers.

Data challenges have been widely acknowledged in the literature. In this study, interviewees concentrated on the accuracy of data (in terms of both incompleteness and incorrectness), the sharing of data (or information), and data (or information) overload. As Innes (2006), for example, notes, intelligence analysis is a field plagued with uncertainty, meaning many of the concerns regarding the accuracy of data cannot be fully addressed. However, it seems many of the issues raised by analysts relate in part to the data standards and practices within police organizations, suggesting at least some of these issues could be overcome with more rigor and scrutiny with regard to how data are obtained and recorded on police intelligence systems (see also Sheptycki, 2004). In terms of data sharing, interviewees repeatedly expressed their frustrations with regard to both the systems they have and their capacity to share information—including interoperability concerns—and the rules restricting how information and intelligence can be shared. In particular, analysts referred to a perceived lack of sharing between some organizational units (Giacomantonio, 2014), which reinforces the need to better appreciate the internal police division of labor (Sheptycki, 2017b). That is, while many have highlighted issues concerning the inter-organizational sharing of information (Whelan & Dupont, 2017), it is not the case that sharing information or intelligence occurs freely within organizations (Tsai, 2002). In this study, we even interviewed select intelligence analysts who were unable to share their products with their line managers because of these security classification requirements. Another interesting observation suggested by intelligence analysts is that, according to them, information overload is becoming a greater problem along with advances in technology but could not offer a solution to this issue. A prominent theme in the research on the rise of so-called big data (Chan & Moses, 2017; Cukier & Mayer-Schownberger, 2013; Sanders & Sheptycki, 2017), it appears the sheer growth in the volume of data is outpacing the capacity of analysts and organizations to meaningfully process and analyze it. What is clear, however, is that further efforts need to be directed toward organizational and technological improvements in the intra- and interorganizational coordination of police intelligence.

By tools, we have referred to the systems and software analysts have available to them to collect, analyze, and disseminate information and the training they receive to use such tools. As noted earlier, research on police databases and systems has repeatedly identified issues concerning the design and functionality of these systems (D. L. Carter, 2009; Ferrara et al., 2008; Sanders & Henderson, 2013). While we have identified many similar issues in this regard to the extant literature, our interviewees tended to lean more toward their perceived lack of training with regard to using the tools available to them. Analysts undertake an initial training course when they first join the organization (in the case of new members) or when they are first appointed as analysts (in the case of existing members), but interviewees referred more to limited opportunities for ongoing training. This lack of ongoing training is particularly problematic as tools change, including updates in existing applications (such as IBM i2 Analyst Notebook, which was used often by analysts in both organizations studied) as well as the introduction of completely new systems. Interviewees felt that the availability of ongoing training for many of these tools needed to improve. Interestingly, according to some analysts, they suggested that more training would produce better outcomes than more staff, which goes to highlight the significance of training as an impediment to intelligence analysis as a practice. As similar recent research has concluded (e.g., Belur & Johnson, 2016), better training for analysts and managers appears to be essential for organizations to truly implement ILP. Not discussed by interviewees, however, were the characteristics that are reportedly needed to be recognized as an effective analyst, including the ability to produce an analytical product, the overall attitude of an analyst (e.g., “can do” attitude), and individual attributes (e.g., confidence); all of which can inhibit the use of intelligence with law enforcement (Evans & Kebbell, 2012).

Influencing decision makers is a key component of ILP. We have concentrated on decision-making with regard to intelligence managers and senior police command and detectives or investigators working on specific cases. What is clear from both categories is that there is a great deal of inconsistency within law enforcement agencies when it comes knowledge of intelligence, including how it is produced and how it can inform decision-making. Interviewees repeatedly suggested that those managers with a greater understanding of intelligence were much more likely to appreciate its value and to therefore use intelligence as the basis of decision-making. In contrast, decision makers with a limited understanding of intelligence were more likely to, intentionally or unintentionally, marginalize analysts, including allocating them much less meaningful tasks. This understanding of intelligence largely comes down to the experience of each manager and detective in question, with interviewees repeatedly suggesting that some units are effectively much more intelligence led than others. Those who have spent time in more specialist units—including dedicated intelligence units as well as units dealing with organized crime and counterterrorism, for example—are much more likely to have a greater understanding of intelligence compared with those who have been limited to routine police operations.

Our analysis therefore suggests that in order to truly understand the role of ILP, we need to better appreciate how it is practiced differently within and between the many organizational units comprising police organizations. As Sheptycki (2017a) recently points out, conceptions of the “police métier” (Manning, 2010) and the formal and informal boundaries within policing organizations shape the function of intelligence in important ways. The vast majority of the extant research on crime analysis or ILP has focused on police organizations as a whole. This has led some to call for greater attention to be placed on the similarities and differences between police organizations (e.g., rural vs. urban police; J. G. Carter, 2013). We argue that there is just as great a need to look further inside police organizations to appreciate the similarities and differences between organizational units and the ways in which they shape the practice of ILP.

Footnotes

Acknowledgments

The authors would like to sincerely thank the anonymous reviewers for their critical input on an earlier version of this article.

Notes

Author Biographies

Morgan Burcher is a lecturer in criminology at Deakin University. Dr. Burcher’s research interests include intelligence analysis, policing, social network analysis, cybercrime, and organized crime. Recent publications have appeared in Trends in Organized Crime and Global Crime.

Chad Whelan is an associate professor in criminology at Deakin University. Dr. Whelan conducts research on cybercrime, organized crime, terrorism and security, and multiagency responses to crime and security problems across organizational boundaries and professional disciplines. Recent publications have appeared in various journals such as British Journal of Criminology, Global Crime, Policing, Policing and Society, and Security Journal.

References

Belur

Johnson

(2016). Is crime analysis at the heart of policing practice? A case study. Policing and Society, 28(7), 1–20.

Brodeur

J.-P.

Dupont

(2006). Knowledge workers or “knowledge” workers? Policing and Society, 16(1), 7–26.

Carter

D. L.

(2009). Law enforcement intelligence: A guide for state, local, and tribal law enforcement agencies. Retrieved fromhttps://it.ojp.gov/documents/d/e050919201-IntelGuide_web.pdf

Carter

D. L.

Carter

J. G.

(2009). Intelligence-led policing: Conceptual and functional considerations for public policy. Criminal Justice Policy Review, 20(3), 310–325. doi:10.1177/0887403408327381

Carter

J. G.

(2013). Intelligence-led policing: A policing innovation. El Paso, TX: LFB Scholarly.

Carter

J. G.

(2016). Institutional pressures and Isomorphism: The impact on intelligence-led policing. Police Quarterly, 19(4), 435–460.

Chan

(1997). Changing police culture: Policing in a multicultural society. Cambridge, England: Cambridge University Press.

Chan

Brereton

Legosz

Doran

(2001). E-policing: The impact of information technology on police practices. Retrieved from http://www.ccc.qld.gov.au/research-and-publications/publications/cjc/e-policing-the-impact-of-information-technology-on-police-practices.pdf

Chan

Moses

L. B.

(2017). Making sense of Big Data for security. British Journal of Criminology, 57(3), 299–319.

10.

Cope

(2004). Intelligence led policing or policing led intelligence? The British Journal of Criminology, 44(2), 188–203. doi:10.1093/bjc/44.2.188

11.

Coyne

J. W.

Bell

(2011). Strategic intelligence in law enforcement: A review. Journal of Policing, Intelligence and Counter Terrorism, 6(1), 23–39. doi:10.1080/18335330.2011.553179

12.

Cukier

Mayer-Schownberger

(2013). The rise of big data: How it’s changing the way we think about the world. Foreign Affairs, 92(3), 28–40.

13.

Darroch

Mazerolle

(2013). Intelligence-led policing: A comparative analysis of organizational factors influencing innovation uptake. Police Quarterly, 16(1), 3–37. doi:10.1177/1098611112467411

14.

Décary-Hétu

Dupont

(2012). The social network of hackers. Global Crime, 13(3), 160–175. doi:10.1080/17440572.2012.702523

15.

Duijn

P. A. C.

Klerks

P. P. H. M.

(2014). Social network analysis applied to criminal networks: Recent developments in Dutch law enforcement. In Masys

A. J.

(Ed.), Networks and network analysis for defence and security (pp. 121–159). Heidelberg, Germany: Springer.

16.

Evans

J. M.

Kebbell

M. R.

(2012). The effective analyst: A study of what makes an effective crime and intelligence analyst. Policing and Society, 22(3), 204–219.

17.

Ferrara

Martenson

Svenson

Hidalgo

Molano

Masdsen

A. L.

(2008). Integrating data sources and network analysis tools to support the fight against organised crime. In Yang, C.C., Chen, H., Chau, M., Chang, K., Lang, S.-D., Chen, P., Hsieh, R., Zeng, D., Wang, F.-Y., Carley, K.M., Mao, W., Zhan, J. (Eds), Proceedings of the IEEE ISI (pp. 171–182), Formosa. Berlin, Germany: Springer.

18.

Fyfe

N. R.

Gundhus

H. O. I.

Ronn

K. V.

(2017). Introduction: What is ILP? In Fyfe

N. R.

Gundhus

H. O. I.

Ronn

K. V.

(Eds), Moral issues in intelligence-led policing (pp. 1-2). Abingdon, England: Routledge.

19.

Giacomantonio

(2014). A typology of police organizational boundaries. Policing and Society: An International Journal of Research and Policy, 24(5), 545. doi:10.1080/10439463.2013.784302

20.

Gill

Phythian

(2013). From intelligence cycle to web of intelligence: Complexity and the conceptualisation of intelligence. In Phythian

(Ed.), Understanding the intelligence cycle. Florence, Italy: Taylor and Francis.

21.

Guest

MacQueen

K. M.

Namey

E. E.

(2014). Applied thematic analysis. Thousand Oaks, CA: Sage.

22.

Guidette

Martinelli

T. J.

(2009). Intelligence-led policing: Strategic framework. Police Chief, 76(10), 132–136.

23.

Gul

(2009). A partial test of the intelligence-led policing model (Unpublished doctoral dissertation). Kent State University, Kent, OH.

24.

Hauck

R. V.

Atabakhsh

Ongvasith

Gupta

Chen

(2002). Using Coplink to analyze criminal-justice data Computer, 35(3), 30–37.

25.

Hutchins

C. E.

Benham-Hutchins

(2010). Hiding in plain sight: Criminal network analysis. Computational and Mathematical Organization Theory, 16(1), 89–111. doi:10.1007/s10588-009-9060-8

26.

Innes

(2006). Policing uncertainty: Countering terror through community intelligence and democratic policing. The ANNALS of the American Academy of Political and Social Science, 605(1), 222–241.

27.

Innes

Fielding

Cope

(2005). The appliance of science? The theory and practice of crime intelligence analysis. British Journal of Criminology, 45(1), 39–57.

28.

Innes

Sheptycki

(2004). From detection to disruption: Intelligence and the changing logic of police crime control in the United Kingdom. International Criminal Justice Review, 14(1), 1–24.

29.

James

(2013). Examining intelligence-led policing: developments in research, policy and practice. Basingstoke, England: Palgrave Macmillan.

30.

James

(2014). Forward to the past: reinventing intelligence-led policing in Britain. Police Practice and Research, 15(1), 75–88.

31.

Joh

E. E.

(2014). Policing by numbers: Big data and the fourth amendment. Washington Law Review, 89(1), 35–68.

32.

Lewandowski

Carter

J. G.

Campbell

W. L.

(2017). The utility of fusion centres to enhance intelligence-led policing: An exploration of end-users. Policing: A Journal of Policy and Practice, 12, 177–193. doi:10.1093/police/pax005

33.

LexisNexis. (2014). Social media use in law enforcement: Crime prevention and investigative activities continue to drive usage. Retrieved from http://www.lexisnexis.com/risk/downloads/whitepaper/2014-social-media-use-in-law-enforcement.pdf

34.

Manning

P. K.

(2010). Democratic policing in a changing world. Boulder, CO: Paradigm Publishers.

35.

Maxfield

M. G.

Babbie

E. R.

(2011). Research methods for criminal justice and criminology. Belmont, CA: Wadsworth Cengage Learning.

36.

National Criminal Intelligence Service. (2000). The National Intelligence Model. Retrieved from http://www.intelligenceanalysis.net/National%20Intelligence%20Model.pdf

37.

National Centre for Police Excellence. (2005). Guidance on the National Intelligence Model. Retrieved from https://www.intelligencesolutions.co.nz/media/1715/guidance%20on%20the%20nim%202005.pdf

38.

New South Wales Police Force. (2016). NSW Police Force annual report 2015-2016. Retrieved fromhttps://www.police.nsw.gov.au/__data/assets/pdf_file/0010/474823/NSW_Police_Force_Annual_Report_2015-16.pdf

39.

O’Shea

T. C.

Nicholls

(2002). Crime analysis in America. Washington, DC: Office of Community Oriented Policing Services.

40.

Phillips

S. W.

(2012). The attitudes of police managers toward intelligence-led policing. FBI Law Enforcement Bulletin, 81(9), 13–17.

41.

Ratcliffe

(2005). The effectiveness of police intelligence management: A New Zealand case study. Police Practice and Research, 6(5), 435–451.

42.

Ratcliffe

(2016). Intelligence-led policing. New York, NY: Routledge.

43.

Ritchie

Lewis

(Eds). (2003). Qualitative research practice: A guide for social science students and researchers. London, England: Sage.

44.

Sanders

C. B.

Henderson

(2013). Police ‘empires’ and information technologies: uncovering material and organisational barriers to information sharing in Canadian police services. Policing and Society, 23(2), 243–260. doi:10.1080/10439463.2012.703196

45.

Sanders

C. B.

Sheptycki

(2017). Policing, crime and ‘big data’ towards a critique of the moral economy of stochastic governance. Crime, Law and Social Change, 68(1-2), 1. doi:10.1007/s10611-016-9678-7

46.

Sanders

C. B.

Weston

Schott

(2015). Police innovations, ‘secret squirrels’ and accountability: Empirically studying intelligence-led policing in Canada. British Journal of Criminology, 55(4), 711–729.

47.

Schafer

J. A.

(2010). The ineffective police leader: Acts of commission and omission. Journal of Criminal Justice, 38(4), 737–746. doi:10.1016/j.jcrimjus.2010.04.048

48.

Sheptycki

(2000). Editorial reflections on surveillance and intelligence-led policing. Policing and Society, 9(4), 311–314.

49.

Sheptycki

(2004). Organizational pathologies in police intelligence systems: Some contributions to the lexicon of intelligence-led policing. European Journal of Criminology, 1(3), 307–332. doi:10.1177/1477370804044005

50.

Sheptycki

(2017a). Liquid modernity and the police métier; thinking about information flows in police organisation. Global Crime, 18, 286–302. doi:10.1080/17440572.2017.1313734

51.

Sheptycki

(2017b). The police intelligence division-of-labour. Policing and Society, 27(6), 620–635. doi:10.1080/10439463.2017.1342645

52.

Taylor

Kowalyk

Boba

(2007). The intergration of crime analysis into law enforcement agencies: An exploratory study into the perceptions of crime analysts. Police Quarterly, 10(154), 154–169. doi:10.1177/1098611107299393

53.

Tsai

(2002). Social structure of ‘coopetition’ within a multiunit organization: Coordination, competition, and intraorganizational knowledge sharing. Organization Science, 13(2), 179–190.

54.

Victoria

Police

. (2016). Annual report 2015-2016. Retrieved from http://www.police.vic.gov.au/content.asp?a=internetBridgingPage&Media_ID=120742

55.

Warner

(2013). The past and future of the intelligence cycle. In Phythian

(Ed.), Understanding the intelligence cycle (pp. 9–20). Florence, Italy: Taylor and Francis.

56.

Whelan

(2012). Networks and national security: Dynamics, effectiveness and organisation. London, England: Ashgate.

57.

Whelan

(2016). Organisational culture and cultural change: A network perspective. Australia and New Zealand Journal of Criminology, 49(4), 583–599.

58.

Whelan

(2017). Managing dynamic security networks: Towards the strategic managing of cooperation, coordination and collaboration. Security Journal, 30(1), 310–327.

59.

Whelan

Dupont

(2017). Taking stock of networks across the security field: A review, typology and research agenda. Retrieved from http://www.tandfonline.com/doi/abs/10.1080/10439463.2017.1356297?journalCode=gpas20

60.

White

J. R.

(2004). Defending the homeland: Domestic intelligence, law enforcement, and security. Belmont, CA: Wadsworth.