Abstract
A growing body of research examines factors that influence the likelihood of solving homicide cases. Much of this research emanates from North America and is based on quantitative analysis of police data. This article explores the views of homicide detectives, complemented by observations of investigations, in both Great Britain and the United States, regarding factors that affect the chances of solving homicides. Although we find some important differences between nations, the qualitative evidence suggests that the likelihood of solving even the most challenging homicide cases in both nations can be influenced by police agency at the individual and strategic level.
Keywords
Introduction
Homicide is generally regarded as the most serious type of criminal offense. In many instances, it is viewed as what Innes (2014) described as a “signal crime,” or an incident that sends a message to the public that all is not well in a particular locality, thereby increasing concerns about safety and security. Thus, although homicide is rare compared with other forms of crime, homicide investigations and their outcomes are subject to considerable scrutiny both from within police organizations and from multiple external sources, including the public, the media, and different levels of government. Indeed, success or failure in “solving” homicides (as reflected in prominent individual cases or in overall rates) is often taken as a general barometer of police effectiveness (Brookman & Innes, 2013). The importance of understanding how and why homicides are solved is highlighted by the fact that, despite major advances in forensic science, homicide clearance rates have declined significantly in the United States (and to a much lesser extent in the United Kingdom) since the 1960s.
A growing body of research has examined the factors thought to influence homicide case outcomes. The English language research on this topic is dominated by studies from North America, the majority of which rely on multivariate statistical methodologies. Such studies offer many benefits, one of which is that they provide estimates of the relative influence of multiple explanatory factors on case outcomes. However, they also have limitations. Their focus tends to be on individual case-related variables, such as the characteristics of the victim or location of the incident, rather than on police-related factors such as the level and quality of investigative resources employed. Alternatively, they focus on broader social or cultural factors such as the cohesiveness of the local community or their attitudes toward the police (Maguire, King, Johnson, & Katz, 2010). One result of this is that most of the variables included in the analysis are presented as “unalterable facts,” beyond the control of the police. This can create a somewhat passive and deterministic view of homicide investigations in which the actions of the police appear to have relatively little influence on the outcome. Such methodologies can also fail to capture the effect of complex and dynamic relationships between the police and particular communities or social groups. Perhaps most important, they leave open many questions about the reasons behind the statistical associations they identify.
A much smaller body of research has relied on qualitative approaches, including field observations and interviews with homicide investigators. While qualitative approaches raise certain inferential challenges, and cannot conclusively demonstrate which factors have an impact on outcomes, they also offer several key advantages. These include the ability to gather richly detailed data, to explore topics for which systematic quantitative data are not available, and to gain a deeper understanding of how and why certain factors may influence outcomes (King, Keohane, & Verba, 1994). In this article, we present findings based on qualitative interviews in the United States and Britain in which homicide detectives were asked to identify and discuss factors that they believe influence homicide case outcomes. By focusing upon detectives’ own insights combined with fieldwork observations of active investigations, we hope to provide a more nuanced understanding, especially about the role of police policy, capability, culture, and practice, including attitudes toward and dealings with different types of community. Some of the factors discussed are equally relevant in the American and British contexts, while others seem much more significant in one nation than the other.
We begin by providing a review of previous research on factors that appear to influence or predict homicide case outcomes. Next, we briefly describe the methods of data collection and analysis used in the present study. In the remainder of the article, we present our findings and discuss their implications.
Homicide Case Outcomes
In the United States, success in homicide investigations, as in all crime investigations, is normally measured in terms of “clearances.” It should be noted that clearing a crime does not require a conviction in court: Investigative success is based on identifying a clear suspect and gathering sufficient evidence to justify a charge of murder or manslaughter. An offense can be cleared in two ways. It can be “cleared by arrest” when at least one person has been arrested, charged, and turned over to the court for prosecution (Federal Bureau of Investigation [FBI], 2014). Alternatively, it can be “cleared by exception” when police have identified the offender but factors beyond their control (such as the death of a clearly identified offender, or the refusal of extradition) prevent them from laying charges. Clearance rates are computed by dividing the number of offenses cleared during a particular year by the total number of offenses recorded in that year: the former figure may therefore include clearances of offenses that were recorded in earlier years (hence the occasional appearance of clearance rates of over 100%).
Until recently, similar—though not identical—rules were used to construct “detection rates” (or “clear up rates”) of homicides in the United Kingdom. However, England and Wales 1 has recently replaced detection rates with statistics on “case outcomes,” which provide details on the current status (e.g., conviction, acquittal, proceedings discontinued, suspect died) of all homicides recorded in any given year. In other words, any outcome achieved, no matter when achieved, is assigned retrospectively to the year in which the relevant crime was recorded (Office for National Statistics [ONS], 2017). This means that unlike the “detection rate,” case outcomes for any one year may be subject to change for many years to come. 2
Even allowing for differences in definition or measurement, rates of success in homicide investigations can vary considerably between jurisdictions and over time. In the United States as a whole, the homicide clearance rate has declined dramatically from above 90% in the mid-1960s to just under 60% in 2016. 3 There are also major variations between individual police forces, with some closing all or most homicide cases and others closing only a small proportion of cases (Davies, 2007; Roberts, 2014). Detection rates are more stable in Britain, with a much smaller decline from an average of 94% in the 1960s to around 90% since 2000. 4
Accounting for Variation in Homicide Case Outcomes
Researchers have examined a wide range of factors thought to influence homicide case outcomes and hence to help explain differences in clearance rates across jurisdictions and periods. The sheer number of such factors is daunting. For instance, in their landmark study of homicide clearances in the United States, Wellford and Cronin (1999) accounted for the influence of 215 independent variables—an exercise not only complex, but also at risk of producing unstable estimates. 5 Some scholars have established parsimonious frameworks for categorizing these many potential explanatory variables. Innes (2003), for example, made a distinction between extrinsic and intrinsic factors, arguing that homicide investigations can suffer due to the “extrinsic complexity” of a particular case or “intrinsic difficulties” associated with the investigation itself. Alexander and Wellford (2017) identified three broad kinds of factors: “demographics,” “overall case characteristics,” and “police practices.” Maguire et al. (2010) distinguished between “environmental” factors (i.e., all of the factors that are external to the police organization, including incident characteristics, community characteristics, and the political environment) and “organisational” factors (e.g., police resources, policies and organizational arrangements, investigative practices, practitioner skills, and available technologies).
These frameworks highlight a broad distinction between sets of factors over which police appear to have little or no control (“extrinsic”/“environmental”/“demographic”/“case characteristics”) and others over which they can exercise some control (“intrinsic”/“organisational”/“police practices”). We argue later that such a distinction can be misleading, as there are complex interrelations between different factors and, in some cases, the police can influence environmental factors more than it might first appear. However, for the purpose of summarizing previous research results, we use a similar framework and terminology here, presenting separately findings relating to “environmental” and to “organisational” factors.
“Environmental” Factors
Maguire et al. (2010) divided what they call environmental factors into two main subcategories: incident/case characteristics and community characteristics. 6 By far, the largest body of research has examined the role of incident characteristics, with many studies concluding that they influence the likelihood of a case being cleared. To begin with, research from several nations has established the general point that many homicides are relatively easy to “solve” in that police do not need to engage in a complex and protracted search to identify the perpetrator. Based on research in England, Innes (2003) found that about half of the homicide cases in his sample could be described as “self-solvers” because police are presented early in the investigation (e.g., through an immediate admission by a perpetrator still at the scene) with sufficient evidence to identify the offender (p. 199). Similarly, based on research in the United States, Puckett and Lundman (2003) identified homicide cases that they refer to as “dunker clearances.” The existence of such a category is reflected in statistics on how quickly cases are solved. For example, Roberts and Lyons (2011) found that an average of 23% of homicides across the United States 7 were cleared by arrest on the same day the incident occurred (referred to as “quick clearances”) and Puckett and Lundman (2003) came up with a similar figure of 22%. Moreover, Wellford and Cronin (1999) found that just over half of their sample of homicide cases in the United States were “solved” within a week, while in Canada, Brodeur (2010) found that in 53% of cleared homicide cases, the person eventually prosecuted was identified immediately by police answering a 911 call. 8
At the other end of the spectrum are cases that pose major challenges for detectives—most often, where there appear to be no “leads” to identify a clear suspect, but sometimes also where there are major obstacles to building a case against a suspect once one has been identified. These are variously referred to as “whodunit,” “complex,” or “difficult-to-detect” cases (Innes, 2003; Puckett & Lundman, 2003). Beyond these generalized, commonsense observations, however, much remains to be understood about the impact of specific factors on the chances of cases being cleared.
As noted above, numerous researchers have explored statistical associations between clearance rates and large numbers of incident-related variables. Unfortunately, taken as a whole, the results of this research are sometimes confusing and contradictory. One of the most consistent findings has been that, after controlling for other variables, homicides involving firearms are less likely to be solved than those involving other weapons or no weapons (Alderden & Lavery, 2007; Litwin, 2004; Litwin & Xu, 2007; Marche, 1994; Mouzos & Muller, 2001; Puckett & Lundman, 2003; Regoeczi, Jarvis, & Riedel, 2008; Roberts, 2007), though even here, a few other studies report contrary results (Bänziger & Killias, 2014; Jiao, 2007; Riedel & Rinehart, 1996). A related finding is that homicides associated with other felonies (such as robbery) have lower clearance rates than those not associated with other felonies (Lee, 2005; Mouzos & Muller, 2001; Regoeczi, Jarvis & Riedel, 2008; Riedel & Rinehart, 1996; Wolfgang, 1958).
The location where the homicide occurred is also thought to be important. For instance, Keel (2008), Wellford and Cronin (1999), Bänziger and Killias (2014), and Jarvis, Mancik, and Regoeczi (2017) all reported that homicides occurring outdoors have lower clearance rates than those occurring indoors. A similarly consistent finding is that homicides in private dwellings or other private places are more likely to be cleared (Addington, 2006; Jiao, 2007; Litwin & Xu, 2007; Mouzos & Muller, 2001; Regoeczi et al., 2008; Roberts, 2007). However, others have found no significant association between location and detection (Marche, 1994; Puckett & Lundman, 2003; Riedel & Rinehart, 1996).
There are mixed findings regarding the characteristics of victims. 9 Perhaps the greatest level of agreement is found in relation to race and ethnicity, which have often been shown to be associated with clearance rates. For example, using data from Chicago, Litwin (2004) found that cases with Latino victims were less likely to be cleared by arrest than those with White victims. Roberts and Lyons (2011) found that homicides involving Hispanic victims were less likely to be solved than those with White or Black victims, while DeCarlo (2016) found that homicides of young African American males who lived in cities with high poverty were particularly unlikely to be solved. In Britain, homicides involving Black victims are less likely to be solved than those with White victims (Brookman, Jones, & Pike, 2017). Regarding other characteristics, Turner and Kosa (2003) found that immigrant, transient, homeless, or unidentified victims were overrepresented among cold cases in the United States (and hence likely to remain unsolved for many years). Bänziger and Killias (2014) found that Swiss homicide cases involving homosexual victims were less likely to be solved than those involving heterosexual victims. They noted a “strong nexus between homosexuality and the clearance rate” (p. 631). On the contrary, several other studies, notably the large multivariate analysis by Wellford and Cronin (1999), have found no significant association between victim characteristics and clearances (see also Addington, 2006; Puckett & Lundman, 2003; Regoeczi, Kennedy, & Silverman, 2000).
A related factor is victim lifestyle, which has also been found to influence the likelihood of detection. For example, Trussler (2010) found that homicide cases in Canada that involved gang and/or drug activity were less likely to be solved than those without these characteristics. Similar findings have emerged in studies in the United States (see, for example, Litwin, 2004; Turner & Kosa, 2003; Wellford & Cronin, 1999). In such cases, police may know the identity of the offender but fail to amass enough evidence to bring charges and clear the case, due to witness intimidation (Maguire et al., 2010). More generally, Rydberg and Pizarro’s (2014) analysis of homicides in Newark, New Jersey found that homicides with victims scoring high on their “deviant lifestyle scale” (determined by whether the victim was a gang member, drug dealer, involved in other illegal activities, or had a criminal history) took significantly longer to clear than those with victims who led nondeviant lifestyles. Similarly, Hawk and Dabney’s (2014) qualitative research in a U.S. homicide unit revealed that investigators were heavily influenced by perceptions of victim deservedness. Investigators prioritized “true victim” cases over those in which the homicide was viewed as resulting from the victim’s involvement in risky behaviors (such as selling drugs or being involved in gangs).
Finally, some researchers have found the victim–offender relationship to be one of the strongest predictors of homicide clearances, leading to the assertion that the more distant the social relationship between victim and offender, the less likely the case will be solved. Hence, “domestic” homicides typically have the highest clearance rates (see, for example, Lee, 2005; Marche, 1994; Roberts, 2007; Roberts & Lyons, 2011). 10
Beyond the research on incident or case characteristics, a smaller body of research has examined the influence of community characteristics on homicide case outcomes. For instance, Borg and Parker (2001) found that a number of community characteristics were associated with clearance rates, including residential segregation, level of education, and residential mobility. By contrast, Roberts (2014) found that jurisdiction-level poverty, residential instability, racial segregation, and population size had no effect on investigative outcomes. Similarly, Litwin (2004) found that, whereas home ownership rates had a significant positive effect on homicide clearance rates, several other community characteristics (including income, unemployment, education, population, and homicide rate) had no effects.
“Organisational” Factors
There is some evidence that clearance rates are affected by the nature and quality of police practices and procedures and other police-related factors (Maguire et al., 2010), but systematic research on the influence of these “organisational” factors is less common than research on environmental factors. Roberts (2014) found that the only organizational variable with a significant effect on homicide clearance rates was police workload. Agencies with larger workloads had lower clearance rates. This finding is consistent with an early analysis of the effects of police organizational factors upon clearance rates by Chaiken, Greenwood, and Petersilia (1977), who also found that larger caseloads were associated with lower clearance rates.
Keel, Jarvis, and Muirhead (2009) surveyed 55 police departments in the United States to test the effects of five sets of factors on murder clearance rates, including three police-related categories: management practices, investigative procedures, and analytical methods. They concluded that formal training of homicide detectives improved clearance rates, as did the use of analytical tools, such as blood splatter and statement analysis. Furthermore, they suggested that managerial oversight could marginally improve agency clearance rates through the mobilization of necessary resources, as could the implementation of witness protection programs to increase levels of cooperation with the police. This is one of the few studies that have highlighted the impact of investigative skills and experience. Earlier evidence on this issue is contradictory. Whereas Marche (1994) found detective experience to be positively related to homicide clearances, both Chaiken et al. (1977) and Puckett and Lundman (2003) reported, counterintuitively, that level of detective experience was unrelated to outcomes.
One of the most comprehensive studies of organizational factors is the abovementioned study by Wellford and Cronin (1999), who examined 798 homicides that occurred in four large U.S. cities between 1994 and 1995. They identified 51 factors that had a statistically significant relationship with the probability of closing a case, 37 of which were said to be to some extent within the control of the police. They highlighted as particularly important a swift response to the scene (i.e., within 30 minutes of the crime being reported) and comprehensive investigative practices.
Other relevant evidence comes from qualitative studies. For example, Carter and Carter (2016) identified seven law enforcement agencies in the United States with high homicide clearance rates (over 80%) and explored through local interviews possible reasons behind this achievement. They were also able to make broad comparisons with other jurisdictions with low rates. They concluded that investigative performance could be improved through developing strong community relations, collaboration with external agencies, high-quality training, and adequate resourcing. In England, Feist and Newiss (2001) examined police files and reports associated with six difficult-to-solve homicide cases. They found that the initial actions of the senior investigating officer (SIO 11 ) were critical in briefing, debriefing, and motivating other officers on the investigation. They also emphasized the importance of carefully evaluating the large volumes of information that can accumulate in the early days of an investigation. Relatedly, based on a qualitative study of homicide investigation in England, Innes (2003) observed that failure to prevent “systemic overload” of information could result in establishing too many lines of enquiry and pursuing the wrong leads, thus reducing productivity. Furthermore, Innes suggested that “compliance drift” (deviations from standard practice or regulation) can creep into the work of detectives during long-running cases, due to pressure to get results or, in some cases, repetition and boredom.
Despite these useful contributions, knowledge about the impact of police practices on clearance rates remains limited. This is particularly striking with regard to the emergence of sophisticated technologies useful for investigating homicide, including improvements in DNA profiling, ballistic imaging, mobile telephone tracking and analysis, and other forensic tools, whose impact on clearances is not yet well understood. It is also noteworthy that there is only limited evidence on the impact of investigator skills and experience, and even less on investigator effort or work ethic.
Overview
There remain large gaps in our understanding of what factors contribute significantly to whether homicide cases are solved. Moreover, existing studies are characterized by widely varying and sometimes contradictory findings, making it difficult to formulate general conclusions. Most of the quantitative research on this issue comes from the United States, where multivariate analysis has produced something of a “laundry list” of incident-related factors thought to influence homicide clearances. These tend both to underplay the influence of factors related to police practices and to leave unanswered broader questions about how and why the factors identified influence solvability. More nuanced kinds of evidence, including exploration of the impact of investigation-related factors, have gradually built up from a small number of small-scale qualitative studies, but synthesizing and drawing conclusions from this rich but diverse body of research is difficult. Importantly, too, no research to our knowledge has examined these issues comparatively across nations. Our aim in this article is to help to fill some of these gaps by reporting findings from our interviews and observations with detectives in both the United States and United Kingdom. Our interviews are useful for exploring detectives’ views, understandings, and experiences concerning factors that they believe affect the chances of homicide cases being solved. Our observations are useful for revealing ordinary working practices and behavioral patterns of detectives in both nations.
Data and Method
This article is based on qualitative data gathered at intervals through interviews and observational fieldwork over a period of 8 years in nine different police areas, five in Great Britain and four in the United States. 12 The selection of both areas and interviewees was based on convenience, rather than random sampling, so the findings cannot be claimed to be generalizable. Nevertheless, the use of qualitative data has enabled us to establish a much more textured understanding of the homicide investigation process than would have been possible with quantitative data alone.
The research in Britain was undertaken in police services of varying size and character, and with a considerable range of homicide rates and detection rates, located in England, Wales, and Scotland. The American agencies were less diverse: all were located on the East Coast, had homicide rates above the national average, and had all at times experienced problems particularly with gang-related murders. In total, 101 interviews were conducted with homicide investigators (57 in the United States and 44 in Britain), as well as over 700 hour of ethnographic shadowing and observation. 13 The latter included shadowing four specific homicide investigations in Britain for a period of 120 hour and five homicide investigations in the United States for a total of 190 hour. Extensive field notes were produced from these observations, including records of numerous informal conversations with detectives and senior officers.
Five of the investigators interviewed in the United States were “managers” (holding the rank of lieutenant, captain, or major) who oversaw the work of detectives. Two of those in Britain were recently retired senior detectives who had held both investigative and management roles. The remainder were detectives currently working on homicides at ground level. Most of the data from Britain were collected between January 2008 and June 2013. A small number of interviews were completed in 2015 and 2016. Data from the United States were collected between April 2012 and August 2013.
Interviews were conducted in private (usually in empty rooms in the criminal investigation or homicide departments). Eighty-seven of the 101 interviews were digitally recorded; for the remainder, detailed handwritten notes were typed up as soon as possible afterward. Interviews lasted an average of 70 minutes, were transcribed verbatim, and were analyzed thematically and comparatively using the qualitative software package NVivo, Version 11. Nodes and subnodes were created incrementally as each transcript was read and reread in depth. For example, detectives’ broad responses to questions about what makes homicide cases difficult to solve became a first-level node, within which 12 subnodes were created, each covering their comments about particular factors they identified (“witness issues,” “crime scene factors,” etc.).
Findings
In this section, we present views and evidence from our interviews and observations with detectives regarding the influence of both environmental and organizational factors on the outcomes of homicide investigations. We begin by summarizing detectives’ perceptions of the kinds of case characteristics that in their experience make some homicides “intrinsically” difficult to solve: to use Innes’ (2003) terminology, the “whodunits” as opposed to the “self-solvers.” We then explore their views about the kinds of organizational and individual practices and characteristics that are most likely to maximize the chances of clearing or detecting such cases.
Detectives’ Views on “Solvability”
When asked to identify the key factors that determine whether a homicide is relatively difficult or easy to solve, detectives in both the United States and Britain tended to focus on what Stelfox (2009) has called the “information profile” of the case. Difficult-to-solve cases were described as lacking physical evidence (such as DNA evidence, closed-circuit television [CCTV]/video evidence, fingerprint, ballistic or other trace evidence) and/or witness evidence. In contrast, cases where such evidence existed (or could be readily found) were viewed as likely to be solved, and solved relatively quickly. For example, a homicide detective from America explained: Physical evidence is still the most powerful in that if you get a print on a gun, you get their picture on a camera, DNA, DNA is absolutely awesome. When you have evidence with their DNA on it, the suspect I mean, it’s awfully hard for them to be able to articulate why it should be on it. (U.S. Homicide Detective)
Many reasons were given by detectives as to why particular homicides had limited information profiles. Among those most commonly cited in relation to a lack of physical evidence were forensically aware offenders, delays in finding the victim’s body (evidence disappears or degrades), victim lifestyle, and the random element of some homicides. Examples of each of these are as follows: With DNA technology moving and progressing people are just circumventing it by wearing protection, by wearing gloves, especially these premeditated crimes, usually the most heinous offenders premeditate their acts before they come out and do them . . . and that’s the difficulty that you run into is that people more and more and more are learning what they need to do to circumvent the DNA evidence, to not leave fingerprints behind and essentially to clean bodies, clean the victims’ bodies, how to dispose of victims’ bodies. (U.S. Homicide Detective) Decomposed bodies are very, very difficult. You find a body in the woods, which I’ve had a few of those over the years, that body could have been there for years . . . complicated because you’re working on basically historical information and trying to put that person there, by that time they’ve changed cell phones five times. Those are very, very complicated but I’m not saying they’re impossible. (U.S. Homicide Detective) She was at a nightclub and she was the designated driver, and she walked outside at the same time these guys were just shooting in front of the club as they drove by, and she got hit in the head. So she has nothing to do with them, and didn’t know who they are, didn’t know anything about them. (U.S. Homicide Detective) This is one where we haven’t got a clue who’s done it; and this is the murder of a gay man in a flat who is believed to be a predatory paedophile, who has got a history of sex offending. Well on the night in question somebody put . . . bins up against the front door of his flat and set them alight . . . set his flat alight and killed him. We haven’t got a clue who has done that, not a clue . . . You have got victims going back years. He has been having gay relationships with young boys on the estate, and it’s a rough estate. He has had young boys back and forth there, having fags [cigarettes] off him, and he is trying to groom them. He’s been loaning money to people, who owe him money. Like, it’s a nightmare . . . We’re still at the stage, even now, this is 3 years old now, still at the stage of there are a number of theories that we just can’t firm up, we can’t discount and we can’t firm up on. (British Homicide Detective)
A further frequently mentioned factor, particularly by investigators in the United States, was the reluctance of witnesses, or others who have valuable information, to speak to the police: At least around here, we have a lot of people that are not cooperative with the police, very much so, I mean, almost to a hundred per cent. (U.S. Homicide Detective) The toughest part is just to get people to tell us what they know. There are people out here all the time that have information on who did what, and it’s hard to fathom why they just hold on to that information. I understand being afraid but when you have an opportunity to make a phone call anonymously to Crime Solvers to share your information without being identified, I just don’t know why people don’t take advantage of it, especially when there’s a reward being offered for the information. (U.S. Homicide Detective) Well, one of the biggest challenges that I have seen over the years . . . is getting the cooperation of the community. Everybody has got this thing now where you don’t snitch . . . So, getting that, and we have tip lines, you remain confidential, and just they don’t really utilize them as I think they should, and I think some of the issue is trust maybe, people in the community don’t necessarily trust the police department as a whole, you know what I’m saying? (U.S. Homicide Detective)
Although they saw it as less widespread a problem, British detectives recognized that some communities in Britain had a strong distrust of the police: The Somalian community are a very, very tight knit community and have, because of where they come from, because of what happens in their own country, have a huge distrust of the police. And then you’ve got the added dimension of the females won’t necessarily talk to the police full stop, and then if they do talk to the police they probably would prefer to speak to a female, which the males don’t like. So you’ve got all of those dynamics. The reason I mention it is because Mrs X, a Somalian lady, lived opposite [the victim’s flat], and she heard loads of stuff going on and she didn’t call the police but she saw somebody fleetingly in the hallway around about the time that we were interested in. So she was an absolute key witness for us. She had three kids all under 3, she spoke really poor English, she had a distrust of the police and didn’t want to get involved. (British Homicide Detective)
As well as distrusting the police, many detectives recognized that members of some communities—typically in multiethnic, socially and economically deprived inner-city areas—were fearful of reprisals from criminal gangs if they gave information to the police. By contrast, “good” witnesses were said to come largely from wealthier areas. For example: A homicide within an area whereby the witnesses are good. And what I mean by that, is the fact is that you’re in a leafy suburb where you’ve got reliable, educated people willing to come forward and give evidence, because they have no fear of giving that evidence. As opposed to a housing estate
14
where lots of people may have seen it, but no one wants to really say anything because they’re worried about comeback. That’s the first thing. (British Homicide Detective, SIO)
It was also recognized that in some areas there was the very real potential for witnesses to be intimidated or even killed to silence them. Again, witness intimidation appeared, from our research, to pose a greater problem in the United States, than in Britain, partly because detectives seemed to rely more upon witness evidence, but primarily because of the more widespread violent gang culture: There’s no incentive to be a witness on a homicide case. There’s no incentive for you to come to court and say, yes I saw X because then, all of a sudden, now you’re worried about somebody shooting you or killing you. You have no attachment to me so you couldn’t care less what—“hey, he got killed, I’m sorry but I’m not coming to court and telling anybody what I saw.” So that’s our biggest challenge. (U.S. Homicide Detective)
The Impact of Police Organizational Factors
The above comments by detectives echo—and give more texture to—some of the key findings from previous research on factors that affect homicide clearances, especially those relating to victim, incident, and community characteristics. What many of the detectives’ accounts had in common was that solvability factors were described essentially as “givens”—in other words, as static factors that could not be changed—suggesting a somewhat deterministic view about the prospects of clearances. However, some detectives believed—and had evidence from their own experience—that effective responses by the police and other bodies can help “beat the odds” and make a significant difference to the outcomes even of the most challenging cases. Our observations of investigations also gave support to this view. We explore below their accounts of how this can be achieved, presented under three main headings: investigator qualities, availability and effective use of resources, and police culture and attitudes.
Investigator qualities
A number of characteristics of investigators were identified as likely to make a difference to outcomes. Particular emphasis was placed on a robust work ethic, resilience, and attention to detail. Relevant experience, knowledge, and skills were also highlighted as important facets of an effective investigator.
The “Columbo-style” work ethic
Most American and many British detectives tended to place a Columbo-style work ethic of meticulous attention to detail and a determination to follow every lead as the defining characteristic of the most effective detectives that they had worked with.
The best SIOs and IOs are real detail people . . . the more boring, sometimes, the better they are. It’s not about . . . being flamboyant and off-the-wall and ideas coming out left, right and centre, it’s about detail. (British Homicide Detective) Well, commitment and dedication . . . resilience, attention to detail. When I say resilience I don’t mean just initially when it happens, but I mean long-term when . . . you’ve got to keep looking through the statements, keep looking for . . . you know, read every line of every statement, every document. (British Homicide Detective)
A specific example of individual qualities of tenacity and resilience clearly making a difference to a case outcome was provided by a British female SIO we interviewed. She had recently detected a cold case homicide that had occurred some 30 years earlier. She had initially been introduced to the case as a review officer, but later took on the cold case when working in Homicide. She campaigned for, and secured, finances to run DNA searches of the National DNA Database and, ultimately, to fund a mass DNA screening exercise that ultimately led to the identification and conviction of the killer. As she put it, “Drive, determination and tenacity, that is the only thing that solved this.”
The reverse side of the coin, identified particularly by detectives in the United States, was the negative impact of officers who were “lazy” or who failed to follow proper procedure: You take that case and you’re full force on it and it’s balls to the walls until you try and close it or until you’ve come to a dead end and you move to the next case. So I had partners that were lazy, I’ve had partners that had the same work ethic as I do, wanted to work overtime all the time. (U.S. Homicide Detective) Not all detectives are the best. There’s various motivations why someone might become a detective, and not all of them good. The majority of detectives are good, they like doing the job. . . . But some people are, “oh, I think I’ll be a detective that sounds quite good.” But not really bringing the detective ethos . . . [that] sense of, of utter pride to complete [the work] . . . you never want anyone to say to you that you could have done more. (U.S. Homicide Detective)
In short, though all detectives recognized that some homicides were indeed much more difficult than others to solve, they also acknowledged that whereas some officers would resign themselves to the fact that they had a “whodunit” that was unlikely to close, others would not adopt a fatalistic attitude and would investigate the most challenging cases with no less rigor and energy than any other: The good detective takes that bad luck and makes it into something; makes that hard case into something, that’s what a good detective is. An average detective would say oh well, there’s a body in the woods, I am not going to be able to close that; a good detective is “I am going to run down everything. And I may not be able to close this, but I am going to run everything down.” (U.S. Homicide Detective)
Experience, knowledge, and skills
Most of our interviewees recognized, and we bore witness to, the fact that detectives with greater levels of experience tended to draw upon this to help them progress challenging investigations: Detective work is so much based on experience. It’s actually key and essential. So no matter how . . . intelligent you are, how articulate you are, you still need that experience. (British Regional Review Officer and Former Homicide Detective)
While shadowing a live investigation in America where the suspect was still at large, the issue of inexperience and its potential impacts upon an investigation became apparent. In this instance, the lead detective who was relatively new to the Squad was struggling to identify the suspects in a fatal shooting. She had exhausted most standard lines of enquiry, including canvassing for witnesses, and generated few solid leads. While she was trying to figure out how to move the case forward, a more experienced detective suggested that she arrange a “cell site dump.” This enabled her to progress the investigation by triangulating suspects’ locations at the time of the shooting, hence narrowing down the pool of potential suspects.
Another aspect of detective work that was widely identified by American respondents as important to the success of investigations was interviewing skills, with a particular emphasis on the craft of interviewing suspects to secure confessions. By contrast, British detectives rarely discussed this. This is partly because, in Britain, interviewing is usually undertaken by specialist interviewers who are not necessarily homicide detectives, and partly because the rules governing interviews are much stricter—for example, a lawyer is almost always present, compared with the situation in many U.S. states where the majority of suspects are interviewed without legal representation and, as we observed, who often waive their Miranda rights. 15
In addition to the skills and experience of detectives, many interviewees emphasized the importance of knowledge, skills, and experience among uniform or patrol officers who are generally first on the scene of homicides. The speed and quality of the initial response was recognized by detectives in both Britain and the United States as particularly important in terms of solvability potential. It was universally acknowledged that during the early “golden hours,” or “first 48,” certain forms of evidence were available that may subsequently prove difficult or impossible to locate and retrieve. This included physical evidence, such as weapons, CCTV footage, forensic evidence (that often degrades with time), and witnesses. Uniformed officers will generally be the first to attend the scene of a homicide and several examples of problematic initial responses were highlighted by detectives. These included failing to gather basic personal details from witnesses/potential suspects, allowing witnesses/suspects to leave the scene, and compromising or contaminating the crime scene.
In some cases, a poor or delayed initial response was the result of the incident not initially presenting as a homicide, or even a crime, as the following British case illustrates: But initially you see, they phoned up, they said there was a drunk . . . a bloke with a head injury, possibly fallen, drunk. It was only when they found, at the hospital, the screwdriver has been stabbed into the side of his head, they thought, “oh my God, he’s been assaulted.” Then it ramped up but of course, by this time, the scene had been left open, the body had been removed . . . well, he was still alive at that point . . . (British Homicide Detective)
In both the United States and Britain, there were a number of examples of “delayed deaths,” even after very violent acts, such as shooting incidents. While homicide detectives were often given the “heads up” that there had been a shooting that might result in a death, they would not be dispatched to the scene or assigned the case, until it was clear that it was a homicide. In such cases, it became critically important that the first officers on the scene handled it appropriately: [He was] shot. He eventually died probably 6 hour later. So in those 6 hour it wasn’t my case. If they’re not dead, it’s going to be handled by the station level detective, whatever area that is in and that was in the actual area. Once he died it was my case . . . So we had to depend on what those station level detectives do . . . they did pretty good work, to be honest with you. (U.S. Homicide Detective)
At the same time, it was recognized that contamination of crime scenes was not always the fault of the officers initially responding. In some, it was recognized that SIOs or lead detectives had not carefully managed homicide scenes, as the following two extracts from both nations illustrate: So the house was searched eight times by the search teams, the police search teams, as a murder scene, and they absolutely . . . during the search, they trashed the scene. So the scene was unrecognisable from what it originally looked like, which really caused me problems in the end. I didn’t realize it was happening; perhaps I should have gone there and looked at it. (British Homicide Detective) The crime scene was horrible, just terrible job on our part. There was definitely blood on the scene that was never collected. It was photographed but never collected. They had a good suspect. Actually, they believed the son was involved in killing the mother. Could never prove it . . . that was one of those things you’re just like, my God, this is bad. I mean you’ve got things that we’ll forever never know. That could’ve been, that drop of blood right there could’ve been the suspect and we’d have DNA to show that that was the son, and done. (U.S. Homicide Detective)
In summary, despite some differences in emphasis between the two nations, overall there was clear consensus that certain individual qualities of investigators were important factors in whether homicides were cleared: In particular, tenacity, resilience, and attention to detail in “working a case” could mean the difference between unearthing a critical piece of information and leaving it undiscovered. Equally, individual detective skills, knowledge, or experience could play an important role in progressing challenging cases. Both these views are supported by concrete examples we have briefly described. It was also widely acknowledged that the knowledge, skills, and experience of the officers who are first on the scene of a homicide can make a critical difference to the likelihood of closing cases.
Availability and effective use of resources
Staffing and time
Having enough detectives dedicated to the case for sufficient time was viewed as critical to the successful outcome of investigations by detectives in both countries—especially on “whodunit” cases. Our research was conducted during periods of budget cuts and staff reductions in police forces across both Britain and America. It is perhaps unsurprising, therefore, that we uncovered numerous examples of detectives feeling under pressure to “do more with less.” In Britain, respondents told us that putting together a sufficiently large team was sometimes difficult, which had detrimental effects on some investigations. Not all forces retain dedicated homicide squads.
16
Instead, major crime teams investigate homicide and other serious crimes and it is generally the case, within this structure, that detectives are abstracted temporarily from criminal investigation departments (CID) to assist with the investigation. This creates an added layer of complexity to the resourcing of murder investigations, as there tends to be a power struggle for resources between divisional commanders of CID—who are mindful of the need for enough officers to investigate other serious but nonfatal crimes—and homicide SIOs who, equally, need to secure sufficient officers from division to bolster their teams. As a retired British detective explained, this battle for resources has implications for both the numbers and the quality of officers sent to homicide from division: I have run jobs where I have got to ring the division [and say] “I want 12 DCs off you now.” Divisional commander will say, “I can only give you six.” You say, “I want 12.” You know, you are trying to wrangle over staff. And then, when they do send them to you, they don’t send you the best . . . and they tend to send you the same ones they don’t want all the time. And you say, “no, I want good, I want experienced people.” (British Detective Trainer and former Detective)
Such problems are exacerbated in cases that remain unsolved for a long period in which pressure mounts to reduce the numbers of officers working on them.
The picture in the United States was somewhat different, in that the challenges of resources related more to the sheer volume of homicides that detectives had to contend with. So while detectives felt that they generally had enough officers with sufficient experience on a squad, they did not always have time, as a squad, to investigate the case before another one came in. This was a major issue for them, as having the time to “work the case” was seen as critically important in terms of being able to solve it: There is a correlation between the closure rate and the time that you have to investigate that case; hands down there is a direct correlation to that, that the more time you have, the more cases you are going to close. (U.S. Homicide Detective) But the problem is when another one comes in, this one will go cold. It’ll sit for a bit and I’ll keep working on it when I can, but when a fresh murder comes in there’s certain things you have to do when a murder happens, that you have to do within the first few days, or else you’ll never get it back. (U.S. Homicide Detective)
Clearly, the longer the investigation goes on, the more difficult this becomes, as new homicides require detectives to take on new cases. One American detective lamented: My main goal from now until my last day in homicide is to close my very first case. It’s almost symbolic, I hate the fact that it’s open. And a lot of these cases we work, we work, we work, we work, but on my case with AR, the whole squad worked it for a couple of days and then everybody got pulled off of it, and it was just me and my partner, and everybody else had to work AS, the 17-year-old girl who was killed last year in her house.
Because detectives were pulled off to work the new case, only two detectives remained working on the original case. This staffing shortage meant that they had to abandon their ordinary way of working the case. As noted by the detective: So it was unique because we don’t work like that; we crush it, we crush the case hard until we can’t either walk anymore because we can’t keep our eyes open, or we run out of leads . . . but with that one it was you got the squad for 3 days but then they have to get back on AS. AS is a unique case; just with everything that’s been put into it. (U.S. Homicide Detective)
Effective use of science and technology
It is now well established that police work in general has undergone a process of “scientification” (Ericson & Shearing, 1986), or “technification” (Nogala, 1995) and homicide investigations are no exception. Nearly all our interviewees pointed to the importance of science and technology in helping to solve murder. For many, scientific evidence brought with it a certainty that other evidence lacked, and this in turn helped them to close cases: If we have it, it’s huge, it just depends. I would probably say at least over half the cases, once you have that . . . So I have a case that’s waiting trial where DNA was huge; just proving that it’s your person. And once you have that then it’s case closed; you can’t argue that, it’s very hard to, at least. (U.S. Homicide Detective) Really it’s about identifying who that sample or stain belongs to, without, you know, any doubt at all. Um, like a billion to one. Twenty, 30, 40 years ago [we were] relying on evidence of human beings seeing things, witness testimony, which is, you know, pretty unsafe really . . . But scientifically with DNA it’s 100% and no-one can argue with it. (British Homicide Detective)
On the contrary, some detectives also emphasized that the advances made in science and technology in recent years do not automatically guarantee improvements in investigative outcomes. What is necessary is the effective use of such resources. As was pointed out, their effectiveness could be compromised in a number of ways, in some cases even to the extent that they could be a handicap, rather than an asset, to investigations. For example, while the computerization of data handling in homicide investigations in Britain was said to have brought enormous benefits, allowing the storage and analysis of large amounts of information, detectives also gave examples of cases that were difficult to solve due to the problem of being swamped by “too much” information. Innes (2003) observed how some homicide investigations result in “information overload” causing information management systems to become saturated and overloaded resulting in “systemic delays” in processing the information (p. 255). This is a problem that has been exacerbated in recent years by the proliferation of digital information (e.g., hundreds of hours of CCTV to interrogate, as well as mobile telephone and computer data to trawl through) as well as by new developments in forensic science: I came to CID in 1988 and there were no computers, no mobile phones, no CCTV, no “Facebook,” no DNA, which meant if you have a murder enquiry there wasn’t that much too look at. Now, all of those are great but all of those are like the world’s biggest haystacks which has got that tiny gem in there, somewhere, and you are expected to look through every single haystack with a fine toothcomb to find that gem, it’s in there somewhere. But I think that the volume of work in most Cat B
17
murders is just huge, huge. And to look at it properly you need more resources and more skilled people. (British Homicide Detective) I think it’s a major distraction. A lot of this media and the opportunity to get phone records because detectives get locked in on, “Well I had to get the cell phone records. I had to get all the chat records and then I need to review it all,” versus like you were talking about canvassing and talking to people, you’ve got to be talking to people . . . people’s memories are fading. This is where I think investigations are shifting away from . . . they’re getting caught up in too much of the information overload of technology and locking themselves down with all these phone records and feeling that this is going to solve their case. (U.S. Homicide Detective)
More generally, interviewees in both countries felt that advances in technology were leading in some cases to “over-reliance on science” at the expense of traditional investigative practice, leading to some deskilling of detectives. For example, some spoke of colleagues failing to “cover the basics” (such as house-to-house or interviewing suspects quickly) on the assumption that science was going to solve the murder: People can shortcut things because they rely, or tend to rely, too heavily on forensics. And sometimes it ain’t going to be there, you’re not going to have it, so now what do you do? And you can’t overlook the basics in a homicide investigation. You still need to cover the basics—you need to be out there, you need to get your interviews done, you need to do a really good crime scene inspection and documentation. (U.S. Homicide Detective) I think there’s a tendency for SIOs to assume from the outset that they’re going to get some great result from either DNA or from passive data generators, and when they don’t they haven’t actually done the basics that they should have done. And so rather than see that more as a bonus and the icing on the cake, that is the cake, and so our SIOs now I think are not carrying out some of the basic enquiries which they should do because they’re just assuming or clinging to the hope that they’re going to get some golden nugget from these new technologies. (British Homicide Manager–Superintendent)
Some managers suggested that a culture of “decision inertia” was creeping into the practices of certain detectives—a tendency to put off arresting suspects while waiting for more evidence from science or technology to emerge: For example, if a name pops into your incident database quite early because a witness has nominated them, when you didn’t have all this stuff to hope for—DNA, ANPR, CCTV—there wouldn’t be any discussion about it. They look good, they’re a suspect, you’d go and arrest them and you would then do it quickly because you know that a priority would be to recover any forensic evidence that they had on their clothing or in their home. Now there’s an excuse to put that decision off because, “Well we won’t arrest today, let’s get the results back from the phone work. Let’s get the ANPR, let’s get the cell dump for the mobile phones back.” And that’s great unless you don’t get any of that back and then you’re 10 days behind arresting that person than you would have been a few years ago. There’s an excuse now to put off some of these things. (British Homicide Manager–Superintendent) It’s great, the technology that we have for cell phones, seeing what people are posting on their Facebook, to be able to blow an alibi out of the water . . . but I think also it can hamper you a little bit, due to the fact that you spend so much time in the office focusing on that, rather than out in the streets. Because streets know who killed the person, the streets are going to know who killed the person before any of us know. (U.S. Homicide Detective) You need to build rapport to gain information. At the end of the day what solves a case is getting out there and talking to people and getting dirty, not sitting behind a desk and waiting for forensics. (U.S. Homicide Detective)
Finally, a number of detectives felt that they were not able to keep up effectively with the latest advances in science and technology as they had not been trained recently: Training is talked about a lot but we’re not sent to training a lot for various reasons, so you kind of have to do it on your own or through word of mouth, you hear what’s worked in other cases. (U.S. Homicide Detective) We are lagging behind and slowly, slowly getting behind the foremost technology. (British Homicide Detective)
There were many occasions throughout the fieldwork when we observed detectives struggling to understand or use various forms of technology that were relatively new to their investigative toolkit. We also observed that there was much uncertainty in some Homicide Units as to what officers were “legally” permitted to access. For example, on one occasion, a detective and district attorney debated whether and how to access a suspect’s Facebook account and were in disagreement as to whether the information could be used as “evidence.” What became clear was that technology was moving at such a pace that legislation (and/or understanding of new legislation) had not caught up. Equally, we observed that in all homicide squads/units detectives would point to particular officers who were known to be able to handle the latest technology (the “go-to” detective for live cell phone tracking, for example) and others who were “old school” and less able or willing to engage such investigative opportunities: I am old school; I don’t have Facebook and I wouldn’t even know how to do it. (U.S. Homicide Detective) There is so much more technology that is probably out there and we are probably missing many things that would probably help us. But that takes money, and it’s just not going to happen. (U.S. Homicide Detective)
Police culture and attitudes
The organization-related factors referred to in the previous two sections all affect the quality—and hence potentially the outcomes—of homicide investigations. However, it was also quite widely conceded, particularly in the United States, that different kinds of cases tend to receive differing levels of investigative effort. The two main influences on this were said to be external pressure on investigators (often manifested in pressure from senior managers) and police beliefs about and attitudes toward different types of people and communities. As our findings below reveal, the two are closely linked.
Low and high profile cases: “True” victims or “any old Schmo”?
Several detectives stated that the public and media “profile” of a case had a major influence on how senior police managers responded, both in terms of the resources they allocated to it and of the amount of pressure they put on investigators to “get a result.” In most cases, the profile of a homicide was governed primarily by the characteristics of the victim. This was put in very bald terms by an American detective: Detective: Obviously if there is a high profile case then that still gets attention, but any old Schmo murder just doesn’t get solved. Interviewer: Any old Schmo. What does that mean? Detective: Well just any low-level person, no one important, not high profile, just an average person. An average drug dealer; no one cares about him or her; it’s a number, but it’s not like the school kid that gets robbed and killed on his way to school, that’s the more important. (U.S. Homicide Detective)
British detectives spoke less openly about these distinctions, though there were hints that some cases received an exceptional level of response for “political” reasons. For example, in one British case involving the murder of a foreign student, the SIO remarked: Politically there was no way we couldn’t give that the Rolls-Royce service. But I’d like to think that we give the Rolls-Royce service anyway.
While “high profile” status had the welcome effect of ensuring the investigation was adequately staffed and resourced, several detectives from both countries stated that it could also create immense pressure on the investigation team. For example, in the above case, the SIO said that he had experienced “lots of internal pressure” from a very inquisitive and probing senior manager, which had made him feel “pretty angry” as the “pressure was just unrelenting.” Indeed, it was even argued by some that such pressure, far from increasing the chances of a timely successful outcome, could have a negative effect on progress: What I see here is command staff involvement where the detectives are allowed to go about their duties day to day but as soon as something starts to get pressure such as media, community pressure, it influences from the top down and it starts to fragment the unit because now no longer are we working on our capabilities . . . The commanders were giving briefs every hour on the hour and they wanted information from us, and no longer were we working on our information; we’re trying to satisfy a need for their information and their needs, and it takes away from, especially the lead people, the primary detectives that are now being distracted. I think that’s a big problem especially the bigger the case. (U.S. Homicide Detective)
The focus of most detectives when talking about the “profile” of cases tended to be on the responses of their bosses and the media, and the consequences of these in terms of internal pressure and the allocation of resources. However, several detectives also discussed the attitudes of frontline investigators themselves toward different categories of homicide and, in particular, the impact of their perceptions of the victim, or of the community from which he or she came, on how they approached the case—and hence potentially on the chances of clearance. This is exemplified in the following comments: Gang-related, drug-related homicides. Where witnesses . . . you’ve got gang members or thugs, gangsters, we call them gangsters, drug dealers, we’ve got those people fighting, shooting each other, whatever, and one’s dead and one’s not. There’s not a public outcry for the killer to be brought to justice. The community isn’t outraged, the detectives working the case, they work the case as hard as they can but there’s nothing will get your attention like in a truly innocent, like a child murder or something. Where your heart’s now into it, not just your head. (U.S. Homicide Unit Captain) Well for one she is a complete true victim; not that everybody isn’t a victim but she was an honest victim. My guy from my last murder is a victim of a murder, yeah, but he is not a “victim” victim, because 3 weeks prior he is running into somebody’s house with guns and robbing. This girl was just getting ready for school the next day. (U.S. Homicide Detective)
What these comments illustrate is that the level of time, resources, and effort put into particular investigations can be influenced both by detectives’ personal feelings about the case, as well as by their managers’ expectations and those of the wider community. The suggestion is that neither the police nor the wider community invest quite as much commitment and concern when one gang member kills another as when someone perceived as a “true” victim is killed.
Police perceptions of communities
As well as the characteristics of the victim, it was clear from our interviews that the perceptions that detectives (and indeed police officers in general) held of the community from which he or she came could also influence the conduct (and potentially the outcome) of investigations. This was particularly the case in the United States, where many detectives clearly expected that getting witness cooperation in certain communities—typically, deprived inner-city, multiethnic areas—would be a difficult process. In some such cases, they responded by simply “going through the motions” when canvassing local residents (almost resigned to not receiving any assistance) or adopting enforcement tactics such as “rounding up” known local criminals to find out what they might know about the homicide. The latter could further alienate citizens, while the former could reinforce a shared view (by citizens and police) that no one would talk, as the following field note and subsequent detective insight, illustrate: We left and knocked on all other doors in the block. One other witness (who lived opposite the 911 caller who the detective had just interviewed), a young Black male, said that he had heard the shot and looked out but didn’t see anything. The detective did not push him for more information but just took his name to make sure he wasn’t disturbed again (or so he said). Other people either didn’t answer or hadn’t heard anything—for example one man said that he had been sleeping (the detective didn’t believe him). I was struck by the fact that the detective seemed to expect and accept this lack of assistance and made no real effort to try to encourage or persuade potential witnesses to help. (Extract from Field Notes Wednesday November 7, 2012, United States) No one is going to say that officially, but that’s the nature. When I came to homicide my sergeant told me at the time, you will know the cases to work, you will know the ones to spend that time on, and the ones not to. So that case needs to be worked, and needs to be solved, that is a tough ass case to solve, because those people think that guy is a hero; no one is going to tell you anything. And so that’s why there is not a hundred per cent closure rate, because a lot of this is just simply people, and if they don’t want to talk to you they are not going to. (U.S. Homicide Detective, Sergeant)
Such beliefs can lead to the (sometimes unconscious) assumption that it is not worth putting a great deal of effort into canvassing for witnesses in certain areas, as the effort will rarely be rewarded. This is despite the fact that detectives were well aware that the key witnesses to some of the most challenging homicides were members of such communities. As one American senior officer explained: Witnesses are often problematic because they come from the same communities as the bad guys but that’s why they know things. [The Courts] want a carload of nuns to be witnesses. (U.S. Homicide Major)
It should be noted that there can be other reasons for not canvassing systematically for witnesses in “no-snitch” communities, principally concern for the safety of residents. To illustrate, during a 3-week fieldwork period in the United States shadowing homicide detectives, the lead author was made aware of two ongoing investigations where witnesses had been killed, either by the suspect themselves or by others acting on their behalf. Intimidation of witnesses from jail was not unusual, with suspects persuading allied gang members, friends, and family members—in one case the suspect’s mother—to threaten witnesses.
In short, there was little doubt that some homicides, due to the status of the victim, and/or of public reactions and those of senior officers, not only attracted greater investment of resources, but also generated in detectives a greater sense of commitment and urgency. This led them, by their own admission, to invest more time and energy into solving these cases. Indeed, as the comment above from a detective sergeant illustrates, this seemed to be part of the culture passed down from one generation of detectives to the next in some of the U.S. homicide squads that we studied.
Discussion and Conclusion
Understanding the factors that can influence homicide clearances is important for many reasons. Homicide clearance is often viewed as a key barometer of police effectiveness (Brookman & Innes, 2013). Unsolved homicides leave potentially dangerous offenders at large, which, when coupled with declining clearance rates, can lead to public unease (Witt & Trussler, 2014). Moreover, the suffering endured by relatives of homicide victims may intensify when the identity of the offender is unknown, denying them the sense of “closure” that can come from a successful prosecution (Riedel & Rinehart, 1996; Rock, 1998).
Our qualitative research not only echoes some of the findings of previous (largely quantitative) studies in this area, but has also added some new perspectives. Perhaps most importantly, our research offers new insights into the nexus between detectives’ perceptions of particular kinds of homicides, offenders, victims, witnesses, and communities; their expectations about the solvability of cases and, in turn, the commitment of energy and resources that they and their departments invest into certain kinds of cases. In addition, as our research covered investigations in both the United Kingdom and United States, we have begun to identify apparent similarities and differences between the two nations in these respects.
In attempting to identify factors that influence the likelihood of solving homicide cases, much of the previous research has used multivariate analysis to test the predictive strength of numerous independent variables. While there are important exceptions, most of these studies focus on what we have called “environmental” factors—principally, characteristics of the victim, the incident, and the area or community within which a homicide occurs. The effects of such factors are generally viewed as beyond the control of the police. This theme is echoed in some of the comments we heard from homicide detectives, particularly in the United States. For example, it was not uncommon to hear detectives poking fun at colleagues who had “caught a difficult case” with remarks such as “good luck with that one.” Although jocular, such remarks may reflect a shared pessimistic view that cases of particular kinds are by their nature almost bound to remain unsolved. These included cases lacking in readily accessible information profiles and, in particular, with an apparent lack of willing witnesses. This in turn reflected negative views about the nature of particular neighborhoods or particular groups of people, which were considered to be either too “anti-police” or too afraid of reprisals from local gangs to cooperate with investigations. The overall result could be almost a self-fulfilling prophecy, where low expectations led to investigations that simply “went through the motions” and were ultimately unsuccessful.
While by no means minimizing the scale of the challenges faced in “difficult-to-solve” homicides, we have identified a number of ways in which our evidence suggests that the mind-sets, policies, and actions of police departments, individual officers, or other key players can affect (positively and negatively) the chances of clearing homicide. Clearly, the methodology we have used does not allow us to demonstrate statistical associations between such factors and variance in clearance rates. However, our findings are based on in-depth interviews designed to distill the combined experience of large numbers of detectives, as well as observations of them undertaking investigations. It was notable that many of those who initially expressed pessimistic views about the “solvability” of particular kinds of homicide, eventually produced examples of cases in which the quality of the response from police or other agencies had led to successful outcomes, apparently against the odds. The key “organisational” factors they identified can be divided into three main categories.
First, it was widely agreed that the quality of homicide detectives can play a central role in whether homicides are investigated effectively. These qualities included a robust work ethic alongside relevant experience, knowledge, and skills. It was further emphasized that the quality of the work of patrol or uniformed officers—who often are the first to attend the scene of a homicide—is also potentially important to the outcome of an investigation.
Second, the availability and effective use of resources is important, especially on “whodunit” or long-running enquiries. Budget cuts across both countries during the period of the research had led to reductions in detective numbers and in specialist training (e.g., in new and emerging scientific or technological techniques). In the United States, the sheer volume of homicides, coupled with the cutbacks, meant that some cases that were potentially solvable were left to go “cold” as detectives were compelled to move on to new cases. In Britain, while the volume of homicide is considerably lower, our findings revealed that difficulties were sometimes encountered in assembling an appropriate team of officers to run the investigation effectively (particularly in areas where there were not dedicated homicide units) and that there were often difficulties in retaining detectives on a case for long enough.
Making best use of science and technology was also highlighted as a significant factor. Despite major advances in science and technology, it was apparent that there were frequently challenges in using it to best effect. These included (digital) information overload, excessive time demands on detectives to make sense of information, systemic delays in processing intelligence or evidence generated from new technologies, and, according to more established detectives, an overreliance on science and technology by some junior detectives at the expense of traditional craft skills. Some detectives also felt they had not been properly trained to keep pace with advances in science and technology.
Third, it was widely agreed that the culture and attitudes of detectives (and other officers) could have an impact on case outcomes. In particular, shared opinions about particular kinds of victims, suspects, and communities could influence not only levels of investigative commitment to particular cases, but also how detectives interacted with members of the community in question. Perhaps the most significant aspect of this is recognition of the importance to homicide investigations of good police–community relations — a theme familiar from broader debates about the vital part played by legitimacy in police effectiveness (Tyler, 2004; Tyler & Huo, 2002). It is not unlikely that a general decline in police legitimacy in the eyes of certain communities has contributed to the downward trend in clearance rates noted at the beginning of this article.
Appreciating the importance of these “organisational” factors opens the door to more informed discussion of potential strategies to improve homicide investigation outcomes. Of course, it would be naive to underestimate the obstacles to achieving these in practice, at least in the short term. Some kinds of organizational change, especially in culture and attitudes, are notoriously difficult and slow to bring about. Nevertheless, some quite simple strategies, if implemented in a determined fashion, can produce positive change relatively quickly. For example, it is possible to enhance detective training, to reward the most committed detectives, or to pair up detectives with different skill sets and work ethic. It is also possible to set up schemes in which different police departments and homicide squads can learn from each other in a systematic way. In the longer term, however, our research suggests that a key focus should be on police attitudes to and relations with particular sections of the public. At present, as we have documented, the low level of cooperation apparent in certain kinds of areas or from certain kinds of social groups, tends to be treated by detectives as a “given,” largely beyond their control. The data from our interviews and observations, together with insights from the broader policing literature, 18 lead us to the conclusion that this may be overly pessimistic and that even at an individual level, detectives who adopt a more positive mind-set and make greater efforts to engage members of such communities, may improve the chances of a successful outcome. Ultimately, however, it is likely that any significant reversal of the declining trend in clearance rates will only be achieved by serious and sustained efforts at a strategic level to engage and gain the trust of communities that are apathetic, frightened, hostile, or otherwise resistant to working with police.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.