Problems and Prospects in Policing in the ‘Post-Ferguson’ Era: Insights from a Southern American State

Professor Ross Deuchar

On 9th August 2014, an 18-year-old black man named Michael Brown was shot and killed by a white police officer named Darren Wilson in Ferguson, Missouri, USA. It later transpired that Brown had been stopped and queried because he was walking in the middle of the road blocking traffic. However, subsequent witness evidence suggested that Mr Brown had been unarmed during the incident and that he was cooperating with officer Wilson’s commands when the officer shot him. In the days and weeks following the incident, members of the Ferguson community engaged in protests and peaceful demonstrations and were later joined by citizens from outside the community. The media reporting of and political engagement with these events have continued to have a lasting impact on law enforcement across the USA in the last few years (Deuchar et al., 2017).

During a recent Fulbright scholarship, I conducted empirical research in a southern American State focused on exploring the perceived nature and impact of the so-called ‘Ferguson effect’ on serving officers there. I wanted to examine the ways in which the media portrayal of American policing and political engagement in the issues emerging from Ferguson had impacted on law enforcement within the State. Specifically, I wanted to explore the perceived repercussions in terms of citizen engagement and the apparent impact on officer morale and policing strategies. As with most of my other research, I drew upon ethnographic research methods (Hammersley, 2006). Across a period of three months I engaged in ride-alongs during police deployments with both departmental officers and those working at County Sheriff level, recording observations of police interactions with citizens and also my own impressions and feelings of what I was observing in fieldnotes. I also conducted semi-structured interviews with a sample of 20 serving officers of various ranks, exploring in more depth their reactions to the media reporting and political rhetoric surrounding policing in the years since the Ferguson shooting.

A substantial body of criminological literature has drawn attention to the links between public perceptions of procedural justice, police legitimacy and the tendency towards citizen engagement with officers (Mazarolle et al., 2012; Lum and Nagin, 2015; Weisburd et al., 2015). The officers I interviewed felt strongly that the media reporting of the high-profile incident in Ferguson had had a profound impact on public perceptions of the police in communities of colour in the State I was working in. They felt it had led to a perceived lack of police legitimacy, and reduced cooperation with officers on the streets. They described the way in which local people had developed an increased tendency to resist officers’ instructions, to film encounters with their cell phones and were even less inclined than before to report crimes or volunteer to be witnesses. As well as blaming the media reporting for this, informal dialogue with officers suggested to me that they felt that the wider political environment over the last eight years had not been conducive towards winning public support for policing. By supporting the Black Lives Matter campaign, they felt that President Obama had subtly reinforced the message that American policing was racially biased.

During interviews, the officers reported that morale was at an all-time low because of the increasing levels of public disengagement with law enforcement they came across in marginalised communities. Many officers felt that the police had become demonized, and that the political and media climate had reduced their ability to win the hearts and minds of those living in communities of colour. Further, many described the way in which they and their colleagues had become less inclined to be proactive in their enforcement for fear of being accused of racism. Tyler (2006) has suggested that, where police legitimacy is questioned, crime is often found to be greater. Although I had no direct observational evidence to substantiate this, some of the interviewees suggested that the increased reticence among officers to confront young black men in disadvantaged communities for fear of how they might be portrayed in the media meant that some criminal activity was now going unchecked. As one County Detective put it, ‘now [cops] are driving by and the criminals know that.’

In spite of the demotivated views of many officers, in one police department I worked in I did detect that the ‘Ferguson effect’ had raised awareness of the need for procedural justice among frontline cops. During my ride-alongs with officers in this particular city, I frequently observed them getting out of their cars to talk to local citizens and this was particularly noticeable in communities of colour.  On several occasions, I saw both male and female white officers actively building rapport with young black men on street corners, and several articulated their strong commitment to community-oriented policing approaches as a means of building trust and upholding dignity and respect. In this same city, there had been some officer-involved shootings of black citizens in the past, but no subsequent public protests or demonstrations akin to those that emerged in Ferguson. Officers attributed this to the strong focus they had adopted on being ‘guardians’ as well as ‘warriors’ within the local neighbourhoods (PTF, 2015).

Incidents like the one in Ferguson, Missouri have become all-too familiar in the United States over the years, but it is clear that the media and political backlash from this particular event has had a profound impact on law enforcement across the country. Indeed, although the southern State where I worked was over 1000 miles away from Ferguson, the impact of the repercussions from Michael Brown’s shooting had been felt strongly by officers I worked with. While my insights suggested that this had contributed to reduced officer morale and proactive enforcement strategies, it was also beginning to prompt an increased focus on the need for positive community engagement and procedural justice, at least in some divisions where I worked. In officers’ minds, the desire to re-focus their efforts on becoming humanitarian ‘guardians’ was perhaps principally driven by a desire to prevent another ‘Ferguson’ incident (Deuchar et al., 2017). In my mind, however, it more importantly represented an attempt to become the type of ethically-minded, socially just law enforcement officers that the public might finally begin to regard as legitimate.

Ross Deuchar is Professor and Director of the Interdisciplinary Research Unit on Crime, Policing and Social Justice at the University of the West of Scotland, and Affiliate Professor of Criminology and Criminal Justice at Florida Atlantic University, USA. Email: ross.deuchar@uws.ac.uk

 

 

 

Rethinking deaths after police contact

Dr David Baker

Between 2004 and 2015 a total of 1,539 people in England and Wales died after contact with the police (IPCC 2015). The term ‘death after police contact’ (DAPC) is used by the Independent Police Complaints Commission (IPCC) and includes categories such as shooting, apparent suicides in, or following police custody (within 48 hours of being released from custody), deaths in road traffic accidents, and deaths in police custody. Police officers are seldom subject to a criminal trial in cases of DAPC, and it is extremely rare for them to be prosecuted if they are. Yet the state is legally obliged under Article 2 of the European Convention on Human Rights (ECHR) to investigate cases of DAPC in an independent forum. The state should demonstrate a duty of care to citizens and ensure their right to life is proactively enabled. Each case of DAPC in England and Wales is typically investigated by two independent organisations (the IPCC and the coronial service) and police are held to account for their actions. This blog post is informed by research into documents produced by the coronial service and IPCC in sixty-eight cases of DAPC in the period 2004-2015. Narrative verdicts recorded by juries in the coronial system and IPCC independent investigation reports into correlate cases were analysed to assess trends and patterns in relation to how these deaths occurred and how they were investigated. The result is a monograph (2016) ‘Deaths after police contact: constructing accountability in the 21st century’. This post sets out some of the conventional knowledge about this issue, and adds some original findings from the research that suggest it might be helpful to rethink the issue of DAPC in England and Wales.

There is no official denial of the real and symbolic importance of cases of DAPC to society. The capacity of the state and society to hold police to account in these cases is seen as a touchstone for legitimate, transparent and consensual policing in England and Wales. Similarly, there is no official denial that deaths in state custody are significant because the state bears a unique responsibility for the welfare of citizens in their care, and a death in custody can often be viewed with suspicion by the public. Moreover, there is no official denial that a disproportionate number of citizens from marginalised groups in society die in these cases. If you are from a Black and Minority Ethnic (BME) group, have mental health issues, or are dependent on substances then there is a disproportionately large chance that you might die after police contact. None of these issues are disputed by the state. It has made numerous official pronouncements stating how important lesson learning is in reducing the number of deaths after police contact (see, for example Fulton 2008, House of Commons Home Affairs Committee 2010, Joint Committee on Human Rights 2004). This much is unequivocal about deaths after police contact. A key question is: are lessons learned?

Much of the societal and media focus on the issue of DAPC identifies police use of force. It is relatively rare for police to shoot individuals dead in England and Wales. During the period 2004-2015 approximately 2 people per year were shot dead (IPCC 2016), compared to an estimated 1,000 per year in the United States (The Counted 2016). Having said that, approximately half of the 68 deaths examined in my research related to the use of physical force, principally restraint. So how did the other half die? A key finding in my research was that in approximately half of the cases, deaths occurred at least partly as a result of police inaction while in custody. Juries in coroners’ courts used benchmarking from official policies to establish that prescribed protocols were not followed regarding the care of detainees. A further finding was that more than half of the deaths examined occurred either as a result of the failings of healthcare agencies, or with healthcare agencies present at the time of death. The police are not only the ultimate emergency service for the public, but apparently also for other emergency services.

I believe it would be helpful to reimagine these deaths as healthcare crises. Many of them are preventable deaths if viewed from a healthcare perspective. Preventable if dealt with efficiently by healthcare services, or by multi-agency working between police and other emergency services. The term ‘death after police contact’ is disingenuous in that often these cases involve healthcare services in addition to police; and frequently there is no contact, instead detainees die at least partly due to inaction. Rethinking these deaths as failures in healthcare could help us rethink how we learn from them to prevent future deaths. A failure to learn lessons in the patterns of deaths after police contact has been repeatedly noted by official agencies and NGOs (see Hannan et al.  2010, Fulton 2008, Shaw and Coles 2007, House of Commons Home Affairs Committee 2010, Joint Committee on Human Rights 2004, EHRC 2014).

The number of people who die after police contact is relatively stable in this country.  Public organisations and regulators have produced an increasing number of reports into this issue, but the number of deaths remains stubbornly unchanged, as does the disproportionality in those who die in such circumstances. We now know more about how people die in custody than we ever did, yet seem to be unable to translate this knowledge into practical applications that reduce the number of preventable deaths. The primary goal of a police officer is to preserve life, rather than necessarily enforce criminal justice (ACPO 2006). Reimagining the issue of DAPC in terms of a healthcare crisis might enable us to focus on the role of police as ‘peace officers’, to see the duty of care of detainees as being central to the police role, and to acknowledge the role of healthcare agencies in being part of a solution to this crisis.

Dr David Baker is a Senior Lecturer in Criminology at Coventry University. Email: d.baker@coventry.ac.uk  

Decision Making Theory and its Implications for Policing

Lee Curley

Detectives are unique decision makers, who have to search for information, weight information and make an appropriate decision. There are four possible outcomes from a decision made by a detective; a true positive, a true negative, a false positive, and a false negative. A true positive is when the correct individual is charged, and a true negative is when an innocent individual is let go. A false positive is when a detective may charge the wrong individual. A false negative is when charges are dropped on a guilty individual. Although, these outcomes may be outside the control of many detectives; with the prosecutor fiscal (in Scotland) and the Crown Prosecution Service (in England and wales) having ultimate control over whether charges are dropped.

The implications of false negatives and false positives can be huge. For instance, if an innocent individual is incarcerated, then an innocent individual’s life is ruined and the real perpetrator is still free to pose a danger to society (Ramsey & Frank, 2007). In addition, if false negatives and false positives occur, then public perception surrounding the justice system may be damaged, which may reduce the deterrent effect of the penal system (Ramsey & Frank, 2007). For example, if citizens believe that they are unlikely to be charged for a crime, they may be more likely to commit such a crime in the future. It is important, therefore, that the decision making processes of detectives are studied.

Rationality versus Intuition

There are two main ways a decision maker can reach an outcome. One way is through being rational (Lee & Cummins, 2004). Rational decision makers use all the evidence available, weight it, integrate it and then make an appropriate decision (Gigerenzer, & Goldstein, 1996). Rationality, however, is halted in environments where there are time pressures, information is emotional, the decision making environment is ambiguous and when the information is complex (Bell, Mawn, & Poynor, 2013; Bright & Goodman-Delahunty, 2006; Cooper, Bennett, & Sukel, 1996; Tversky & Kahneman, 1974). Therefore, researchers, such as Snook and Cullen (2008) have suggested that it is not realistic to expect decision makers to be rational, and rather intuitive processes may be more likely. Consequently, the implications of intuitive decision making within detectives should be researched more.

When referring to intuition in this piece, I am referring to cognitive short-cuts called heuristics (Tversky & Kahneman, 1974). These heuristics are fast and frugal strategies that allow decision makers to use a limited amount of cues to navigate the environment around them (Gigerenzer & Goldstein, 1996). In that sense, intuitive processes are less complex, when compared with rational decision making models, such as Bayesian analysis (Gigerenzer & Goldstein, 1996).  Simon (1956) proposed that the environment does not allow for fully rational processing to occur. Consequently, decision makers have a limited, yet efficient, cognitive system that allows them to make judgements in the noisy environment that surrounds them. The question this blog hopes to answer is, what implications can intuitive decision making have for decision makers within the police?

Detectives and Tunnel Vision.

The most widely cited irrational intuitive process relating to detective decision making is tunnel vision. Tunnel vision can be defined as a collection of heuristics and cognitive flaws that cause legal and forensic decision makers to centre on one suspect, and then search for guilt confirming information, and disregard disconfirming evidence (Martin, 2004). Findley and Scott (2006) suggest that tunnel vision is a negative decision making strategy employed by detectives.

Through investigating several case studies, Findley and Scott (2006) proposed that a common theme that seemed to arise in miscarriages of justice was the utilisation of tunnel vision. They suggest that tunnel vision, which may be accidental, can have an asymmetrical effect on officers’ search for, and interpretation of, information. For instance, an incorrect eyewitness identification could cause detectives to believe that a particular suspect is guilty (Findley & Scott, 2006). This may then effect how later evidence is perceived, with ‘neutral’ information being seen as guilt confirming (Findley & Scott, 2006). This conclusion has been supported by Ask, Reinhard and Marksteiner (2011) who found that individuals judged criminal evidence to be less credible when it disconfirmed a prior assumption of guilt.  The more “elastic” the evidence was the more susceptible participants were to the negative evaluation of disconfirming evidence. For example, evidence such as eyewitness testimony is easier to stretch to fit prior beliefs in comparison to DNA evidence. Ask, Rebelius, and Granhag (2008), also, found that police trainees were susceptible to judging such ‘disconfirming’ evidence as more unreliable than consistent information.

Additionally, Findley and Scott (2006) suggest that the perception of guilt may then bias investigators into interrogating the suspect in a manner that may lead to a false confessions; that is, detectives may ask more leading and guilt directed questions. Hill, Memon and McGeorge (2008) found that a belief of guilt did lead to a self-fulfilling prophecy. In other words, interviewers who believed that the target individual was guilty, were more likely to ask guilt related questions, and thus prove to themselves that the person was guilty. Further, the self-fulfilling prophecy is a vicious cycle as a pre-decisional belief of guilt changes how individuals search for and evaluate information, thus causing prior beliefs to be reinforced.

Snook and Cullen (2008) suggest that the negative perception of tunnel vision is premature; heuristics may have positive implications for detectives. They argue that it is irrational to believe that police officers should investigate every potential suspect, evaluate all the evidence, and search all possible lines of enquiry.  It is not heuristics that lead to negative outcomes, rather, erroneous information leads to such consequences (Snook and Cullen (2008). Heuristics may actually help detectives to navigate through a noisy crime scene and focus on relevant information, rather than all the information. Therefore, heuristics may decrease the cognitive load of detectives and fact finders, which may allow them to be more efficient decision makers.

Conclusion

In conclusion, fact finding and decision making in regards to policing happens in a complex environment, which may lead to intuitive processes occurring. However, the implications of these intuitive processes are still unknown. Some research suggests that intuitive processes are suboptimal and should be prevented, whereas alternative research proposes that heuristics evolved to allow us to make decisions in a complex environment. Ironically, it could be said that more information is needed. Future research should aim to highlight the implications that heuristics may have in detective decision making. My current research investigates heuristics and decision making strategies within a juror population, I look to see which variables promote accuracy, what causes cognitive fallacies and how individuals evaluate evidence. However, I do hope that for my post-doctoral research, that I can utilise my expertise in decision making and statistics and apply it to a policing context, and thus help to fill in some of the gaps I have highlighted in the literature.

Lee Curley is a PhD student at Edinburgh Napier University. Email: Lee.Curley@napier.ac.uk Twitter: @Psycurlogy

A Uniform Not Uniform

Dr Camilla De Camargo

Police officers are distinct and unique actors in public spaces. They experience a peculiar familiarity with wider society: they often do not personally know the citizens in the areas they patrol but everyone knows that they are part of the police by their uniform. The image of the police has many different aspects and features: it has been designed and re-designed since the inception of the modern police force in 1829 and continues to change year on year. The police uniform is an important part of this image work and image management (Mawby, 2002: 5); it embodies not only the physical exterior, but also the character of individual police forces.

Adaptations to Uniform Policy

The visual representation and symbolic meanings attached to clothing is constructed, reconstructed, managed, controlled and communicated, not only by the uniform designers, but by police officers themselves to make the meaning of their uniform personal. As a consequence, paradoxically, the uniform lacks uniformity in style. Let’s consider the word ‘uniform’ then – it is, according to the Oxford Dictionary, ‘[r]emaining the same in all cases and at all times; unchanging in form or character’.  Yet, for various reasons, different versions of the police uniform exist (discussed a little later). It is important therefore to consider how adaptations and modifications can affect public perception of the police. West Midlands Police have said they would consider allowing their Muslim female officers to wear the burka or niqab if they wish, in a drive to recruit more officers from BME backgrounds. The Met and Police Scotland have previously approved the hijab but the burka would be  a giant leap in uniform adjustments. Indeed, Conservative MP David Davies  criticised the burka as being “a symbol of oppression for women and not something that a modern police force should be supporting” and of course, who can forget the furore surrounding the burka/burkini and the police trying to enforce its ban in France. Our concern however is not with meanings of the burka (or niqab) per se but rather how their incorporation into official police uniform might affect police image work.

So what is a uniform?

Scholars have debated what actually constitutes a ‘uniform’. Simply, ‘(it is) a distinctive dress worn by members of the same body’ (Randell and Gray, 1995: 16). Examination suggests that a uniform is: ‘a prescribed set of clothes which enables the observer to identify the wearer’s organisation or affiliation’ (Bhugra and DeSilva, 1996: 393).

Society learns to distinguish members from non-members of different groups through their uniforms. The attire informs the actor and the audience what to expect from the individual and the organisation. By requiring the donning of a uniform, and thus suppressing personal choice of clothes, the police institution ‘binds the individual to his or her peers, underscores common membership, and encourages a sense of loyalty among members and faithfulness to the same rules’ (Rubenstein, 2001: 87). If an officer does not abide by these rules, the alienation from the group is clear: if the transgression is serious enough, police officers are required to remove (and return) their uniform and accompanying accoutrements that are similarly embedded with the organisation’s authority.

A National Uniform

The police institution’s failure to roll out a national standardised uniform is due to managers not ‘buying into the concept’ (Police Review, 2007: 4), yet this negates the notion that a uniform is actually uniform in style across constabularies. The former ACPO Head of Procurement cited the ‘initial cost of a switch in uniform and a desire to maintain the identity of local forces are some of the factors discouraging people from the national uniform’ (Davies, cited in, Police Review, 2007: 4).

The police uniform, though essentially similar between different forces, has never been exactly the same in detail, colour or insignia nationally. These dissimilarities also apply across ranks and divisions due to the presiding Chief Officer’s preferences for each force, their access to different supply chains locally, and their individual budgets. The 1934 Dixon Committee on uniform consistency advised that ‘all police should be capable of turning out, in both garments and head-dress, approximating closely enough to a uniform pattern as to respects to material and style’ (cited in Clarke, 1991: 16emphasis added). Although the police uniform is symbolic and highly recognisable, it is debatable whether not having a set standard uniform for the same ranks causes problems with perceived legitimacy.. Young (1991: 72) contends that the police world does, unsurprisingly, ‘value uniformity’, and as the language suggests, ‘embodies the essentials of a system obsessed with physical and ideological concepts of order and discipline’, which is ‘massively symbolised by the uniform’.

Conclusion

If the police uniform is so iconic and important, in regards to perceptions of power and authority, is it not crucially important that all the police forces in England and Wales wear exactly the same uniform? Perhaps we’re too caught up in idealistic images of our British police in times of tension and unrest; after-all, we often see children falling out of the school gates, ties askew, shirts crumpled and untucked and wearing various items of non-regulation paraphernalia. Yet we know, from a glance, that they all go to the same school. But alas, school children are not hired to protect us, symbolically or otherwise! Research findings from my PhD showed that there are significant disparities between forces – in some London boroughs for instance, there is a shift back to the more ‘professionalised’ look of white shirt, tie and smart trousers; further up North however, there is more of a lean toward ‘militarised’ all black clothing. Other findings show that there are problems with legality and complaints when officers choose to modify their outfits; recognition and trust is important in this relationship, a problem particularly when PCSOs are mistaken for PCs (see De Camargo, 2017 – forthcoming). Across many UK cities, the public must make clear distinctions between PCs, PCSOs, special constables, private security, street wardens, city-centre night marshals, environmental crime officers… and this list is far from exhaustive. This presentation of ‘hybrid’ police may actually ‘heighten anxiety in the community’ (Cooke, 2005: 233). The ‘putting together’ of one’s uniform is personalised as well as regulated… to an extent. Various accoutrements, uniform versions (outdated or otherwise) and huge discrepancies between forces all add up to the uniform not actually being uniform. So while it is admirable that West Midlands Police are using (positive) discriminatory hiring techniques to increase the amount of BME officers to build better relationships with the diverse communities they serve, allowing a blanket covering of identity (and further uniform modification) could potentially further damage the already fragile and fractious police/public relationship.

Dr Camilla De Camargo is a Lecturer in Criminology and Sociology at Kingston University. Email: C.Decamargo@kingston.ac.uk, Twitter: @DecamargoC

 

Small Area Estimation for Targeting Hot-Spot Policing Interventions

David Buil-Gil and Angelo Moretti

Hot-spot policing strategies involve more focused police attention on small geographical areas where crime (and disorder) is more prevalent. During the last two decades, multiple evaluations have shown that police patrols targeting micro-locations with relatively high crime rates within a territory are effective in reducing offending, mediating what has been called a “micro-deterrence” process (e.g. Braga and Bond, 2008; Braga et al, 1999; Braga et al, 2012; Braga et al, 2014; Sherman and Weisburd, 1995; Sherman et al, 2014; Telep et al, 2014). The efficacy of hot-spot policing for targeting such areas depends, therefore, on available data geocoded at low-geographical level. Mapping crime is the first step for a smart policing intervention. Police-detected crimes are easily mapped at low-geographical level, as are calls for police service. However, an appropriate hot-spot policing approach cannot depend solely on known or officially-detected crimes, but also on non-detected offences, citizens’ perceptions of disorders, fear of crime and attitudes towards the police, among others. The latter variables are usually based on surveys with small samples that cannot be easily mapped at low-area level. The following sections are going to detail a new generation of statistical techniques that are helpful for obtaining small area estimates of survey variables; estimates that can play a vital role on a multidimensional-based hot-spot policing.

Some of the first approaches to hot-spot policing focused mainly – and sometimes only – on detected crimes as the unique data source for targeting small area patrols and evaluating those interventions (e.g. Sherman and Weisburd, 1995). However, more recent studies have established the need for including other variables in the databases and maps used by police officers to decide where to target major efforts and how to assess their impacts (e.g. Braga et al, 2014; Ratcliffe et al, 2015; Weisburd et al, 2011). Hot-spot policing interventions, whilst having crime reduction as their main objective, also tackle citizens’ perceptions of deviance and aim to decrease feelings of unsafety. Trust and police legitimacy also need to be analysed when assessing hot-spot patrolling (Weisburd and Telep, 2014). Officially-detected crimes collect three to five times fewer offences than victim surveys (Coleman and Moynihan, 1996) and might be influenced by criminal policy and police propensity to record offences (Aebi, 2010). It is therefore necessary to assess survey-detected victimisation as well as other variables related to subjective perceptions of disorder and trust in the police. As noted, the main source of data for obtaining information about these necessary variables beyond detected-crimes is victimisation surveys.

Victimisation surveys are sample surveys designed to obtain data on personal experiences with crime and deviance. The Crime Survey for England and Wales (CSEW), formerly the British Crime Survey, is the main source of data for analysing these variables in the territories included in its sample. However, the CSEW sample is mainly designed for being representative of the entire English and Welsh territory. With a few exceptions, the survey yields a minimum sample of 650 households for each of the 42 Police Force Areas (PFA), but the sample is not representative of small geographical areas within each PFA. In other words, the small number of households sampled in each of the 181,408 output areas included in the survey – zero in numerous cases – makes it impossible to support reliable direct estimates that can be used for police-targeting purposes. Output areas are small geographical areas designed by the Office for National Statistics to have similar population sizes (125 households on average). Whilst some police forces and local authorities run their own surveys, they are very few and their samples are not large enough to run direct estimates of adequate precision either. We must therefore rely on indirect estimates. In order to obtain indirect reliable estimates of the variables of interest at low-geographical level, small area estimation techniques are an option with great potential for the future (e.g. Taylor, 2013; van den Brakel et al, 2013; Whitworth, 2012).

Direct estimation methods use only area-specific sample data and sampling weights to produce small area estimates. This means that when the area sample size is small these estimates may suffer from low precision and are therefore unreliable for patrol targeting purposes. As noted, crime surveys usually sample few (or even zero) people from each micro-location/area of interest (e.g. output areas). Small area estimation aims to develop statistical techniques which produce efficient and precise estimates at small area level (also when the sample is not large enough to support direct estimates). In order to calculate estimates of adequate precision for small areas, indirect estimation methods make use of auxiliary information from other sources of data (e.g. Census) and produce linking models that borrow strength from related areas or time periods. Thus, this type of statistical inference is ‘model-based’.

According to Rao and Molina (2015), small area models can be classified into two types: area-level and unit-level models:

  • Unit-level models, which relate individual or unit values of a study to unit-specific covariates, can only be used when data is available at individual or unit-level. Due to confidentiality criteria, data from victimisation surveys is rarely released at individual level, thus making unit-level small area estimation impossible.
  • Area-level models relate small area direct estimates to area-specific auxiliary information available from other sources of data. Auxiliary data from the Census and other administrative records can be used in area-level small area estimation to produce reliable estimates at small area level of crime surveys’ variables.

Weisburd and Neyroud (2013) argue that a wider involvement of science is necessary in the arena of evidence-based policing. Officially recorded crime data mapped at small area level appears insufficient for targeting hot-spot patrols, as they only collect a small proportion of crime and can be biased. Victimisation surveys are today a useful tool for obtaining further information on crime and deviance, but their small samples cannot be used to produce reliable direct small area estimates. Model-based small area estimation techniques are today a potential option for obtaining further reliable information at low-area level for a more complex, efficient and evidence-based hot-spot policing.

David Buil Gil is a PhD student at University of Manchester researching small area estimation techniques for victimisation surveys data. Email: david.builgil@postgrad.manchester.ac.uk. Twitter: @DavidBuil

Angelo Moretti is a PhD student at University of Manchester researching multivariate small area estimation methods for poverty and wellbeing indicators. His research interests cover also multivariate statistical techniques for data dimensionality reduction. Email: angelo.moretti@manchester.ac.uk

Algorithmic policing and freedom of information – how far should transparency extend?

Marion Oswald

Algorithmic transparency is the new watchword – from the UK Parliament’s Science and Technology Committee (House of Commons Science and Technology Committee, 2016), to the European Data Protection Supervisor (EDPS, 2016) and even Angela Merkel (Connolly, 2016).  With the current controversies over fake news online, the spotlight has fallen mainly on the private sector – internet platforms, search engines, social media – but for how long?

The operational deployment of algorithmic tools (computational methods that analyse data sets in order to draw conclusions, increase knowledge and make links) within the U.S. criminal justice system has provoked considerable debate.  Third party proprietary systems used for parole and sentencing decisions have been accused of hardwiring discrimination into the system (ProPublica, 2016) and raise issues of natural justice and procedural fairness (Oswald & Grace, 2016).

Predictive database-driven tools for offender management in the probation and prison contexts have been used in the UK for several years (NAMS, 2015).  However, the extent to which algorithmic tools have been adopted within UK policing for investigative operations and intelligence analysis appears to vary significantly between forces and be less than transparent.  For instance, a search on Kent Police’s website for predictive policing tool ‘PredPol’ brings up no results, despite coverage of the force’s deployment of the algorithmic technology in the UK press (see for instance O’Donoghue, 2016).

My freedom of information-based study (with Jamie Grace, Sheffield Hallam) into algorithmic analysis of police intelligence in the UK (Oswald & Grace, 2016) suggested that a relatively small number of forces were using computational or algorithmic intelligence analysis.  Although, due to the limitation of such studies, this is unlikely to represent a complete perspective, the results may still give support to James’s conclusion in October’s blog (James, 2016) that the institution is ‘failing to make the best use of its intelligence’ and so is ‘not working smart enough.’

Responses from UK police forces to our freedom of information request indicated that algorithmic tools were used both at the ‘macro’ level – for instance, assessing crime patterns – and at the ‘micro’ level such as for decision-making or risk assessment relating to individuals.  Detail was lacking however, with no details as to the specific crimes, activities, schemes or laws that were the focus of the tools.  For instance, it would not be possible to determine from the responses whether the tools were used to assist decisions pursuant to the preventative disclosure schemes ‘Clare’s Law’ and ‘Sarah’s Law’.   The reasons for using such technology were expressed in general terms, and the exemptions under the UK’s Freedom of Information Act engaged by many forces may disguise other uses of such tools, or, importantly, gaps in operational capability.

In response to our freedom of information request, thirty-two responses used some form of exemption.  Twenty-seven responses cited the Section 23 absolute exemption: ‘Information supplied by, or relating to, bodies dealing with security matters’ – MI5, MI6, GCHQ and so on – outnumbering the use of the Section 31 exemption (Law enforcement) cited by only five responses.  The Section 31 exemption was combined with the Section 23 exemption four out of the five times that Section 31 was used.  Section 23 was most often cited on its own rather than being combined with other exemptions.

Most (although not all) responses that cited Section 23 combined it with ‘neither confirm nor deny’ wording, which should be used when either confirming or denying would reveal exempt information in itself.  Section 31 is a qualified exemption, requiring the public body to engage in a public interest balancing exercise to decide whether or not to release the requested information.  There were considerable inconsistencies in the way that forces handled this balancing exercise, with some forces providing rather generic arguments either way, and one providing only a bland statement that a test had been completed.

So what can we take from this?  The inconsistent use of FOI exemptions may allow sensitive information to be inadvertently exposed or gaps in capability inferred, which could be damaging to law enforcement and national security interests.  However, if one thing can be learned from the debate surrounding the Investigatory Powers Bill, it is that the adoption of new technology by the State without appropriate transparency around methods can result in damage to public trust as well as legal challenge.

Operational details and methods must be protected.  Yet balanced transparency is also vital. Examples of such balanced transparency exist, for example in Canada, where, while a detailed understanding of the Violent Crime Linkage System (ViCLAS) ‘is quite rightly not in the public domain due to investigative sensitivities – but official, clear, easy to find information on what the system is, who can use it, how to go about access and related research is available‘ (Dawson and Stanko, 2016).

The Investigatory Powers Bill is in itself an example (in some eyes a flawed one) of an attempt to achieve balanced transparency and oversight.  As the use of algorithmic tools within UK policing appears to be at a fairly early stage, now would seem to be the right time to assess the legal underpinning and the governance framework, and to ensure that appropriate transparency is built into contracts with third party software suppliers.  Such an exercise may help to demonstrate the effectiveness of these tools, as well as increasing public trust and so could contribute towards James’s call for the institution to make the best use of its intelligence.

Marion Oswald is a Senior Fellow in Law, and Head of the Centre of Information Rights at the University of Winchester, and a solicitor (non-practising). Email: marion.oswald@winchester.ac.uk. @_UoWCIR @IRPandPJournal

Drowning not waving: after 40 years of neglect, what now should be done to develop police intelligence practice in Britain?

Dr Adrian James

I have been studying police intelligence practice for about 16 years (see James, 2016, 2014, 2012, and 2003). It has been a fascinating, intriguing, and frustrating journey in almost equal measure. Perhaps the thing I have found most frustrating is that after more than 40 years of scholarly research, police-led research and ‘official’ inquiries into criminal intelligence work, that practice looks very much the same today as it did in 1975.

The Audit Commission (1993) was a significant actor in stimulating interest in the subject, but insider-led scrutiny of criminal intelligence work began with the Baumber Inquiry (ACPO, 1975). This was followed by the Pearce Report (ACPO, 1978); the Ratcliffe Report (ACPO, 1986), the 1997 report Policing with Intelligence (HMIC); the Bichard Inquiry (Home Office, 2004); the  National Intelligence Model Review (to which I made a small contribution) (ACPO, 2014) right the way through to Building the picture: An inspection of police information management (HMIC, 2015). These are some of the most significant inquiries but I doubt this is an exhaustive list.

            Of course some things have changed. There is increasing specialisation in criminal intelligence work as well as a proliferation of intelligence assessments and sundry other analytical products. But, largely, these changes are presentational. Even if they have shaped practice at the margins, I question whether they have influenced policing in anything other than the most superficial of ways. Certainly, what really mattered to Baumber (and to a great extent to both Pearce and Ratcliffe) in terms of the development of a suitably professional and capable intelligence workforce – a collective vision for intelligence; common understanding of the intelligence basics; improved training and selection procedures that put the best and most experienced staff into those important positions – have not been delivered. I argue that is because, despite the rhetoric to the contrary, the institution has never truly accepted the case for them.

In previous publications, I have attributed this to human factors; the negative impacts of institutional conservatism, managers’ self-interest and professional subcultures (see James, 2016). I have also reflected on the extent to which there is an identity crisis in the intelligence world; left to its own devices has it lost sight of its raison d’etre? Perhaps, it is able only to focus on its own part of the business, without concerning itself with the external realities; the police organization’s unswerving commitment to action over reflection, its rejection of existentialism in favour of pragmatism. It may be more accurate to say that an indifferent executive has not allowed it to concern itself with those external realities by sidelining it from meaningful involvement in the decision-making process – that certainly was a significant factor in the story of the NIM (see James, 2013). I remain convinced that these realities are hugely significant factors in this environment and are detrimental to the development both of the work and the workforce but the more I reflect on the issues, the more I wonder if these are just very obvious symptoms of bigger problems rather than the problems themselves.

Recently, I have been exploring that idea with a colleague from the Finnish Police University. Our dialogue increasingly has focused on classic analyses of bureaucracies. Weber (1947) argued that bureaucracies are wired to become increasingly specialized precision instruments. Largely, their aim is not to fit better with the pieces of a larger puzzle but to achieve excellence in their own right. When the task is unclear, the bureaucracy loses its identity and floats without purpose. Perhaps, that is where the intelligence world finds itself today. Weber said that to exercise control, bureaucratic administrations relied on a combination of technical and experiential knowledge, which itself was a product of the “‘striving’ for power”, characteristic of bureaucratic organisations (1947 p.339). By that analysis, arguably, what we are witnessing here in the intelligence milieu is a natural development of the institution in which significant efforts to reshape the status quo through ‘revolutionary’ schemes such as: the first recorded intelligence-led policing initiative, the Aberdeen Policing Experiment; the Unit Beat Policing experiment of the 1960s, the Kent Policing Model (the precursor of the NIM) and the NIM itself (see James, 2013), have consciously (or unconsciously) been disdained for fear that in the long run they may have unknown and unintended consequences, more detrimental to the institution or to the individuals who wield power within it or to the communities it serves (not necessarily in that order).

Both Robert Reiner (2012) and Ben Bowling (2007) have argued persuasively that we should recognise the limits of policing. Bowling suggests that ‘good enough’ policing is as much as we reasonably can expect from the institution. Should we apply that same caveat to police intelligence practice? I am in complete agreement with Reiner and Bowling on the need for a kind of reality check on police capabilities but it seems to me that ‘good enough’ is only an acceptable aim when it can be demonstrated that the organisation and its individual members actually are doing their best, working efficiently, effectively, and ethically to provide the services that stakeholders and communities expect and need. For all the energy expended by the institution in the context of crime and criminality, the evidence suggests that it may be working hard enough, but in failing to make the best use of its intelligence, it is not working smart enough. As the reader has seen, identifying the root causes of these inefficiencies will be difficult; as the last 40 years have shown, resolving them may be much, much harder.

Dr Adrian James is a Senior Lecturer in Criminal Investigation at the Institute of Criminal Justice Studies, University of Portsmouth. Email: adrian.james@port.ac.uk