From Ethnomethodologist to Data Lover: How I Became a Keen Advocate for Using Administrative Data

Abstract

This article is a personal reflection on a professional journey from ethnographer to keen analyst of administrative data. Using a knitter’s analogy, I describe using the thread of administrative information, armed with one stick (social science methods) and the second stick, practice-relevant, theory-informed research.

This article is a personal reflection on a professional journey from ethnographer to keen analyst of administrative data. Using a knitter’s analogy, I describe using the thread of administrative information, armed with one stick (social science methods) and the second stick, practice-relevant, theory-informed research.

Main Text

Throughout My Career I Cared about Administrative Data …

As a young social science researcher in New York, my role was to collect data from the rolls of leather-bound books across New York state courts. To assure geographic coverage, it was my job to drive to obscure courthouses to fling open the public and legal records of prosecutions of drug offenders. The New York Drug Law Evaluation Project published its findings in March 1978, and it is worth starting this reflection with a quote from that report: “…passing a law is not enough. What criminal statutes say matters a great deal, but the efficiency, morale and capacity of the criminal justice system is even more of a factor in determining whether the law is effectively implemented.”¹

Joining academic life in the 1970s, my academic research highlighted violence against women and girls in particular, influenced by the rise of the second wave of feminism in the 1970s and 1980s. My academic learning focused initially on deep description and social observations, exposing women’s victimization and its deep impact on the kinds of violent crime women experience. Social science data had to be “collected,” recorded, then analyzed. Not surprisingly, there was little accessible “official” data on violence experienced by women and girls, despite the fact that the homicide statistics routinely registered women’s killers as typically ex- or current partners. As the feminist research clearly showed (and shows today), many of the experiences of gendered violence were obscured in the official record. Furthermore, few academic researchers had access to recorded official information (police recorded crime), which published information on the number of crimes but not much information on what the substance of these crimes were.

My 5 years of experience as the director of the Economic and Social Research Council (ESRC) Violence Research Programme (1997–2002) opened my eyes to the availability of official evidence of so-called hidden violence. Projects exploring experienced violence—in schools, in prisons, as employees in churches, hospitals, nightclubs, and much more—uncovered not-so-hidden records about the impact of violence on social institutions, social policy, and most importantly, people. I took a decision to join government to champion ways of using administrative data to help drive improvement to criminal justice. It was my decade and a half in the Metropolitan Police and the Mayor’s Office for Policing and Crime in London that cemented my knowledge about how administrative data is a force for good in driving better policing and better justice.

Knitting as a Metaphor: Using the Threads of Data with Two Sticks

I am also a knitter. Thread throughout this short piece will be references to knitting, the skill of turning a thread of yarn into many different functional (and artistic) garments. All you need for a cardigan is two sticks and thread. My role in the Met Police and the Mayor’s Office (as head of evidence and insight) was devoted to turning the thread (here, recorded crime complaints) into usable insight about policing (for the purposes of strategic planning and workforce management) and policing’s justice outcomes. As a social scientist, I began to work with crime statistics as social science data—arguing, for instance, that its geographical concentrations or impacts on vulnerable populations are not just social science hypotheses and theories but appear as strong features in recorded crime data. Although there continue to be debates about the comprehensiveness of these crime data, improving crime records is important as much for effective, efficient strategic policy as it is for locally grounded (and informed) engagement with communities’ needs. Servicing a wide range of police functions, initiatives, and most importantly through the analysis of police performance on crime on a strategic basis, my role enabled a continuous dialog based on “what the crime reported information” tells us.

How Good Is the Thread?

In 2014, UK National Statistics removed its kite mark “official” from police-recorded crime from Scotland, England, and Wales. Routine statistics on recorded crime published by police earn their status as “official” from the UK Stats Authority if these are judged to be trustworthy, of good quality and of good value. In 2016 Police Scotland’s kite mark returned (NI’s was never affected). Unfortunately for England and Wales, policing’s own benchmark for their work—recorded crime—has not regained this important status as “official” crime data. No doubt in my experience some policing operational knowledge is poorer as a consequence.

I learned in my over 15 years inside that the culture of tolerating and working around poor-quality records was saturated within police working culture. An internal commitment to improving crime recording was not only important to the social scientist in me but is required to improve safety and security. There continues to be a struggle with more accurate recording of crime information in England and Wales—time pressures, a lack of understanding as to why some of the detailed information is important to strategic analyses, or just badly supervised work—that influences the information linking victims' reports to official records of crime.

As an advocate for more accurate crime reporting, this was an affirmation that the thread of information is important. My role as an inside advocate for a more data-driven approach to understanding crime problems in London was not always easy as a result. I learned that crime data could be observed alongside the records of the ambulance service or medical information. They could be placed in the context too of what Londoners told the Deputy Mayor Office for Policing and Crime’s Public Attitude Survey, which routinely asks residents about their experience of and expectations for policing. As someone who prepared briefings for the Senior Officer’s Management Board monthly performance reviews, I brought to their attention the state of crime in London, whether it be a rise or fall in violent crime, or to an analysis of a particular police operation (e.g., “what was the impact of the operation on crime”?). The data were not always welcome news to the Commander in charge (did the kind of crime targeted in the police operation decrease? Were local people more trusting of policing in their area after police initiatives? Is London getting safer?). It took at least a decade before the data itself gradually became trusted to tell a story about the wider narrative of safety and security through “what the data tell us.” Officials inside the police service and the Mayor’s Office have to trust the quality of the analysis data provide. Today, this information has become the starting point for oversight discussions. The thread is becoming more robust.

Knitters Routinely Use Two Sticks

I return to my knitting analogy (two sticks as my tools) to emphasize two principles I relied on as basic doctrines, and these might assist those currently working inside public sector organizations. Stick One: Rely on the data you have. Advocate relentlessly for its improvement (what’s missing? What kind of questions do strategic decision makers ask that cannot be answered by the information you have? Can you help improve the standard information that is routinely collected?). Stick Two: Use the data to explore both performance (in the widest sense) and strategic policy (operational, data-driven) conversations. Link these to the best wisdom that academics provides. (What drives better services or outcomes for people? What kinds of blockages impede “better” police responses, encounters with the public, or intervening earlier in dangerous situations or people? Where are people reporting more kinds of crime, and have these changed over time?)

These two principles led to improvements in the kinds of information available over time. In particular, the changes in the monitoring of hate crime have led to better action and conversations about the experiences of those targeted by racist or religious violence. I believe that for the public and for officials, there are data useful for steering the kinds of conversations that need to be held about making London safer (see www.london.gov.uk). I am confident that there are now data (in the public domain) that are driving these conversations about crime and justice in London. The more transparent these data are in the public domain, the better and more informed are the public and others in being able to have more intelligent conversations about what we—the public and government—do differently to improve safety and security in London.

Knitting Whole Stories: Using Crime Reports to Explore Reported Rape in London

Any knitter will tell you that creating garments is not straightforward. The design, the yarn, the gauge of the needles (the “sticks”) are all influenced by the way individual knitters knit. The display of the final product is for all to see, for knitters often wear their work. This last part of this article offers one more example of a finished product leading to better conversations about justice. I will use a recent study of a largely hidden crime; rape, as my example. Research suggests the low proportion of rapes (reported, for example, to the Crime Survey of England and Wales) are reported to police. In 2005, I was asked to undertake a review within the Metropolitan Police, and over a period of eight years, the exercise was repeated seven times. Last year, the Mayor’s Office for Policing and Crime repeated the exercise using rape reports from 2016, so that the cases could be traced through the whole journey through the criminal justice system using crime reports as the administrative data source.

These data show a remarkable consistency, and some change in the characteristics of the kinds of cases coming to police attention (see https://www.london.gov.uk/mopac-publications). In 2016, those reporting rape were predominately female (89%), two in three victims were under 30 years at the time of the offense, and overwhelmingly 95% had at least one marker of vulnerability (e.g., mental health, under 18, learning difficulties, and so forth). Only one in seven reported rapes were submitted to the Crown Prosecution (CPS) for charging considerations, and three in one hundred of the reported rapes resulted in a conviction.

The 2019 study shows that there is a rich seam of information in crime records. Further, exploring the justice journey was accomplished through matching the CPS administrative record with that of the police record. Insight into the strongest predictor of case attrition (reported rapes that do not complete a full justice journey) shows that it is the victim her/himself who is likely to be the actor who decides for whatever reason to stop cooperating with the retention of the crime complaint as a criminal investigation. It does show, however, if police succeeded in conducting a Video Recorded Interview with the victim (at the start of the investigation), then victims are 6 times more likely not to withdraw from the proceedings.

There is a richness of information here. This was accomplished through a detailed hand coding of the administrative crime record. For the purposes of better knowledge about victims and harm, however, there needs to be a way of harnessing information in a more accessible electronic format. Independent researchers will ask routinely about what kinds of crime happens to victims—where, when, in what context, and from which offenders. This information can weave into a narrative which, set alongside policing and community efforts, can set a more robust way of challenging rape in London. It is possible—returning to my metaphor—to construct a rich tapestry from two sticks and a thread.

In Closing

Using police reported crime as administrative data is more difficult if the researcher is on the wrong side of the “firewall.” I was advantaged for a number of years to have access to crime data as a way of working. I strongly advocate better linking of administrative justice data. The Administration Data Research (ADR) UK/Ministry of Justice Data First Programme is an important step in this direction. Decades on from my wrestling of leather-bound court records, I find myself writing and speaking about the use of administrative justice data—in the improvement of policing services and security and wider justice. I continue to be interested in knowing about people’s understanding and reality of personal safety and security and how this intersects with law, social order, policing, and justice. The advantage today is the growing interest from academics and government officials to harness and to improve routinely collected information, mobilizing a commitment to harness data as assets for a better social world. As the National Audit Office states in its 2019 publication Challenges in Using Data Across Government, government needs to up its game on the use of administrative data for well-informed decisions in government programs and services.² Indeed. Through this journal we will find other kindred spirits, techniques, approaches, and insight into how we can do this better together.

Biography

Professor Betsy Stanko OBE is the External Advisor to the Data First Programme, an ADR UK/Ministry of Justice collaborative project. Retired (April 2016) as Head, Evidence and Insight, Mayor’s Office for Policing and Crime in London, she had a second career in the Metropolitan Police following twenty-five years as a professor of criminology in the UK and USA.