This post by Alex Luscombe and Alexander McClelland is part of the “(In)Security in the Time of COVID-19” series. Read the rest of the series here.
In the fall of 2019, the world saw the emergence and global spread of a new coronavirus (SARS-CoV-2) capable of causing acute respiratory syndrome (COVID-19) in humans. First appearing in Wuhan, China, COVID-19 quickly spread around the world, eventually reaching Canada. As scientists began work on a vaccine, governments sought ways of slowing the spread of the virus through emergency legal interventions, including travel bans, school shutdowns, and mandatory closures of non-essential businesses. Additionally, governments across Canada sought ways of enforcing these restrictions, granting a range of new COVID-19 related powers to police and bylaw officers nationwide.
In direct response to the announcements from Canadian government officials that police and bylaw officers would be fining, arresting, and potentially jailing Canadians for allegedly violating the new emergency orders, we launched the Policing the Pandemic Mapping Project (PPMP) in April 2020. PPMP is a Canadian data justice initiative that tracks and publishes data and critical commentary on enforcement practices in response to COVID-19. As criminologists, we launched this project out of a concern that widespread patterns of injustice and inequality that we had seen time and time again in other areas of policing would similarly characterize the enforcement of rules related to COVID-19.
Shortly after launching the project, PPMP garnered extensive national and international attention, engagement, and support from journalists, policymakers, activists, and concerned citizens. In a country where meaningful criminal justice data is too often kept out of reach from the public, PPMP offered the rare prospect of being able to analyze and make sense of emergent trends in pandemic policing in Canada in semi-real time. What started as a weekend side-project aimed at drawing attention to the potential harms of COVID-19 policing on social media evolved into a fully-fledged project with funding, a small research team, and collaborative partnerships with organizations like the Canadian Civil Liberties Association.
Over the past twelve months of doing PPMP, we have learned a lot, but not just about viruses, people, and law enforcement. We have also learned a great deal about data and data activism. Rather than reflect on our empirical findings and policy ideas, which we have written on extensively elsewhere, here we reflect on the lessons we have learned from doing data activism. Although we are criminologists rather than historians, we hope that these insights will be of value to readers of Active History, especially those interested in the emergent practices of “data rescue” and “guerilla archiving”. Below we reflect on four major lessons that we have learned over the past twelve months of co-leading our first data activism project in Canada.
Lesson 1: Just providing the data alone is not enough
PPMP started very simply: a map of media and government reports of police enforcement actions related to COVID-19 across Canada that users could interactively explore from their web browsers. Shortly after this, we also provided the same data in table format to allow people to more easily filter and extract the raw data and analyze it themselves. We soon realized, however, that this was not going to be enough. Our use of mapping was strictly descriptive and provided only limited insight into the enforcement of laws pertaining to COVID-19. Our goal was to raise awareness about the potential harms of mobilizing police in response to COVID-19, contributing to larger ongoing discussions about the role of police in society, and providing a map and table alone was hardly going to achieve this.
To partially address these limitations of our project, we pursued a more multilayered approach that extended beyond collecting and sharing COVID-19 enforcement data. We published a project white paper that situated our project in established theories of counter-mapping, social inequality, and anti-racism; we began producing periodic “enforcement reports” that commented on larger situational, demographic, and geographic trends in the data; and we began actively engaging with news agencies, activist organizations, and policymakers across Canada to disseminate our findings and policy perspectives.
These tactics of knowledge mobilization have proven to be tremendously successful. Over twenty thousand people from around the world have visited the PPMP website, consumed the project’s reports, downloaded the raw data, and engaged with some of the project’s findings in various other publications. PPMP reports have been widely disseminated online, and reported on in mainstream media outlets, with over thirty online and in-print articles referencing the project to date, including the BBC, CBC, the Globe and Mail, Global News, CTV News, and the Toronto Star, among others. Journalists regularly inform us that the project has become an extremely useful resource for their work.
Although the exact impact – if any – our project has had on policing practices and policy is more difficult to measure, we do not believe that it has been negligible. We have had members of a police services board contact us, noting that our reports were used to make decisions. Several months ago, a city councillor contacted us directly, noting that the police in their city were pursuing an education-first approach, citing PPMP’s data and commentaries as a major impetus.
Lesson 2: Always communicate the limitations
PPMP is a database of primarily quantifiable characteristics of when COVID-related fines were given out, the value of the fine, the law invoked by the acting police or bylaw officer, where the enforcement event took place, and so on. While these data can reveal a great deal about patterns in COVID-related enforcement, as readers of Active History will be well aware, they do not tell a complete story. The data points we collect for PPMP are in large part constrained by what we are able to collect. As the project relies on secondary sources of information – news articles and government communications primarily – we do not have control over the information that is made public. Demographic information on the age, race, gender, and socioeconomic status of those receiving fines, criminal charges, or facing arrest as a result of a COVID-related infraction is available only in fragments. Some cities have taken steps to ensure that such information will be difficult if not impossible to obtain even in the future. The City of Toronto, for instance, has explicitly stated that it will not be collecting any information about race in COVID-related incidents. This was announced around the same time that observers in the US and UK used publicly available data to measure disparities in COVID-related enforcement in those countries.
And, of course, these are not the only limitations of our project. Even if we did have more complete information about the demographic characteristics of the people being fined, charged, and/or arrested for alleged COVID-related violations, there would still be a great deal of contextual information missing. Why was someone fined? What was the nature of the interaction between the person being fined and the police officer? Who else was around at the time the fine was given out? Who was in the same area but not given a fine? How was the discretionary decision to punish someone made by the police or bylaw officer? These kinds of observations and insights are vitally important to the study of police responses to pandemic-related behaviours, but PPMP cannot provide these. It was never our intention that PPMP would serve as a one stop shop for information and analysis on the policing of COVID-19. Rather, PPMP was, and still is, only one possible intervention among many, and certainly not the most important by any means.
A final limitation of PPMP that we wish to comment on, one that we did not fully appreciate until we had launched the project, is that efforts to create a central repository of information can in turn interact with the source of the data the project is drawing on. As data activists, we have come to appreciate just how entwined we are with the very processes that we are documenting. Early in the project, we managed to operate in the shadows of media reporting, centralizing data that journalists had obtained directly from official sources and freedom of information requests. As time went on, however, and our project became more publicly known, many media outlets began treating us as a direct source of information on the policing of COVID-19, as if we were a kind of unofficial open government portal working on behalf of government agencies. Journalists now regularly email us requesting the “latest numbers” in their region, when in reality we have the exact same question for them! In response to this unintended consequence of our project, we have begun to collect more information directly from the source (e.g., from police agencies and municipal/provincial governments). Doing so, however, has come at the expense of completeness, as there is only so much our small team can do.
Lesson 3: You cannot control how the data are used once they are out there
When we launched PPMP, we decided that the raw data should be shared publicly, allowing anyone to use it as they saw fit. This has not, however, been easy, as doing so means that some people will use the data for purposes you may not like or agree with. For example, early in the pandemic, findings from our project were shared by a number of libertarian twitter accounts, and we were even invited onto an alt-right radio talk show (which we declined). While some unforeseen uses of the data have taught us a great deal, informing our own policy ideas and perspectives in ways that we are hugely grateful for, others have simply made us wish that we had never released the data in the first place.
There is, however, no way around this, and nor should there be. Releasing the data for anyone to use means that people are going to use the data for reasons that you may not agree with. Although it depends on the project, the kind of data one is dealing with, and so on, our position with respect to PPMP has been that we cannot censor or control access to the data solely because someone wishes to use it to support a cause or movement we may personally disagree with. To do so in this particular case would be entirely antithetical to our goals as data activists and would be more akin to precisely the kind of government information control that we are trying to challenge and overcome.
Lesson 4: You might make mistakes, just be transparent!
The data sources that PPMP relies on are frequently incomplete and often contain mistakes. Although we always seek to verify media reports by triangulating them with other sources, this is an imperfect art. Verifying information can include searching for the origin of reports of an enforcement incident, such as government sources. In many cases, however, verification may not be immediately possible based on available information. The process of verification thus happens over time, especially as more reports come out about a particular incident, law enforcement agency, etc. Balancing the need to verify information with the need to release information in semi-real time has meant that sometimes not everything in our database is factually correct. We have done our best to communicate this to the public in all of our reports, scholarly publications, and social media communications.
Those on the consuming end of PPMP have also done a great deal to help us verify and fact check our data. In one case, a police chief in a major city in Ontario wrote to us directly, upset that we had incorrectly attributed a COVID-related fine to his police force when it was in fact handed out by a city bylaw officer. We fixed the data entry issue immediately and thanked them for pointing this out to us. Importantly, we have worked hard not to become defensive in these situations, viewing them instead as opportunities for further verification and quality control by people outside of our immediate team.
Acknowledging that mistakes are inevitable and correcting them whenever they are pointed out to us has helped to bolster the rigour of our data collection efforts through the crowdsourcing of quality control. Being upfront and transparent about potential errors and being responsive to those who point them out is a major aspect of good data activism.
Despite the many challenges and limitations, PPMP has been an extremely rewarding and – we hope – useful intervention in the public good. In launching PPMP, we wanted to offer a useful source of information that could act as a supplement or counterpoint to other kinds of information and analysis, such as qualitative and historical analysis into the policing of COVID-19 in Canada. As “guerilla archivists”, we wanted to begin the work of collecting this information so as to retain a durable record that scholarly and non-scholarly communities alike could analyze in the years and decades to come. PPMP is about both scrutinizing and documenting the current moment, both for the now and for the future. We have learned a great deal over the past twelve months and, as the project goes on, will no doubt learn a great deal more – about people, about laws and social control, and about data and data activism. We hope that the lessons we have reflected on here are of value to current and future data activists in Canada and the Active History community more broadly.
Alex Luscombe (@alexlusco) is a PhD Candidate in the University of Toronto’s Centre for Criminology & Sociolegal Studies. Alexander McClelland (@alexmcclelland) is an Assistant Professor in Carleton University’s Institute of Criminology & Criminal Justice.