Data Power Conference

Picture of a conference workshop

<< Back to Programme

Panel Session 2a): Data and Governance (Chair: Clive Norris)

 

Big Data and Canadian Governance: A Qualitative Assessment

Joanna Redden, University of Calgary

In this paper I argue that investigating how big data analysis is being incorporated into government processes requires a qualitative approach to move from mere observations of technical properties and applications to a sociology of big data assemblages (Sassen 2002), one that views data uses as constitutive of an assemblage of actors, institutions, hierarchies, capabilities, and networks (Kitchin 2014). And, crucially, one that places emerging government uses of big data analysis within its wider informational context. I do so by providing an overview of my investigations of government uses of big data in Canada, and my interviews with politicians, civil servants, data consultants, non-profit advocates, and corporate consultants, and also upon policy documents and research reports. Analysts argue that big data analysis should be used to complement other modes of research, however in practice access to alternative modes of information can be limited by political factors. In Canada there has been increasing government use of big data analysis, more social media monitoring, increasing efforts to make more data open to the public, in combination with increasing cuts to significant statistical services such as cutting the long form census, cuts to key information bodies such as the National Council of Welfare, greater control of access to information, limits on journalistic investigation, and barriers to public servants speaking publicly. This context is important because while some sources of information are being eliminated or silenced, others are being pursued which have significant implications for responses to issues such as poverty.

 

Data sovereignty through Representative Data Governance: Addressing Flawed Consumer Choice Policy

Jonathan Obar, University of Ontario Institute of Technology and Michigan State University

In 1927, Walter Lippmann published The Phantom Public, arguing for what he referred to as the ‘fallacy of democracy’. He wrote, “I have not happened to meet anybody, from a President of the United States to a professor of political science, who came anywhere near to embodying the accepted ideal of the sovereign and omnicompetent citizen” (Lippmann, 1927, 11). Beyond the challenges of omnicompetence, Lippmann argued, had we the faculties and the system (how large an Ecclesia?) for enabling millions to realize popular rule, to control all areas of government ranging from the military, to infrastructure, to education and healthcare, none of us would have time for work, family or enjoyment. The realization of this ‘unattainable ideal’ would leave society at a standstill.

Repurposing Lippmann, this paper suggests that current and proposed data privacy legislation derived from OECD Privacy Principles (e.g. efforts in Canada, the EU and the US) advances a flawed consumer choice model that perpetuates a similar ‘unattainable ideal’ – personal data sovereignty. Had we the faculties and the system for enabling every digital citizen the ability to understand and continually manage the evolving data-driven Internet, to control the data being collected, organized, analyzed and sold by every commercial organization, government agency and data broker, to understand and provide informed consent to every privacy policy - would we have time to actually use the Internet? To live? To work? This is the fallacy of personal data sovereignty in a digital universe increasingly defined by big data.

Providing individuals the opportunity to access and control their data is not enough. A plan for personal data sovereignty should express the true possibilities of its subject. If it is true that the fallacy of personal data sovereignty is similar to Lippmann’s ‘unattainable ideal’, then perhaps the imperfect, yet pragmatic solution to the fallacy of democracy may apply – representative governance.

Through a combination of policy and case study analysis, this paper aims to demonstrate the limitations of legislative efforts that favour informed consumer choice models of personal data privacy. A policy analysis of three legislative efforts (drawing from the OECD’s Privacy Principles) favouring an informed consumer choice model is conducted. These efforts include: Canada’s Personal Information Protection and Electronic Documents Act, the EU’s Data Protection Directive, and the U.S. Consumer Privacy Bill of Rights. Three case studies are analyzed to provide justification for the policy critique: Noam Galai’s ‘Stolen Scream’, Max Schrem’s europe-v.facebook.org, and Hunter Moore’s revenge porn business. A discussion of the strategies of various infomediaries, early representative data sovereigns (for example, the U.S. company Lifelock), will follow.

Works Cited

Lippmann, W. (1927; 2009 edition) The Phantom Public. New Jersey: Transaction Publishers.

 

Data Power and the Digital Economy: Actual Potential and Virtual

Jonathan Foster and Angela Lin, University of Sheffield

In this paper we argue that the capacity to produce value as a by-product of the capture, aggregation and analysis of data and by doing so act upon consumers’ actions is a form of data power. This data power has three phases: actual, potential and virtual. Actual data power involves an increase in the capture, aggregation and analysis of data about consumers’ actions e.g. actions prior to and subsequent to a transaction can also be tracked, as well as any other online and mobile activities. Based on an analysis of data about consumers’ actions, potential data power involves an increase in the range of potential actions that corporations can use to structure consumers’ current and future environments e.g. dynamic pricing, recommendations, personalization. Virtual data power involves an increase in the possible types of power that can be brought to bear on consumers’ actions. We also argue that it is the transformation of data power from one phase to another that plays a constitutive role in changing the relations between corporations and consumers in a digital economy. In summary we argue that the capacity to derive value as a by-product of the capture, aggregation and analysis of data increases the ways in which corporations can structure the field of consumers’ actions, thereby making consumers subject to data power. To what extent the emergence of data power makes consumers and the public further subject to capital is one of the further questions to be addressed.

 

Big Data and Power: What's New(s)?

Josh Cowls and Ralph Schroeder (Oxford Internet Institute)

Much social big data is owned or controlled by private entities, whose business models hinge on the utilisation of this resource for profit. At the same time, big data approaches are often characterised as being ‘truer’ or more accurate than traditional research methods such as self-report studies or surveys. In this paper, we will examine the implications of the private ownership of big data and its powerfulness as knowledge in relation to a specific domain: news reporting online. Recent studies have tracked the types of stories a news organisation covers and an audience’s propensity to view and share them, and have discovered meaningful patterns (Boczkowski and Mitchelstein, Bright and Nichols). These studies draw attention to some dangers inherent in access to and control of data (in this case by news producers, but also data analytics providers) and presumptions about its accuracy. We argue that the uses of big data in this case create asymmetries of power: news organizations and experts know about the disjunction between what news people read in the aggregate and what news is published, but the public does not. This disjunction creates a number of threats to a well-functioning public sphere: since governments increasingly rely on measurement of public opinion to make policy, the accuracy of what is on the news agenda is becoming a key battleground. Further, this accuracy may be compromised by a bias towards more quantifiable digital sources. Access to this knowledge and scrutiny of its representativeness therefore needs the urgent attention of research.