Book Review – The Panoptic Sort: A Political Economy of Personal Information, by Oscar H. Gandy, Jr.

The academic field of surveillance studies has (thankfully in my view) become more crowded during the past few years in response to the increasing use of data technologies for social control. In the early 1990s, when some of us (e.g. me) were naively celebrating the liberating potential of the internet, Oscar H. Gandy, Jr. was critically examining earlier incarnations of data systems and practices that contributed to the entrenchment of existing systems of domination and social injustice. First published in 1993, his book The Panoptic Sort was a groundbreaking account of the history and rationalization of surveillance in service of institutional control and corporate profit at the expense of individual privacy and autonomy. In the a second edition, published by Oxford University press in 2021, Gandy updates his original book for the context of today’s increasingly ubiquitous technologies that collect, process, and commodify personal information for instrumental use by corporate interests.

Gandy explains the panoptic sort as “an integrated system that is involved in the identification, classification, assessment, and distribution of individuals to their places in the array of life chances, which are determined in the play of tensions within late capitalism.” It’s about using data and computational processes to sort people into categories used to grant or deny  resources, opportunities, and privileges. Importantly, Gandy emphasizes that technology is instrumental but not determinative. As technologies of the panoptic sort become socially normalized they reinforce existing power structures, but the imperative “resides not in the machines but in the people who use them” (p.51). 

Throughout the book Gandy applies critical theory to assess “the character and consequences of the development and implementation of the panoptic sort” (p.16). The panoptic sort is a form of strategic technical rationalism that feeds on itself, mining past information about individuals to divide them into categories of value and risk. One consequence is the undermining of the public sphere (see: Habermas) as “a ghostly afterimage that will appear in different forms to different individuals according to their profiles” (17). Gandy draws on Max Weber’s understanding of the rationalization of economic production and the “iron cage” of capitalism, where transactions and relationships are guided by measurement, calculation, and instrumentalism, enabling centralized control and bureaucratization. Individual privacy, dignity, and autonomy are irrelevant to the panoptic sort; they are in fact barriers to the desired production of economic value.

Anthony Giddens’ theory of structuration comes into play as people begin to understand the rules in their information environment, and the resources they may have to navigate within it. To the extent that the rules and resources come to be seen as stable and legitimate, they develop into what we think of as “normal” institutions and social systems. An example (not explicitly discussed by Gandy, but representative of the phenomena he describes) would be the website privacy and cookie opt-out notice made nearly universal since adoption of the EU’s General Data Protection Regulation (GDPR). Most of us understand that websites seek to identify our behavior and interests, and thanks to the GDPR and various technical means to limit online surveillance, we now have some resources to limit collection of our personal data. This is a limited form of individual agency, but one that underscores the structures of website tracking in the communicative environment where we now basically live.

Gandy presents an insightful critique of Marxian theory, or more precisely, ongoing critiques of Marxian theory. Citing Dan Schiller’s analysis of the information economy, he argues that the debates over Marx’s labor value theory bring us closer to recognizing the surplus value of “nonproductive labor” in the capitalist exploitation of personal data we generate by merely existing, collapsing the dispute over productive vs. nonproductive labor. And as the panoptic sort divides society by assigning individuals to categories that privilege or limit their life choices, it establishes new social hierarchies that defy Marxian theory concerning class consciousness and collective interest.

Foucault’s panopticism is naturally invoked as a reference point and “powerful metaphorical resource for representing the contemporary technology of segmentation and targeting, which involves surveillance of consumers, their isolation into classes and categories, and their use in market tests that have the character of experiments.” As Kevin Robins and Frank Webster suggest, these technologies are freed from “the architectural constraints of Bentham’s stone and brick prototype.” As a result of new information technologies and practices, our socio-technical environment functions as a panoptic machine with unprecedented efficiency, power, and reach (p.24).

I emphasize Gandy’s explication of his theoretical grounding, because it provides essential support for his observations on the antecedents of the information technology-based panoptic sort, and its evolution to the systems used today. He briefly covers the bureaucratization of content analysis and data reduction present in government information gathering, such as the U.S. Census, and use of government data by private corporations. The book then turns to the “corporate data machine,” from employee selection, drug testing, insurance risk data, financial information, real estate and utility records, entertainment and leisure activities, education and legal records, and very importantly as a point of entry for today’s panoptic sort, everything people do as consumers. 

Collection of a wide range of personal data is of course a long-standing practice by commercial entities. But by the 1990s, advances in technology allowed for the integration of data from many formerly discrete computer systems, and facilitated automated processes of behavior consumer profiling and behavior modeling drawing from hundreds of data sources. These new processes provided corporations with an unprecedented power to identify, classify, and predict consumer behavior, and to target them with persuasive messages tailored to their profiles. They also increased the demand for consumer data as a commodity and the emergence of a data brokerage industry. Gandy provides details on the growth of several key players such as Amex, TRW, and Equifax. 

Gandy then reports on his own research (conducted in 1990 for the 1st edition) into public perceptions and concerns about privacy, using multiple sources including surveys, focus groups, and interviews with 1,250 U.S. adults. He provides an extensive analysis of the results, and while much has changed since 1990, the study serves as a valuable benchmark for perceptions of privacy today. 

The final chapter from the original 1st edition turns to challenges of data protection and the inequalities of power inherent in the panoptic sort. Giddens’ structuration theory is again discussed in relation to the evolution of privacy law, and the individual’s rights in a social system dominated by corporate surveillance that increasingly affects their life choices. Gandy provides a very thoughtful discussion on the meaning of individual autonomy as an ideal that in relationship with others is always contested. “The extent to which an individual conforms to the expectations of others is the extent to which she or he has given up some autonomy,” he writes. “The panoptic sort threatens the autonomy of the individual by increasing the range of activities that may be brought under the watchful eye of significant others” (p.208). Constant surveillance and evaluation also reduce the nurturing of a person’s individuality, a process of experimentation and self-discovery that can depend crucially on access to privacy.

Interestingly, Gandy offers a refreshed take on Warren and Brandeis’s landmark Harvard Law Review article, published in 1890, which discussed the right of privacy as “the right to be let alone.” If personal data is considered exclusively for its economic value, for example as suggested by Richard Posner in a often-cited 1978 Georgia Law Review article, private life under the panoptic sort is untenable, as personal data of the individual has no market power unless collected and processed en masse with the personal data of others. We might well ask if being let alone is even possible under the panoptic sort, and what this could mean for regulatory intervention.

Finally, the 2nd edition of The Panoptic Sort includes Gandy’s new afterword reflecting socio-technical developments in today’s world of social media, Big Data, artificial intelligence, and the Internet of Things. Like Shoshana Zuboff, Gandy calls for mobilizing a collective response of resistance to “this historically unique form of power” (p.278). The challenge for those who would resist the panoptic sort will be developing a policy response that can protect each individual’s right to be let alone, and restore transparency, accountability, and trust in corporate data practices. This might seem like a tall order. But as Nick Couldry and Ulises A. Mejias argue in another recent book, The Costs of Connection, surveillance capitalism will come to dominate more and more of our social and economic lives unless we are successful in the effort to rein it in. Reading The Panoptic Sort would be a good way to start.