Sunday, July 08, 2018

New Paper Posted: "Data Driven Governance: Building Data Driven Accountability Based Regulatory Systems in the West and Social Credit Regimes in China"



I have just posted a draft manuscript for comment.  Entitled (for the moment) Data Driven Governance: Building Data Driven Accountability Based Regulatory Systems in the West and Social Credit Regimes in China, the draft has two principal objects.  The first is to consider the parallel efforts of both Western states and China to develop data driven accountability fueled governance systems. The second is to suggest the scope of the challenges that such system construction will likely encounter.

The development of data driven governance has provoked substantial angst and uncertainty everywhere.  There is good reason for this angst, but perhaps not for the usual reasons conventionally advanced. Data driven governance systems (including the quite ambitious project of Chinese social credit) grounded in accountability and managed through the self-reflexive operations of an analytics that incorporates social, economic, political or religious objectives through algorithm represents a new form of governance, with its own language, its own structures, and its own ecologies. It exists still within traditional systems of law and regulation and was originally understood as a technique for the implementation of the policies and objectives of those systems.  Those traditional systems have developed their own language, modalities, ideologies and structures within which the integrity of the system can be maintained.  Yet in this "new era" of governance, data driven governance already exhibits signs of producing its own language, its own structures and its own modalities for enhancing and protecting system integrity within ideological parameters in the context of  which the traditional language and forms of constitutional political government operated through complex bureaucracies intertwined with judiciaries and popular representative organs may n longer be particularly relevant.

This "new era" of governance thus not not necessarily call for yet more efforts to "tame" data driven governance within the cage of traditional government and its structures and methods of operation.  Instead it may require the development of new sensibilities, new interpretive language, and the recognition of new classes of system operators whose injection into the process of governance may profoundly affect the way societies understand and engage with governance organs. This trend may be understood (and encouraged) within those organizations at the vanguard of these changes (within vanguard enterprises in the West (and public security apparatus) and embedded within certain organs of collective organization in China). Yet among those deeply embedded within conventional governance-power systems it has produced resistance or efforts at domestication, which pepper scholarly journals and the regulatory efforts of state and international organs.  Yet rather than or in addition to resistance and domestication, it may now be time to turn to the business of building principles of Demokratie, Sozialstaat, Bundesstaat und Rechtsstaat into and through the language of data and data analytics to ensure that algorithmic governance, like that of the law-regulatory systems that preceded it, will operate under appropriate ideological constraint. And if the politician, the lawyer, and the bureaucrat will not engage in these projects, then it is likely that the engineer, the econometrician and manager may. Power relations will not be the same thereafter.

The Abstract and Introduction follow. The draft may be accessed HERE. Comments and reactions most welcome!

Larry Catá Backer[1]

Abstract:  Data driven governance systems are transforming the regulatory landscape of both states and other governance institutions. Grounded in principles of accountability and embedded in incentive based systems for reducing risk and managing behaviors through mechanisms of choice and markets, these governance systems may well reshape the way states and other governance organs are constituted and operate. This short essay has two objectives.  The first is to examine the challenges that social credit, ratings or assessment systems pose for effective implementation. Social Credit itself refers generally to a new mode of data driven governance through which data analytics are used to create and operate algorithms that provide a basis for rewards and punishment for targeted behaviors. More specifically, social credit references the specific project of the Chinese state to create a comprehensive legal and regulatory mechanism grounded in data driven metrics that they have named "social credit." To that end, Section II considers first the difficulties of separating the role of social credit as a set of techniques and as a means of advancing ideological principles and objectives, in the context of Chinese efforts.  Section III then examines some of the ways in which Western efforts at social credit institutions have sought to meet similar challenges. The section first explores the context of social credit systems in the West, and its operationalization, principally in the private sphere and through the use of market mechanisms for behavior management. It then examines the way that social credit might be used in the West as a technique of governance and as a means of embedding international standards in domestic behavior. The essay concludes by suggesting that social credit represents the expression of new forms of governance that are possible only through the correct utilization of big data management.  The shift in regulatory forms also point to significant shifts in the relationship between law, the state and government.  Accountability regimes grounded in behavior standards enforced through data driven analytics may well change the focus of public law from constitution and rule of law to analytics and algorithm.


I. Introduction

                  About a decade ago, when the attention of influential thinking about governance was occupied elsewhere (Napolitano, 2011; Reyes, 2013), one might have noted a curious development in the nature of the forms of governance and its objectives within Western liberal democracies.

Surveillance has morphed from an incident of governance to the basis of governance itself. It is both government (apparatus) and governmentality (its self-conception and complicity, the prisoner becomes his own keeper). In this sense, surveillance has become the new regulatory mechanism. And law is becoming its servant. And the state, either as the traditionally conceived apex of political order, or as the repository of large aggregations of power within an international state system, now serves as a (but not the) nexus point for the regulatory power of technique. It is in this sense that we can speak of the “death” of the “state” or the “rise” of a transnational political system, or the “death” of the public/private divide or even the construction of non-public autopoietic systems. (Backer, 2008).

These changes, one might think, had the potential to change significantly the relationship of the state to law, and of the character and role of law in the governing of states. Yet an initial consideration might have dismissed this trend as irrelevant to the development of the productive force of law and its system. The phenomenon wasn't law; it had been the object of an abstract and remote elite political philosophy since the 1970s (Foucault, 1997); and it appeared most valuable for the extent to which one could pronounce this area "eccentric" rather than for any value where it counted--for tangible value for academics concerned about the collective intellectual movements of their field. Indeed, “it is debated whether this increase in scholarly attention for governance (purely) mirrors a rise in governance as a social phenomenon or (merely) indicates it is a fashionable research topic.” (Mascini & Erp, 2014).

Still, changes appeared to signal a new era of management that would fuse the authority of public and private institutions in new and uncharted ways. The trend was especially evident in the governance of behavior traditionally beyond the reach of states—transnational economic activity (Ruggie, 2013).  There was a sense that the appropriate approach to the management of behavior (by states or private institutions) was increasingly centered on the ability of decision makers to deploy data within algorithms to develop finely tuned systems of reward and punishment to manage appropriate behavior, to hold individuals accountable, and to contribute to social development (e.g., Shiller, 1993).  Due diligence and the construction and operation of monitoring systems to provide accountability through standards developed by law (or markets) appeared to produce that blending of public and private—political and economic systems—that might overcome the difficulty of extending law and rule of law beyond the state (UNHRC, 2008; Jayasuriya, 1998). A intuition emerged, especially among scholars, that “corporate human rights needs to be addressed at a variety of jurisdictional levels—national, regional, transnational and international—by a variety of actors—states, international organizations, corporations and NGOs.” (Simons & Macklin, 2014, p. 271). 

That move toward systems of discretionary decision making built on data-algorithm-consequence models would have to further the command of law and the public policies of which law was an expression (an especially potent  idea in the management of human rights impacts of enterprises, e.g., Deva, 2012; Simons & Macklin, 2014, pp. 178-271).   It was management that counted, perhaps more than law, and institutions that served principle through the management of market driven behaviors, not political institutions. Within this context, it appeared increasingly clear that rule of law was moving toward data driven systems implemented through development of compliance practices of individuals and enterprises, and overseen by administrators exercising constrained decision making authority for the public good.  Regulatory governance appeared to push institutions not toward law based government but to accountability based governance (Scott, 2004; Backer, 2019). Accountability refocused government from the state and form law, to regulation, and the metrics required to bring those subject to standards to account. “Decentered approaches to regulation emphasize complexity, fragmentation, interdependencies, and government failures, and suggests the limits of the distinctions between the public and private and between the global and the national.” (Levi-Faur, 2011;  p. 6).  And it also expanded an already quite substantial breadth for regulating—there was nothing beyond the power of accountability, and thus of management, through regulation if useful (Ibid., p. 9 “the expanding part of governance is regulation, that is, rulemaking, monitoring, and enforcement” Ibid., p. 16).

None of this, however, appeared to disturb the supremacy or coherent integrity at the heart of law or in the construction of public rule systems, even as spaces for data driven governance seeped into the regulatory state apparatus. But suddenly all that appeared to change. The trigger was the action by China, which appears to have ascended to the position of principal global driving force in political theory and action, when the Chinese State Council published its 2014 Notice concerning Issuance of the Planning Outline for the Construction of a Social Credit System (2014-2020) (the “State Council 2014 Notice”) (PRC, 2014). It proposed using the technologies of big data and big data management along with the possibilities of artificial and machine learning to develop comprehensive data driven structures for management around algorithms that can produce real time reward-punishment structures for social-legal-economic and other behaviors. This project, a development of ratings and rewards systems, means to unify and integrate systems of monitoring, of transparency and of compliance within the traditional law-administrative regulation construct of state systems, appears to be one of the most innovative and interesting efforts of this decade.  In the process, of course, social credit, or data driven governance and accounting-punishment-reward systems can significantly up-end the now century’s old structures of rule of law by effectively making its structures irrelevant.

                  Social credit can be understood in two senses. First, Social Credit itself references the specific project of the Chinese state to create a comprehensive legal and regulatory mechanism that they have named "social credit." Second, it refers generally to a new mode of governance that recombines law and governance, and the public and private spheres in new and hybrid ways that will likely transform the structures and principles on which legal, governance, and societal regulatory systems are now understood and through which they acquire their legitimacy.  In both senses, the structures of social credit are similar. In each case the system seeks to rate or score or assess the object of regulation through a process that requires the acquisition of specific and relevant data, which is then interpreted through the application of an algorithm to produce an assessment or a score or a measure which can be used to assess compliance with underlying objectives. Those scores than serve to guide the application of legal or administrative decisions—they can trigger rewards or suggest punishment (Backer 2017).

                  The triangular relationship between governmentalization (of both public and private institutional actors with managerial power), the mass of the population (which is its object and now its foundation), and the ‘statistics’ (that both define and serve to manage the mass of the population) is the essence of the problem of transparency in the twenty-first century. (Backer 2013).  At its limit, the enterprise of social credit suggests both the emergence of a new field of law as well as the negation of the privileging of law within economic and political structures. On the one hand, one might be tempted to see in the social credit enterprise a notion of the dissolution of the constitution of law within itself; that is that the structures of legality, and its constitution, will have consumed itself. What will emerge from that self-consumption will be the methods and systems that it had once generated and which had been deployed in the service of the constitutional project—that the success of the constitutional notion will ultimately consume it so that where once there was constitution there will only be mechanics; where once there was principle, there will only be data; and where once there were norms, there will be “statistics.” (Sokhi-Bulley, 2011;  Cf., Desrosieres, 2011). This is bound up in the more fundamental idea of the end of law and the irrelevance of lawyer except as technician of a new system the lawyer no longer controls. On the other hand, the success of social credit may require and indeed may be dependent on the simultaneous development of a law for the digital and data age. That is, in the digital age, society (however constituted) is even more in need of law's nomos and narrative. That nomos and narrative may vary depending on the societal and political context, but it must nevertheless develop alongside the re-constitution of the principles, customs and manners of governance. To understand social credit, one must understand the evolving structures of the relationships, in law and politics, of the relationships between states, its masses, and the institutions through which it operates. In that respect, data driven governance systems are transforming the regulatory landscape of both states and other governance institutions. Grounded in principles of accountability and embedded in incentive based systems for reducing risk and managing behaviors through mechanisms of choice and markets, these governance systems may well reshape the way states and other governance organs are constituted and operate.

                  This short essay has two objectives.  The first is to examine the challenges that social credit, ratings or assessment systems pose for effective implementation. To that end, Section II considers first the difficulties of separating the role of social credit as a set of techniques and as a means of advancing ideological principles and objectives.  In this section, social credit is examined as an aspect of big data management with substantial governance and normative effects.  In that context, a number of issues are identified: social credit as a project of informatics, as systems of control and management, and as a governance mechanism. The section seeks to examine the proposition that to understand the shaping of law today (and soft law as well) one must understand social credit. The implications for the structure of government and for the exercise of social and political leadership might be profound. Section III then examines some of the ways in which Western efforts at social credit institutions have sought to meet similar challenges. The section first explores the context of social credit systems in the West, and its operationalization, principally in the private sphere and through the use of market mechanisms for behavior management. It then examines the way that social credit might be used in the West as a technique of governance and as a means of embedding international standards in domestic behavior.

The essay concludes by suggesting that social credit represents the expression of new forms of governance that are possible only through the correct utilization of big data management.  The extent to which state authorities in China are willing to utilize big data management will shape the form, scope and direction of the governance possibilities inherent in social credit initiatives at the local, provincial and national levels. But not just China; the quite visible move toward social create in the West, albeit in a fragmented and functionally differentiated way among public and private institutions, also point to significant shifts in the relationship between law, the state and government.  Accountability regimes grounded in behavior standards enforced through data driven analytics may well change the focus of public law from constitution and rule of law to analytics and algorithm (cf., Morabito, 2015; Logan, 2010).  In both China and the West, it is likely that a new language will be required to frame these emerging structures of control.


[1] W. Richard and Mary Eshelman Faculty Scholar and Professor of Law and International Affairs, Pennsylvania State University; Board Member, Foundation for Law and International Affairs; Coalition for Peace & Ethics. The ideas in this essay was first presented  at the Conference: The Chinese Social Credit System 2017, held at Shanghai Jiaotong University 23 September 2017.  It draws on and expands the essay produced for that event and published as 测度、评估和奖励:中国和西方建立社会信用体系的挑战?(Cutting-edge measures, assessments, and rewards: The challenge of establishing a social credit system in China and the West?), “互联网金融法律评论(jiflsjtu微信公众平台。前沿栏目·第三季第21篇(总第182篇).  (Shanghai Jiaotong University  “Internet Financial Law Review (jiflsjtu). My thanks to Flora Sapio (University of Naples), Sun Ping and  Tong Zhiwei (华东政法大学英文版 ; East China University of Political Science and Law), and Duoqi Xu, Shanghai Jiao Tong University KoGuan School of Law; Director, Shanghai Jiao Tong University Research Center for Internet Law Innovation for their very helpful comments  on earlier versions of this draft and for their support and engagement in this project.


No comments:

Post a Comment