The role of soft law in the digital state: Supporting or eroding public values?

This is part of the blog series 'Preserving Public Values in The Automated State' edited by Malavika Raghavan, Alexandra Sinclair, Giulia Gentile and Orla Lynskey

Malavika Raghavan, PhD researcher, LSE Law School

Embedding digital technologies into public service provision raises many questions for citizens in a “digital” state.  What do we do when the computer says no? How do we find out why the computer said no? And who is responsible for delivering legal entitlements if the system shuts us out? 

These questions reflect administrative justice concerns that arise in the digital state, where people increasingly interact with digital technologies to access welfare support, health care, cross national borders, or even file taxes. However, as revealed by blogs from this series on the DigiPublicValues project, basic expectations that the State should act accountably, transparently and through lawful procedures have been complicated by the adoption of privatised, digitalised systems to deliver such services. These realities invite attention to the rules that apply to such systems: who sets them, how do they evolve, and do they truly prioritise the public interest?  

In response, this blog calls attention to the role of soft law instruments as an important force in shaping the “rules of the game” in public digital systems. 

What is soft law?

 “Soft law” refers to the numerous written instruments produced by the bureaucracy in the course of administration, often used to set out policies and procedures to guide the action of subordinate bureaucrats. Such documents have proliferated in modern bureaucracies around the world, resulting in countless “official” documents such as administrative directions, manuals, guidelines, memoranda, circulars, handbooks, ‘FAQs’ and so on. 

The legal force of these documents is limited. They are administrative instruments that bind officials within the bureaucracy, but do not amount to legal “rules” that give rise to rights and obligations enforceable in courts by third parties. In the past, these instruments have been called “quasi-legislation” to highlight their informal status as “soft law”, in contrast to primary and secondary legislation that constitute “hard law” in the traditional hierarchy of legal norms.  

The distinction between hard and soft law is however proving increasingly hard to maintain in practice, with the rising production of and reliance on such documents in modern bureaucracies.   Soft law documents contain important details about how digital systems operate, their procedures and basis for making decisions, review and complaint mechanisms etc. These are essential legal matters that can be critical information for all those interacting with the state to access public services. Apart from the salience they hold for officials, as  “official” government documents, such soft law instruments can also be seen as authoritative by third parties and individuals who interact with the bureaucracy. They are often the main official documents released with details about the operation and governance of digitalised systems.

Soft law’s role in governing public sector digital systems

As Prof Sanchez-Graells’ post in this series notes, the dominant approach to public sector digitalisation has been to procure information technology (IT) through contractual arrangements. Despite the far-reaching implications of digitalisation reforms and the substantial cost incurred to the exchequer, they have been adopted in most countries without significant parliamentary debate, legislation or public consultation.   The choice to do so belies a simplistic understanding of IT as a “tool” or a matter of implementation that can be left to the administration—rather than as creating substantive legal issues that are deserve wider democratic deliberation. 

As a result, digital systems have often been deployed accompanied merely by administrative instruments as their legal scaffolding, or by secondary legislation (drawing on primary legislation in the relevant public service area, for e.g. social welfare). Examples of this trend abound. Soft law plays a strong role in India’s national direct benefit transfer (DBT) system, that I analyse in my on-going doctoral research. The DBT system consolidates data on welfare payments and coordinates the digital delivery of payments to recipients’ bank accounts. It was created by executive instructions (from bureaucratic bodies like the Prime Minister’s Office) rather than an overarching legislation, and such soft law instruments are also central to the system’s  on-going governance. 

Even in countries where secondary legislation underpins such systems, they often defer the substantive detail about digital technologies, their use, effects and safeguards to soft law instruments released by the bureaucracy. In the UK’s Universal Credit (UC) system, although the welfare entitlement conditions are set out in social security legislation, there is limited detail on digital design, delivery or requirements despite UC being digital-by-default. In the Netherlands, a controversial data-intensive system called the Systemic risk indication (SyRI) was deployed to profile welfare beneficiaries and generate a predictive “score” of their propensity to commit welfare fraud. The legal basis for the system lay in secondary legislation which was found to be unsatisfactory in the legal challenge that ensured (following which use of SyRI was halted).

These instances reflect the need that soft law is meeting in the governance of digital systems, that are constantly changing and dynamic in ways that evade ex-ante legal definition. Although conceptually understood by legal scholars as internal documents that guide officials in their functioning, in practice soft law are also used by officials to coordinate private sector actors who deliver public services, and inform individuals using digital systems to access services. 

Socio-legal implications (and socio-technical) implications 

The growing production of soft law materials is part of a larger phenomenon, that scholars have termed “proceduralism”. Soft law is however gaining outsize influence in the context of digitalised systems given the contractual nature of their procurement, and limited “hard law” governance in the area. In areas where such formal legal regimes are skeletal or non-existent, soft law is often “the rule”. 

This raises broader concerns for democratic accountability. For instance, one concern arises from soft law not being legally enforceable “rules” but non-enforceable administrative directions. This means that where the policies and procedures they set out are not followed, there is limited room for recourse through courts for individuals or non-state third parties. 

Deeper concerns also arise upon consideration of the forces that shape soft law in digitalised systems, and the values that they may channel. As my doctoral research is revealing, soft law documents in digitalised systems often change in tandem with technical changes and documentation—to ensure that administrative procedures and policies are aligned with technical processes. Such changes are often made to accommodate technical updates, improvements and newer technologies into the chain of delivery. Even where other (non-technical) factors drive such change, they must be implemented technically through digital systems and processes. 

This centres the role of technological processes and related documentation in shaping legal change—through soft law. Higher authorities including courts often rely on soft law to inform their (re)interpretation of the law. As officials iterate soft law to manage technological change, there is a need to assess whether the objectives and values informing technology design are compatible with those underpinning public services. More careful “bottom up” analysis is urgently needed to understand whether soft law regimes serve to bolster or erode public values in privatised digital systems.