Soft, But for Whom? Law and the Creation of Accountability – UK Constitutional Law Association

*Editors’ note – this post is part of a series on ‘The Rule of Guidance?’. The other posts in the series will be available here.

Five years on, and with a potential new pandemic on the horizon (if not already here), researchers from all disciplines have been very busy evaluating how the SARS-CoV-2 pandemic was managed by different governments around the world. In legal and political studies, for example, there is a growing body of literature which considers the encroachment on civil liberties effected by so-called ‘lockdown’ policies and their overall impact on the rule of law.

Strict lockdowns, implemented through a mix of primary and secondary legislation and enforced by police, have been contrasted with pandemic management policies in which governments mostly limited themselves to providing information and public health advice to the citizenry, thus avoiding the encroachment upon basic liberties. Sweden is often discussed as the primary example of this ‘soft law’ approach to the management of the SARS-CoV-2 pandemic. The discussion of the Swedish case is still ongoing, not only as to the overall effectiveness of such a strategy in reducing deaths and long-term health impacts of the virus, but also as to the underlying factors that made the Swedish authorities prefer soft over hard law instruments.

In the United Kingdom, the use and ‘misuse’ of soft law during the pandemic has been the subject of a growing number of contributions in legal studies. Some of the posts making up this blog series on the ‘The Rule of Guidance?’ provide a valuable addition to this growing body of literature. Known issues about the use of soft law tools include, at least: 1) the mixing of law and guidance, resulting in a high degree of confusion in the general public (and sometimes in law enforcement as well) as to what was required and what was merely advised; 2) the lack of procedural safeguards in the adoption of guidance by the government vis-à-vis legislation (as highlighted by Ghazi in her post); 3) the effects on the separation of powers of the adoption of ‘interpretive guidance’ by the government – that is, guidance in which the government puts forward its preferred interpretation of legislation as ‘the law’.

In this final blog post of the series, I will instead question the legitimacy of the use of soft law tools in domestic administrative law from a different angle than the one already discussed by Himani Bhakuni in her post. I will start by explaining the traditional reasons that establish a prima facie preference for the use of soft law over legislation in regulatory theory. I will then use the example of mask mandates in the UK to explain how the choice of soft law over hard law might not be necessarily freedom-maximising for everyone, as its advocates say. I will then unpack the relationship between law and accountability. I will conclude by arguing that, as a result, the policy alternative between hard and soft law instruments is much more complex than is normally thought.

On the preference for soft law in regulation theory

Where does the preference for the use of soft law as a regulatory tool stem from? A review of the relevant literature suggests two main justificatory routes, embracing either input or output legitimacy. Starting from the latter, there is a wealth of empirical case studies in different disciplines which indicate a higher degree of efficacy of soft law norms compared with legal ones. John Drury mentions some of them in his contribution to this blog series. Soft law, in other words, would tend to work better in regulating behaviour than hard law, at least in some settings. This would be due either to the fact that, in some cases, soft law norms are directly produced (or co-produced) by those to be subjected to them; or, more generally, because the mere fact of recommending rather than mandating would elicit a higher degree of agency in – and thus compliance by – the addressees of the norms (as Drury argues).

In terms of input legitimacy, there are well-known arguments in political and legal philosophy for the preference of soft law over legislation. One of them is framed straightforwardly around the principle of individual autonomy, which I consider a fundamental characteristic of law qua normative institutional practice in my own work. Soft law, it seems intuitive to affirm, is by default always morally preferable to legislation because even only the threat of coercion which accompanies legal norms in case of non-compliance violates individual autonomy (and dignity).

Another well-known argument for preferring soft forms of regulation to legal ones stems from the application of the principle of horizontal subsidiarity, understood as the governing standard of the distribution of powers upon the different units of the social fabric. In a nutshell, while vertical subsidiarity demands that decision-making powers are allocated to the state institution that is closest to those affected as far as possible, horizontal subsidiarity – as embraced by Hayek for instance – requires a fortiori that the state should exercise its regulatory powers only when society (in its different components: trades, associations, families, all the way down to individuals) cannot effectively self-regulate. In other words, we should first allow for self- or even co-regulation: that is, regulation that is at least co-created by the state and its addressees, rather than imposed unilaterally by the former on the latter.

Face masks mandates and lack thereof: freedom, but for whom?

These two, convergent, arguments appear to ground a prima facie preference for soft law regulation over legislation in public policy, at least in some areas. One of them is personal health, where there is an established line of legal, political, and moral thinking – usually developed on the basis of John Stuart Mill’s ‘harm principle’ – cautioning against unilateral interventions by the state into how to deal with own’s own body and mind.

A clear example of this could be seen in the UK government’s approach to face coverings in the first two years of the SAR-CoV-2 pandemic. After an initial reluctance to recommend mask wearing to the public (arguably also due to initial hesitation by the WHO), in May 2020 finally the UK government urged the public to wear face covering if social distancing was not possible. Then, in June 2020, mandatory face mask wearing was introduced and quickly extended to virtually all non-private settings. The mandates were then lifted in July 2021 as part of the easing of lockdown policies, only to be reintroduced again in November 2021 for two months to help contain the rapid spread of the Omicron variant.

The government’s messaging on this particular issue has been widely criticised. What is relevant for our purposes are the two points in time (July 2021 and January 2022) when mandatory mask orders were lifted. One the one hand, this was cause for celebration for many members of society opposed to the imposition of masks and of mandatory measures more generally. Among these, former UKSC judge Lord Sumption has gone on record on multiple occasions to say that the British government should have left it to the personal choice of individuals whether to take any type of precaution or not against the SAR-CoV-2 virus, characterising the hard law measures (including lockdowns and masks mandates) as fundamentally illiberal.

On the other, several of the most vulnerable members of UK society – either individually or through their representative associations – lamented the removal of mask mandates publicly. In particular, they highlighted how, far from representing an increase of their freedom, the removal of mask mandates and corresponding consequential drop in their use constituted the exact opposite. This is because their ability to leave the house safely was going to be severely impacted by the lack of the obligation on others to wear a mask, say, on public transport or in other settings (shops, etc), making public spaces substantially less safe for them (given what we then knew about how the virus spread through aerosols).

Law and the creation of mediated accountability

Whose freedom, then, is protected by governing a public health emergency via soft, and not hard, law? It may well be true generally (as Drury says) that soft law interventions are better observed and more effective in relation to public health outcomes. But the UK’s experience of (legal) mask mandates versus (guidance-encouraged) mask-wearing provides an important counter-example. Mask wearing by the public went significantly up when legal mandates were introduced and significantly dropped when these were removed.

What happened here? Drury, correctly, highlights the signalling or expressive function of legislation, through which the salience of certain behaviours and expectations – either positively or negatively – are crystallised and given institutional strength by the law. In this regard, making mask-wearing obligatory communicated (after contradictory messaging) its importance in slowing down the spread of the virus and helping saving lives – particularly of those more at risk from it. Who wouldn’t want to protect their loved ones from a deadly virus?

But there might be another important element as to why making mask wearing obligatory mattered, which never comes to the surface in the literature. Legal norms, as opposed to informal ones, create ‘mediated’ forms of accountability – that is, ‘accountability that is mediated via an external authority’. What does it mean? As Brennan et al. put it, it means that:

[r]ather than individuals bearing responsibility for holding one another to account, the external authority has a recognized right and responsibility to hold individuals to account on their behalf.

The case of mask wearing mandates (or lack thereof) makes for a perfect illustration. As long as there are mask mandates imposed by law, individuals can simply ask the relevant authority (the police, or whoever else is responsible for enforcing legislation in that particular setting in the first place – the train manager, the hospital security staff, and so forth) to bring non-compliers in line. The two essential elements of accountability – answerability and enforcement – are in the exclusive purview of state authorities, and individuals at risk can (should) count on the state’s machinery (or its proxies) to see the legal mandates implemented and fully complied with.

Things are completely different when mask-wearing is not legally mandated, but merely recommended (if at all). In this situation, individuals – and particularly vulnerable ones – are left by the state to fend for themselves, so to speak. Accountability is unmediated and entirely left to the personal and social dynamics between individuals. This impacts upon both of its constituting parts: answerability – which is not anymore to the external authority, but to each other – and enforcement, which becomes private or self-enforcement. This exposes the weakest members of society to harrowing dilemmas – do I ask that person on the bus to wear a face mask as it is recommended (but not mandated) by the government and risk being on the receiving end of an adverse, maybe even aggressive reaction, or do I simply shut up and accept the increased risk of catching COVID-19?

Conclusion

Sadly, and predictably given what I have just illustrated, the reaction of many clinically vulnerable members of our society to the removal of mask mandates in the UK and the use of soft law instead was to confine themselves to their homes. This is still the case for many of them, almost five years in.

The use and misuse of soft law during the first years of the SARS-CoV-2 pandemic has been rightly criticised on multiple fronts in the UK. In this short contribution, I have endeavoured to highlight a further and neglected aspect that must be considered by policy-makers when deciding whether to use hard or soft law tools to implement policy, particularly in the context of a public health emergency. For limiting public intervention to merely recommending or advising a certain course of action – wearing a mask, say – does not necessarily increase the freedom of all, as advocates of the ‘minimal state’ hold.

In this regard, existing discussions, in both the theory and practice of regulation, about the pros and cons of using legislation to govern behaviour focus almost exclusively on (the threat of) sanctions and on its effect on individual liberty. But law routinely performs a host of further functions which cannot be reduced to the modality of commanding. One of these further functions is that of modifying the way in which we hold each other accountable. Law creates a system of inter-personal accountability (and relative enforcement) that is mediated by an external authority and which offers (or should offer anyway) protection to the weakest members of society. Therefore, the decision whether to use hard or soft law tools in public policy is much more complex – and with more significant trade-offs – than it is commonly appreciated, and it should be approached as such.

My gratitude goes to Isra Black, Paul Scott and to a colleague who wishes to remain anonymous for very helpful comments on a previous version of this post.

Paolo Sandro is Associate Professor of Public Law and Legal Theory at the University of Leeds. He is currently an Academic Parliamentary Fellow in the Parliament, Public Administration and Constitution Hub in the House of Commons.

(Suggested citation: P. Sandro, ‘Soft, But for Whom? Law and the Creation of Accountability’, U.K. Const. L. Blog (12th February 2025) (available at https://ukconstitutionallaw.org/))