1. Regulatory policy 2.0

Challenges raised by technological transformation

During COVID-19, society’s reliance on digital solutions has never been greater. Governments have harnessed digital technologies to support the public-health response to COVID-19 worldwide and secure the continuity of their operations and service delivery. These interventions include outbreak monitoring, case identification, contact tracing, evaluation of interventions and communication with the public. More important still, with the pandemic considerably accelerating and reinforcing prior trends, digital technologies have reshaped the way in which we work, keep in touch, go to school and shop for essentials goods. The deterministic influence of digital and emerging technologies on societies will only grow in the years to come (see Chapter 6).

In this context, governments face an increasingly daunting task in administering their responsibilities as regard technological transformations (Box 1.1). Governments will need to foster innovation and accommodate technology-driven disruption while upholding a level of protection for people, businesses and the public interest at large perceived as adequate. They will need to develop agile and future-proof regulatory approaches while at the same time balancing the need to provide stability and predictability for businesses. They will need to enable greater experimentation under regulatory supervision while also reforming regulation and rules to avoid the creation of uneven playing fields and stranding legacy assets.

Against this background, regulation is essential to mitigate the risks of technological transformation while promoting useful innovation, experimentation and entrepreneurship. Yet, while pace of digitalisation and its impacts on society and markets have been widely addressed by the OECD and others, far less attention has been paid to the actual consequences for the rule-making activities of governments.

It is clear that sweeping technological advancements are creating a sea change in today’s regulatory environment. The governance and regulatory challenges raised by innovation can be broken down into four main categories (OECD, 2019[2]):

• Difficulties to cope with the accelerated scaling of digital technologies;

• Erosion of the usual delineation of markets, challenging the scope of regulators’ mandate and remits;

• Enforcement challenges;

• Fragmentation of regulatory frameworks across jurisdictions while most emerging technology generate strong cross-border effects.

Beyond these regulatory challenges, governments also face growing public and political pressure to develop new approaches to innovation, in order to avoid hindering the development of useful innovation while ensuring that the downside risks are effectively managed. There is, for example, an increasing expectation to change the way online platforms are regulated. These (belated) calls for regulation have triggered the emergence of many reform proposals and the introduction of new regulations. Unhelpfully, little attention has been given in turn on how regulatory practices should evolve. Yet, with the benefit of hindsight, it is clear that rushing into regulation and bypassing standard regulatory policy protocols is not an optimal strategy. It entails real risks of regulatory failure, including:

• Unnecessary barriers to the spread of innovations that serve the public interest (e.g. through unnecessarily opaque, incomplete, redundant or overlapping regulatory landscapes) ;

• Failures to mitigate the downside risks that innovation can entail.

The first type of regulatory failure can be difficult to detect but the resulting opportunity costs may loom large for economies and societies, in particular in a context of slow productivity growth. While the second type is often much easier to diagnose, a major concern is that its consequences might not be reversible in some cases (e.g. gene editing), entailing potentially severe societal damage.

Both failures highlight a strong need for governments to proactively engage with innovation and use regulatory policy as an effective tool to navigate the associated challenges and, ultimately, choose the right regulatory (or alternative) approach. The traditional regulatory policy tools provide important opportunities to consider, consult, question and test the approaches that may help achieve general policy objectives. They can support governments in choosing between regulatory and alternative approaches to promote digital innovation while mitigating the risks. Results of recent cases studies on technological transformation highlight that a variety of regulatory approaches have been implemented by governments (OECD, 2021, forthcoming[3]) These range from explicitly preventing the development and adoption of digital technologies, to adopting a “wait and see” approach to discover which perceived risks materialise, or piloting of innovative approaches such as the adoption of fixed-term regulatory exemptions (e.g. regulatory sandboxes) for innovative entrants that maintain overarching regulatory objectives, such as consumer protection.1

Regulatory policy tools are key to help governments identify strengths, weaknesses and needed adaptations of existing laws. While interesting regulatory practices are emerging across countries in this area (Box 1.2), more would need to be done to maximise the opportunities brought about by technological transformation and mitigate the risks of regulatory failure. As regards IRC in particular, much more can be done to co-ordinate government policies and develop effective regulatory action across borders. The regulatory divergences across jurisdictions are, to some extent at least, an illustration of the lack of regulatory co-operation. Such differences should not be downplayed, as they hold the potential to undermine the effectiveness of action, further undermine peoples’ trust in governments and generate barriers to the spread of beneficial innovations. Similarly, the outcomes of a survey launched by the OECD shows a raising awareness among countries of the need to adapt RIA practices to deal with the concerns raised by technological transformation but few initiatives have been effectively launched so far.

Benefits and potential pitfalls in using technology for regulation

Digital transformation offers major opportunities for governments to strengthen their regulatory capacity (see Chapter 6). The technological evolution can offer new innovative approaches to resource-constrained governments and regulators, and to support more effective and efficient rulemaking. For instance, Artificial Intelligence is increasingly being piloted to support reviews and consolidation of regulations, because of its ability to cover massive amounts of legislation and look for links between texts from different sectors, administrative levels, etc.2 Increased capacity to gather, manage and analyse data also means that the evidence base for the regulation-making process can be significantly broadened and strengthened. Risk analysis and risk assessment can be considerably improved at both strategic and operational levels. Technological change can also enable major improvements in regulatory delivery both in terms of improving the rapidity of responses and precision of targeting (data interconnection, Machine Learning etc.) – and to monitor complex, widespread, remote objects and phenomena. Real-time monitoring, e.g. through the use of remote sensors can also enable the effective implementation of performance-based regulation (e.g. in pollution control – see also Chapter 6, section on Outcomes-focused instead of process-focused regulation).

At the same time, new technologies can also present some potential pitfalls, e.g. when increased technological capacity and decreasing costs lead to the impression that, in a given areas, “total control” may be possible (through a combination of real-time, widespread monitoring, use of the Internet of Things, etc.). The question is not only whether this is actually feasible, but whether it is desirable. First, quite obviously, a massive extension of remote surveillance raises major concerns in terms of individual rights and freedoms, and the fact that technology makes the “surveillance state” easier to implement and more “effective” should make democratic governments all the more cautious about engaging too far in such approaches. Second, excessive reliance on increasing the regulatory reach through technology and automation presents serious problems even from the perspective of regulatory effectiveness, and even assuming that concerns about individual rights are set aside (for the sake of argument) or can be effectively addressed.

Indeed, while excessive reliance on remote surveillance can cause serious vulnerabilities, over-enthusiasm for a massive extension of control is also based on fundamentally flawed premises concerning both the “fitness-for-purpose” of remote controls, and their impact on compliance. On the former, it assumes that remote surveillance and data will be adequate, and reliable. In fact, a large number of risks, that regulation aims to address, are very difficult or impossible to control remotely (e.g. many food-safety risks relating to personal hygiene, handling, etc., cannot effectively be remotely monitored). Moreover, while undertaking to extend control as much as possible reflects a vision founded on distrust (i.e. it assumes that there is always a need to control more, i.e. that violations of rules are likely otherwise), the tools of remote control are themselves eminently vulnerable to fraud, hacking etc. In other words, if the extension of control is seen as necessary because of distrust in the actors of the current system (be it economic operators or state employees), then technology-enabled “extended control” is not a logical proposition, since the same possibilities of dissimulation and manipulation exist (or, arguably, are increased).

Thus, technology should not be seen as a substitute to sound regulatory design and delivery, but as a new instrument to implement good regulatory practices in an even more effective way.

Adapting regulatory management tools

Every regulation is inevitably an experiment. Some regulations deliver the desired benefits, some do not and need to be amended or repealed. Regulatory management tools such as RIA, stakeholder engagement and ex post reviews of the stock of regulations provide important opportunities to consider the whole portfolio of potential solutions, analyse their impact, consult with all stakeholders, monitor compliance with and evaluate performance of regulations. They have been playing the central role in the efforts to achieve regulatory outcomes since the very beginning and will remain a crucial part of regulatory policy in the future. However, these tools need to be adapted and their implementation must be improved to help governments navigate the challenges and the opportunities brought by transformative changes and choose the right regulatory (or non-regulatory) approaches in achieving government objectives.

According to the results of the previous iREG surveys, and confirmed by the one carried out in 2020, RIA is still in many countries carried out late in the regulation-making process, in many cases only used to justify the solution that has already been selected. This goes against the main purpose of RIA – a rigorous process that critically examines alternatives to a given policy problem, providing for inherent trade-offs and ultimately highlighting the option that will maximise the benefits for society while minimising costs. The process of impact assessment needs to go hand-in-hand with the process of developing a policy proposal and start at the inception of this process when identifying the problem to be solved and the objectives to be achieved (OECD, 2020[8]). How to achieve this, how to make RIA a firm part of the daily work of civil servants and policy-maker will be one of the main challenges for the upcoming years.

To identify the right option, officials need to take into account the whole portfolio of potential solutions, both regulatory and non-regulatory (OECD, 2020[8]). Administrations need to get rid of the “regulatory reflex” – opting for a regulatory solution wherever there seems to be an issue requiring a government intervention. Guidance and training on the use of alternatives to regulatory tools need to be improved, officials must be motivated to examine alternative solutions wherever this might be beneficial and regulatory oversight bodies need to play a stronger role in promoting the use of such solutions. Using approaches such as outcome-oriented regulations, co-regulations, standard-setting or leaving space for self-regulation has to become more common in order to permit more flexibility in response to the fast-pace of technological changes. International experience and instruments need to be viewed more systematically as a fundamental pillar of domestic rulemaking. For this, learning governments’ approaches to similar challenges in other jurisdictions or at the international level can be a valuable step in identifying the right regulatory or non-regulatory option.

One of the likely effects of the COVID-19 pandemic is that regulators might be operating in an environment with increased uncertainty in the future. The techniques such as problem definition or analysis of impacts might therefore need to be adjusted as well. When defining the problem to be solved, officials will have to take into account that various problems might have significantly different impacts on various groups of stakeholders (e.g. youth, elderly people, women, SMEs) or sectors of the economy (tourism, hospitality). In the phase of evaluating impacts, the uncertainties regarding potential effects of the adopted measures must be taken into account as well. RIA, envisaged as a tool to maximise the benefit/cost ratio, could also represent one tool helping select the solution that will lead to the desired effects and outcomes (OECD, 2020[9]).

Stakeholders might play a crucial role in identifying optimal solutions. Stakeholder consultation is not only about identifying optimal solutions, but also giving citizens the opportunity to weigh in on trade-offs and value preferences to help manoeuvre the new ethical and distributional challenges that arise from new technologies and technologically-driven business models. For this, it is important that stakeholders are systematically consulted and engaged in the regulation-making process from the early stages (see Further shift from public consultations to stakeholder engagement).

Furthermore, once the immediate pressure from the crisis is over, regulations adopted through fast-track procedures will have to be subjected to careful ex post, or post-implementation reviews (PIR) in order to examine their effectiveness. It will be necessary to adapt the ex post review processes to make sure that a significant backlog of regulations to be reviewed is avoided. This would involve prioritisation while making sure that measures adopted as temporary will not become permanent. The use of sunsetting clauses and expiry dates will probably become more frequent.

It is also particularly important that governments possess robust and adequately resourced regulatory oversight bodies, which will play a crucial role in ensuring that better regulation habits do not fall in priority in a time of crisis. They will potentially play an important role in extracting lessons learned and promoting the adoption of innovative approaches to regulatory management; as well as helping prioritise ex post review efforts and ensuring that relevant evidence from implementation is collected and assessed.

Increasing regulatory coherence, aligning regulatory governance with strategic governmental goals

Regulatory coherence refers to the use of good regulatory practices in the process of planning, designing, issuing, implementing, and reviewing regulatory measures in order to facilitate achievement of domestic policy objectives, and in efforts across governments to enhance regulatory co-operation in order to further those objectives and promote international trade and investment, economic growth and employment.

At the national level, regulatory management tools help to ensure that regulations are aligned with the overall political objectives expressed in government’s manifestos, programme statements, coalition agreements, etc. RIA is a crucial tool in achieving regulatory coherence. For this, it is necessary that, in the phase of problem identification and especially defining the objectives of the government interventions, officials take into account the existing political environment and to what extent the examined alternative solutions fit into the more general objectives of the government. Some kind of “light RIA” might be carried out already at the stage of defining government objectives to make sure that they are achievable and, in some cases, not a sub-optimal ways of dealing with a given problem. This aspect should then play an important role along with fitness-for purpose and cost-effective ratio in selecting the efficient solution.

The COVID-19 crisis has put in the spotlight the threats and harms to health and economy, but does not reduce the urgency of addressing environmental threats to global prosperity and well-being. The need to address the climate change and sustainability will thus remain once this crisis abates. As mentioned earlier, regulatory policy is rarely used to its full potential in achieving current priority goals for governments across the globe, such as the Sustainable Development Goals, fighting climate change, aging population or supporting inclusive growth. To change this, it is necessary to focus on tools that have potential in improving policy coherence and helping governments in achieving their goals as well as, drawing from the experience of the COVID-19 crisis for instance, ensure that business costs related to regulation are kept limited. This would include the ones of sustainable development, through optimisation of policy responses to the defined problems, be it social injustice or climate change.

Sound public policies grounded in evidence – and implemented effectively – will be crucial for the achievement of the 2030 Agenda. By fully reflecting SDGs in the regulatory framework, governments can enhance public sector’s capacity to consistently formulate, implement, and monitor policies coherent with the 2030 Agenda for Sustainable Development across sectors. To this end, governments should make use of regulatory management tools at all stages of the regulatory policy cycle. While embedding SDG considerations at the ex ante impact assessment stage is crucial for the development of new legislation positively impacting SDG achievement, conducting retrospective reviews through an SDG lens is equally important to ensure the existing stock of legislation is in line with the goals of the 2030 Agenda. Engaging relevant stakeholders, like NGOs and research institutions, helps to ensure the latest scientific insights regarding SDGs are reflected in policy making.

Crises such as the COVID-19 pandemic require a coherent policy response. This, of course, includes introducing coherent regulatory measures. Such coherence must be achieved among various institutions at the central level of government, across different levels of administration, including between federal, state, regional and municipal levels of government but, nonetheless importantly, also at the international level through IRC (see Chapter 4).

Indeed, the inherently global consequences of the COVID-19 pandemic highlighted that no policy maker could act in isolation, that to be effective, domestic regulators and policy makers had much to learn from foreign authorities and to gain from joining up approaches. Governments therefore made important efforts to share knowledge and experience in designing the right policies in this context, to guarantee the resilience of global value chains especially in essential food or medical products, and to ensure the interoperability of services such as international travel or internet access. At the same time, the urgent needs for co-operation at the heights of the crisis highlighted the importance of embedding IRC systematically in regulatory frameworks to be able to mobilise in time for future emergencies, and to work together towards preventing future crisis (see Chapter 4).

Again, regulatory oversight bodies need to play a more proactive role in assessing whether the quality of draft regulations are coherent with government objectives (see Chapter 3).

Removing regulatory “sludge” and “deadwood” for better recovery

Setting up clear and predictable administrative procedures is important to achieve regulatory objectives. At the same time, procedures can be an important source of administrative burdens (or regulatory “sludge”4) which might be difficult to justify in the times when setting up and operating a bank account might be done remotely without even visiting the bank. Governments should over time review their administrative procedures and, to the extent possible, try to streamline them to make it easier to comply with the obligations set by these procedures. In many countries, the COVID-19 pandemic forced governments to speed up digitalisation of government services.

To make it easier for regulated subjects to comply with administrative procedures, governments should create one-stop shops (OECD, 2020[10]) through which it will be possible to deal with most of the government services, if possible using remote access and ensure the application of the once-only principle (asking for information only once and sharing it between administrative services), aiming at reducing the administrative burden on citizen and businesses. Also, excessive reporting might not be necessary whenever data might be collected, processed and shared among government institutions using modern technologies (OECD, 2019[11]).

The COVID-19 pandemic contributed to the digital transformation of government operations in many aspects. What seemed to be – or was claimed to be – impossible had to be done in a matter of months or even weeks. Public communication and service provision between governments on one side and businesses, citizens and organisations on the other revealed in some cases unjustified complexity of certain administrative procedures but also surfaced the value of “Government as a Platform” ecosystems to help identify user needs and develop new services in response to crises. It has also proven that digital transformation and digital government efforts must in the future go hand-in-hand with the revision of administrative procedures and design of services with a view to their simplification and reduction of unnecessary administrative burdens as government becomes more proactive and user-driven.

In a post-COVID world, it will also be necessary to restart economies that were partially put on standby during lockdown periods. The recovery programmes should include systemic reviews of regulations, especially those affecting businesses. Even during the crisis, governments waived hundreds of regulations and administrative procedures. Governments have, for example lifted rules against restaurant deliveries and all sorts of permit regulations that were preventing people from working while under lockdown. A more systemic review of unnecessarily burdensome regulations should be put on the agenda of all administrations which want their businesses to flourish again and see new ones created (ref to forthcoming Working Paper on competitiveness impacts of regulation). Special regimes for SMEs and/or better analysis of impacts on small and microenterprises and considering exemptions for those companies should also form part of the recovery strategy.5 The new technologies, possibilities in collecting and processing large volumes of data, might make this process more structured and evidence-based. It is however, necessary that administrations clearly define the data they will need for monitoring and reviewing regulations already at the stage of development of regulations.

Why we need to consider behaviours in regulatory governance

Behaviour change is one of the goals of regulation. This can be achieved through trying to promote a certain action (i.e. purchasing healthier food) or outcomes (i.e. a healthier society) or dissuade others (i.e. limiting the rise in health care costs). To achieve this goal, regulatory policies are traditionally derived from highly generalised and powerful deductive models of human behaviour and decision-making (Lunn, 2014[12]). These models often rely on a number of assumptions of human behaviour, such as the “rational actor” model for economic regulation. Policy makers often then use these assumptions of humans’ decision-making to build policy responses.

However, the field of BI has demonstrated that social context and behavioural biases systematically influence people’s abilities to act in predictable ways (OECD, 2019[1]). BI offers a clear methodology for policy makers to analyse policy problems based on lessons derived from the behavioural and social sciences, collecting evidence of which solutions work (and which do not), and applying these findings to improving the outcomes of public policy. These applications have largely been towards changing individual behaviour, often at the implementation stage of policy making to improve non-behaviourally informed policies. New research is exploring BI applied to changing the behaviour of organisations, including both regulators inside government and regulated entities in the market (OECD, 2019[17]); (OECD, 2020[14]).

A new paradigm is emerging called “behavioural public choice” (Lucas and Tasić, 2015[18]); (Viscusi and Gayer, 2015[19]) that raises normative arguments in favour of applying BI to the institutions, processes and tools that form the framework of regulatory governance. The roots for this theory can be found as far back as Niskanen (1971[20]), noting biases in bureaucratic processes, and more recent political economy arguments, such as public choice, that present evidence behind government failures based on failures to rationalise (i.e. (Tullock, Seldon and Lo Brady, 2002[21]), as cited in (Viscusi and Gayer, 2015[19])). What behavioural public choice theory adds is a model whereby psychological biases can be used to analyse government failures and offer different solutions in response.

The core argument for behavioural public choice theory is the realisation that governments are increasingly using behavioural sciences to intervene via policy decisions, while not taking into account that policy makers and regulatory themselves are subject to the same biases and barriers as any other individual (Viscusi and Gayer, 2015[19]). This ultimately impacts the efficiency and effectiveness of institutions, processes and tools. In fact, governments tend to “institutionalise rather than overcome behavioural anomalies” (Viscusi and Gayer, 2015, p. 40[19]), including failures in risk perception and risk assessment that can lead to an over application of the precautionary principles leading to excessive regulation or a failure to realise that regulation is needed at all.

For example, cost-benefit analyses treat losses and gains equally. However, a well-founded axiom of behavioural science is that humans weigh losses more heavily than equivalent gains (Kahneman and Tversky, 1979[22]), which can lead to more risk-adverse behaviour. As Viscusi and Gayer (Viscusi and Gayer, 2015[19]) note, if for instance, the Hippocratic Oath for doctors is “first, do not harm” then this will set the tone for medical regulators to more emphasis on losses than gains, which may prevent some drugs from being approved. In other cases, regulatory agencies may develop “tunnel vision” that result in policy makers sticking with certain processes and tools, resulting in inconsistent and inefficient outcomes.

In short, the inability of individuals within government to rationalise because of behavioural factors is an important source of government failure (Lucas and Tasić, 2015[18]). The behavioural public choice perspective is therefore a call to symmetrically apply behavioural insights to both the regulators and the regulated entities (Thomas, 2019[23]). In other words, since government is created and run by humans, who experience the same biases and barriers as anyone else, there is good reason to look at the ways behavioural science can help government run more effectively.

This suggests a couple different implications for policy making. First, decisions by regulators are, at least some of the time, systematically biased and this can result in sub-optimal policy decisions. Second, more than just biased decision making, it suggests that there are behavioural barriers that prevent regulators from using effectively the tools and processes of regulatory policy making to support better regulatory policy outcomes, such as intention-action gaps or other friction costs. Thus, regulators and the regulatory process can benefit from identifying and reducing behavioural biases and barriers, and supporting follow-through on good intentions. This paradigm gives new entry points for BI to help support better regulatory policy making that moves beyond implementation towards supporting the broader governance of public organisations through change and reform management.

Using BI to support the use of regulatory management tools

Behavioural public choice has important implications for the use of regulatory management tools in promoting better regulation. OECD (2021, forthcoming[16]) research suggests several new approaches to identifying behavioural biases and barriers that affect use of regulatory impact assessment, stakeholder engagement and ex post evaluation, as well as proposes possible behaviourally-informed solutions to overcome them. These can be summarised into four categories, based on OECD (2019[1]):

1. 1. Attention: Help regulators focus on using regulatory management tools effectively. This includes reducing cognitive limitations and myopia through the smart use of reminders and defaults.

2. 2. Belief: Support regulators in updating their belief about the utility of the tools, and therefore their motivation to use them. Implementation intentions can help bridge intention-action gaps, as well as leveraging loss aversion by highlighting the risk of not using the tools.

3. 3. Choice: Assist regulator’s choice about specific ways to use the tools by addressing problems of group think and status quo bias through diversified work teams and deliberation structures that encourage debate.

4. 4. Determination: Encourage regulators to continue using these tools over time by addressing barriers such as perverse social norms and perceived lack of autonomy in decision making. Demonstrating improvement and changes to the tool over time can provide structure while enhancing autonomy.

The above list is non-exhaustive and derived from discussions with the regulatory policy making community in an attempt to place focus on important areas that have been highlighted as pertinent to the use of regulatory management tools. They provide a good starting point for research into practical ways to use BI to improved use of these tools.

More broadly, there is also opportunity to use the BI methodology itself to improve the analysis conducted as part of regulatory management tools. The point of these tools is to gather and use information to better inform regulatory decision making. The BI methodology is a complementary tool that similarly seeks to gather evidence, but does so through an experimental approach. This seeks to “de-bias” the evidence collection, analysis and decision-making process by providing evidence of actual – as opposed to assumed – human behaviour, and use that evidence to make policy.

It can also help uncover “behavioural failures” (Viscusi and Gayer, 2015[19]) that depart from the individual rationality assumed in economic models and can provide additional justification for intervention in certain cases (Congdon, Kling and Mullainathan, 2011[24]). This allows policy makers to consider behaviourally-informed solutions alongside the traditional policy responses in a way that augments and elevates the decision by opening the door to other regulatory and non-regulatory options. The may be especially powerful in the context of ex post review, which can permit more time to conduct robust experimental evaluations of potential behavioural issues that can ultimately drive new design cycles.

Ethical business regulation

A related approach, Ethical Business Regulation (EBR), is becoming an increasingly popular way of achieving better regulatory delivery principles. EBR is a behaviourally-focused tool that seeks to use flexible, outcomes-oriented approaches to delivering regulations. The idea underpinning EBR is that “regulating with rules” is bound to always under-perform, since the primary driver of behaviour inside corporations is not rules or deterrence through enforcement but rather culture (Hodges, 2015[34]). EBR does not exclude the use of rules and sanctions, particularly for businesses and individuals that behave clearly unethically, but acknowledges that rules and deterrence are bound to achieve disappointing results if they are seen as the primary instrument to reach regulatory objectives.

From an EBR perspective, regulators assess the internal culture of firms, and engage in a co-operative approach if this culture is fundamentally sound and conducive to achieving the goals of regulation. This creates positive incentives for firms that are “on the edge” to change their culture, and the regulator’s role is to provide guidance on how to do so. For firms that refuse to follow an ethical path, the regulator reverts to strict enforcement of rules and sanctions. In a sense, EBR reformulates a flexible and responsive approach that has long been used by a number of regulators, but in a more coherent and “culture-focused” framework (see Box 1.5).

Finally, ongoing research by the OECD Secretariat on regulatory delivery institution shows that path dependency is still largely the predominant force that shapes regulatory delivery structures, institutional mandates, resources etc. There has, to date, been relatively few experiments of systematically reviewing and revising these, re-assessing existing structures and resources against current risks and their respective salience. Identifying duplications and conflicts of competence, as well as mapping all resources involved in a given regulatory field, can be important steps towards further improvements in regulatory delivery (see Chapter 6 for more details).

The need for risk-based, professional, flexible regulatory delivery

One of the challenges to both the development of new technologies and the effective regulation of new, technologically-enabled or technologically-transformed economic activities, is the over-reliance on excessively prescriptive and detailed rules. The impossibility of achieving an “optimal” precision of rules has been convincingly demonstrated decades ago, and the unavoidable distance between rules, compliance and results has likewise long been known (Diver, 1983[35]); (Baldwin, 1990[36]); (Ogus, 1994[37]); (Baldwin, 1990[36]). What this means in practice is that rules, because the full extent and characteristics of all possible situations can never be predicted in advance, will always be either “over-prescriptive” (forbid or mandate too much, ruling out some activities that would actually be desirable) or “under-prescriptive” (fail to forbid some harmful activities). This impossibility of “optimal” rules means that risk-based, accountable, professionally-grounded discretion at the regulatory delivery and enforcement stage is unavoidable if one is to achieve the desired regulatory objectives (Box 1.6 and (Wetenschappelijke Raad voor het Regeringsbeleid, 2013[38])7).

Applied specifically to the challenges raised by technological transformation, the impossibility of optimal rules means that may either completely fail to prepare for the new challenges and risks created or increased by these technologies, or on the contrary prohibit a range of new activities and products that could in fact be desirable – or, quite possibly, both at the same time. This is particularly visible in the case of administrative procedures, such as permitting and licensing rules. The emergence of “ride-hailing” apps led to the question of whether they were allowed under applicable rules governing taxis and drivers-for-hire, and how they could or should (or not) apply for a permit, based on existing requirements and processes. Rather, good regulation in such a case would mean asking what the risks from such activities can be, how they can best be addressed, whether the current processes are fit-for-purpose, and what should be changed in order to ensure the best possible public welfare outcomes given a new situation.

Risk-based regulation (see Chapter 6) means analysing risks to understand their roots and mechanisms, assessing the levels of different risks, and managing risks through a variety of measures. This, in turn, means using the risk angle both when developing an drafting rules, when allocating resources to different institutions or sectors, when determining licensing requirements or targeting inspection visits, or when applying sanctions. Risk is a cross-cutting notion that should be used as the foundation of the regulatory system. Moving away from an approach that is tightly focused on the letter of the rules, and attempts to constantly reduce margins of discretion of regulatory staff is necessary to avoid such pitfalls, where regulation leaves society worse off through a combination of barriers to innovation and ineffective management of risks (see Chapter 6). Regulations address a number of different potential harms (bodily, environmental, financial etc.), not all of which are of equal seriousness – in particular, reversibility or its absence creates a key difference. Likewise, regulation addresses many hazards – industrial pollution and explosions, food poisoning, building fires and collapses, marketing fraud, tax evasion etc. Again, not all of these are of the same severity, and the likelihood of each of these actually happening varies greatly. Thus, comparing the level of priority of regulating different, but also different economic sectors or establishments, based on the harm caused, is inherently difficult.

Risk can allow to consider allocation of resources at a strategic level (between different domains such as environmental protection, food safety, state revenue, technical safety etc.), even though this is rarely done – as well as to prioritise regulatory interventions in a given domain, between different economic sectors and establishments, which is a much more frequent practice. In this way, risk can function as a kind of common measurement unit, allowing easy conversion and comparison of the relative “value” of different regulatory interventions in terms of lives saved, environmental impact, economic impact etc. – but this is only possible if a common approach to risk assessment across regulatory domains and sectors exists.

Comparing the relative levels of risk, and deciding on the appropriate type and intensity of regulatory response, requires having gone through risk assessment – i.e. estimating the relative level of different risks in terms of combined probability and severity of harm. To allow full comparison across different regulatory domains, not only should there be a unified approach for risk assessment – but also a method to convert different types of harm. While this is theoretically possible (there exist many approaches in law and economics to estimate the economic value of life, health, the environment etc.), it is rarely done in practice with that level of precision. Most often, comparisons of risk levels are done within a given category of harm – e.g. potential losses of life, or potential financial losses. In any case, regardless of the level and scope to which risk is applied, it is an instrument of comparison, and thus prioritisation.

Finally, while risk prioritisation done solely by sector or type of activity can be a useful first step of improvement in situations where risk assessment is starting “from scratch” and with limited or no data to support the exercise, it is not optimal, and insufficient in the longer run. In advanced economies, and where data needed for risk analysis and prioritisation are available to regulatory delivery authorities, a more differentiated approach to risk assessment and targeting can be expected – e.g. so as to be applied to each business entity or object (facility, establishment) individually, based on inherent characteristics and track record.

Why risk matters: the importance of prioritisation, and proportionality

This should not mean, of course, that regulators and regulatory staff are vested with arbitrary powers, and that economic operators and citizens do not have appropriate protections. The use of discretion should aim at a responsive, compliance-promoting approach, and be grounded in principles of risk-proportionality, accountability and transparency (OECD, 2014[41]) (OECD, 2018[29]) (OECD, 2018[29]). Only if decision-making principles and criteria are clearly communicated can discretion be exercised in a way that safeguards rights and the rule of law.

One way to move forward is to support regulated citizens and business as the bearers of risk in managing the risks they face. This involves working in partnership with all stakeholders able to produce sustained change. Inspectors in Britain’s Health and Safety Executive have long relied on an approach where law is the last resort and they seek to engage with regulated businesses and promote safer practices through a variety of behavioural approaches (i.e. advice, comparisons with others, indication of potential risks and costs, etc.) (Hawkins, 2003[42]). Results show better outcomes (in terms e.g. of fatal and major accidents) than before the change of approach and/or in other sectors not using this new approach to the same extent.8

Harnessing innovation for regulatory delivery

The “delivery” stage of regulation is critical, particularly through smarter enforcement and inspections, which are based on risks and focused on outcomes. As exposed, designing “optimal” rules has always been impossible, and is even more so in a context of rapid change, increasingly complex economic networks etc. – and technology does not change this fundamental point. “Regulation as Code” is a new name, and uses new technology (and increasing computing power), but corresponds to a restrictive view of rules and regulations, which might lead to consider regulatory discretion as an imperfection to be eliminated, and be seen as a way to achieve perfectly predictable and unambiguous rules. In fact, it might not always be feasible nor desirable to have fully “explicit” or “unambiguous” rules, as exposed above (Box 1.6). In many cases some of the ambiguities in existing legislation result from unavoidable political compromises or from the willingness to leave some form of flexibility and discretion to deal with uncertainty (which is inherent when dealing with events and situations which, by definition, have not taken place yet). In addition, transposing or transforming a significant share of rules into code would imply a fundamental shift in regulatory governance, enforcement, and legal proceedings –which, in turn, raises critical questions around the democratic legitimacy to implement and interpret rules outside of clearly determine executive and judicial powers. Outside of narrow, specific cases where this may be suitable (e.g. specific aspects of tax or customs law, of financial regulation, or of eligibility rules for subsidies), such an approach is bound to yield sub-optimal (or even poor) results. Rather, harnessing technological advances to improve regulatory delivery recognises prior good practice, and tries to implement it more broadly and effectively through new technology.

Growing evidence suggests that governments are increasingly incorporating new technologies in a number of ways (see Chapter6, section on Highlights: major initiatives and innovations in risk-based regulation). This involves, in particular, better collection, management, and use of data, so as to assess and respond to risks more effectively, as well as improved use of communication tools. Being a smarter government requires a more forward-thinking approach to the use and integration of information, technology, and innovation in the activities of governing and delivering services (see Box 1.7). Digitising regulatory delivery also raises a number of risks and barriers that should be effectively addressed by governments.

Better communication strategies

One of the key causes of risk-aversion, and of losses of trust in the regulatory system, is disappointment related to excessive expectations. All citizens, consumers, media etc. expect too much of regulation and of the regulatory system. To avoid losses of trust and legitimacy, regulators (and their political supervisors) need to be honest and transparent about the limits of regulation: no regulation or regulatory system can entirely avoid risk, regulators are not all-powerful; safety lies primarily with business operators and consumers, and there are trade-offs between more regulatory stringency and other elements of public welfare (Coglianese, 2012[48]). The solution to this problem is the adoption of comprehensive risk communication strategies, which provide the essential links between risk analysis, risk management and the public. A coherent risk communication strategy may be ensured through the development of a risk analysis process, combined with open dialogue amongst all interested parties, (i.e. actively informing the public about the desired objective while disclosing the associated risks in a transparent fashion). To communicate risk effectively, there is a need to understand the target audiences and the challenges they are likely to face in assessing the risk and acting on it. In the current complex communication environment with a multitude of platforms, communicating risk in a controlled and co-ordinated way may of course be a challenge. This is why it is essential for regulators to steer regulated subjects and the public more generally around various issues that exist to find credible sources of information.10

In addition, regulators tend to assume that citizens have no appetite for acknowledging the existence of risks, and therefore often opt for statements that are principally intended to be reassuring, while supposing that the public is risk averse. Recent research has however shown that citizens are a great deal more relaxed about risks than often supposed, whereas the public is often rightly sceptical when risks are played down.11 Regulators should communicate risk, e.g. by organising all sorts of public consultation that are also designed to provide citizens the opportunity to influence the decision-making process, and to prevent excessive disappointment (some of which may unavoidably happen). Citizens’ participation in the regulatory process, and ensuring that they have a good understanding of risks, could prove decisive in setting their expectations, and, subsequently, to avoiding any overreactions – as overreactions are mostly due to a lack of information from, as well as insufficient honesty and transparency of regulators about the limits of regulation. However, even in this case, merely disseminating information without communicating the complexities and uncertainties of risk may be insufficient to ensure effective risk communication.

Further shift from public consultations to stakeholder engagement

Stakeholders have a right to express their views as part of the process of developing, implementing, and reviewing regulations (OECD, 2017[50]). Their desire and willingness to influence regulation-making will probably grow in the future. At the same time, to get the best available input and to make the stakeholder engagement process inclusive, it will become more and more important for governments to play a more active role in reaching out and engaging with groups of stakeholders that may have been underrepresented so far, whether it concerns ethnic or sexual minorities, underprivileged groups, or micro-businesses (for more OECD work on Open Government, see http://www.oecd.org/gov/open-government/ and, in particular, the Recommendation of the Council on Open Government12).

To make sure that stakeholders provide meaningful input to the regulation-making process, policy makers need to engage with them regularly and sufficiently early. Regular engagement with stakeholders, be it businesses, NGOs, representatives of certain groups of the society (e.g. youth), etc. is indispensable for creating an environment of mutual trust. Providing complete feedback from the consultation process, i.e. how stakeholders’ input is reflected in the regulation or, if not, providing tangible reasons for why it is not the case. Discussion fora where views on the quality and performance of the regulatory framework are regularly exchanged help administrations to understand the needs of the regulated subjects but also to explain the purpose of existing or new regulations. In some cases, these fora provide not just an opportunity for stakeholders to “complain” about the quality of regulations and regulatory burdens but also to jointly look for solutions (see for example European Medicine’s Agency iSPOC system) or, if necessary, for administrations to explain why certain solutions cannot be accepted. This helps support mutual understanding of what the government is trying to achieve through regulations and potentially increase trust of stakeholders in government regulations and therefore increase compliance with regulatory measures and achieving regulatory outcomes.

Administrations need to be aware of who will be affected by regulations and how. Any groups of stakeholders which might be disproportionately affected should be identified and also consulted with. Preferably, stakeholders should be mapped already at the inception of the process, before the regulations are being drafted. It is the government’s responsibility to give all stakeholder groups an equal opportunity to express their views. This might mean actively reaching out to those who might not have the necessary resources for getting engaged or might not be sufficiently informed on the opportunities to be consulted.

Demonstrating good governance of regulatory institutions

To fulfil their functions, regulatory institutions need to make and implement impartial, objective and evidence-based decisions that will provide predictability to the regulatory regime, inspire trust in public institutions and encourage investment. The governance arrangements of regulatory institutions are critical to the delivery of their functions and their performance as well as to (re-)gaining trust of citizens and businesses. The issue of the governance of regulators is thoroughly discussed in Chapter 5.

The fundamental importance of ethics

Rules have inherently limited power to exert change, and seeking to make regulatory systems ever more precise and rigid, and enforcement ever stronger, is no guarantee of positive results in achieving intended outcomes. Flexibility and agility are essential to avoid excessive burden and barriers to activity and innovation, and avoid bureaucratic defensiveness and situations where the unwavering implementation of rules can lead to absurd situation, and harm trust and legitimacy. Just as the implementation of rules needs to be underpinned by transparent criteria (in particular grounded on risk, see Chapter 6) and by strong professional ethics in regulatory bodies – supporting ethical approaches within business operators is likewise essential, and goes beyond the enforcement of formal compliance (Hodges, 2018[51]).