Sandboxes and the role of policy experimentation
Bringing new digitally enabled products and services to market: Sandboxes and the role of policy experimentation
Digital technologies and data have brought incredible convenience and spurred many innovations, but they also pose great challenges for regulators. Digitally enabled and innovative products and business models often differ significantly to those in traditional markets, and in some cases, they do not fit well with existing regulatory frameworks (OECD 2019). For example, digital business models that operate across traditional sectoral boundaries, or those that transform the roles of users (e.g. ‘prosumer’ models), may not easily fit into existing regulatory frameworks.
Where digitally innovative firms face regulatory uncertainty or an absence of regulation, they may be less likely to set up and attract the funding and resources needed to scale up (Pelkmans and Renda 2014, Armstrong et al. 2019). This is compounded by the fact that the risks of future regulation fall disproportionately on the first mover. Digital business models are often completely new to market, and could therefore pose unforeseen risks for consumers and the wider economy (Carney 2017). At the same time, regulatory uncertainty may deter users from adopting a particular digitally enabled service.
Policymakers across the world have recognised the regulatory challenges associated with digital transformation, and have responded in a variety of ways, ranging from ‘wait and see’ and ‘test and learn’, to banning digitally enabled business models outright. Between these two extremes, some policymakers have opted to experiment.
Policy experimentation and regulatory sandboxes
Digital technologies and data can enable a more effective, risk-based approach to regulating digital innovations. One example is the Digital Health Innovation Action Plan from the United States Food and Drug Administration (FDA), which applies a risk-based approach to regulating software-based medical technologies, including mobile medical applications (United States Food and Drug Administration 2018). The analysis of data can also help respond to potential regulatory breaches in real time (OECD 2018a).
Outcome- or performance-based regulation is another form of policy experimentation. This approach requires outcomes or objectives, rather than the means by which they must be achieved, potentially enabling firms the freedom to innovate while remaining within the spirit of the law. Australia, for example, has adopted performance-based guidelines for the use of autonomous vehicles (Australian National Transport Commission 2018).
Regulatory sandboxes are an increasingly popular mechanism of providing room to experiment (Attrey et al. 2020). Regulatory sandboxes are a structured form of regulatory flexibility that enable selected firms to test innovative products or services with minimal regulatory requirements. Regulatory authorities typically administer regulatory sandboxes.
Regulatory sandboxes gained prominence in the financial services sector (‘fintech’), but have since expanded to a wide range of sectors, including transport (drones, autonomous vehicles), energy (smart meters), health (mobile health apps), and the ICT sector itself (5G deployment) (OECD 2018b, Research Centre for Autonomous Road Vehicles 2018, Office for Gas and Electricity Markets 2018, CRC 2020). Cross-cutting policy issues, such as data protection, are also being addressed by sandboxes (ICO 2020). Often, the skills and competences to administer a sandbox differ from those required for traditional regulatory approaches.
Regulatory sandboxes come in many shapes and sizes, but they share common characteristics
Countries approach regulatory sandboxes quite differently. This fragmentation has led some stakeholder groups to call for harmonisation of sandbox criteria to avoid regulatory arbitrage (European Banking Authority 2017). Nevertheless, many emerging regulatory sandbox programmes share some common features.
- Genuine innovation or novelty. Firms often must demonstrate that their business idea is a genuine innovation, for example by using a new technology or the innovative use of an existing technology.
- Identifiable consumer or social benefit. Some sandbox programmes ask applicants to demonstrate how the proposed innovation can lead to consumer benefit (e.g. higher quality or lower prices), or how the business model addresses an otherwise unmet societal need.
- Need and readiness for sandbox testing. Many sandboxes require that firms identify the specific regulation that constrains it, as well as demonstrate that their product or service is ready to be tested in a ‘live’ market or controlled environment.
- Defined time, sectoral or geographic limits. Regulatory sandboxes typically include limitations to the scope of the testing. Such constraints are usually temporal, but can also include sectoral or geographic limits.
- Safeguard mechanisms. Most regulatory sandboxes include safeguards to achieve overarching regulatory objectives, including consumer protection, safety, and data governance.
Sandboxes can bring benefits, but they are not always the best approach
Regulatory sandboxes aim to support the entrance of innovative, often digitally enabled, products and services to market. While an emerging regulatory approach with few systematic evaluations, regulatory authorities and participants point to a range of benefits of sandbox programmes, including:
- More innovative products and services make it to the market
- Better financing options for innovative firms as sandbox acceptance provides a signal to potential investors and
- Greater knowledge-sharing between frontier innovators and regulators.
For firms, regulatory flexibility can enable live-market testing and market entry that would not have otherwise been possible. This can reduce the time to market for innovations, driving consumer benefits and broader spillovers in the marketplace. Reduced regulatory uncertainty and the ability to conduct testing can also help to facilitate financing for innovative firms.
While regulatory sandboxes can bring benefits, a range of potential challenges also emerge:
- Risks can be difficult to predict
- Difficulties scaling up and addressing cross-sectoral innovations and
- Heavier burden on regulators in terms of time, money and skills.
Early or first-to-market innovations are, by definition, untested and their potential risks can be difficult to predict. Further, some digitally enabled innovations such as platform-intermediated lending have not yet been tested in an economic downturn (Carney 2017). While most sandboxes include extensive safeguards, digital innovations can introduce risks, however small or well-managed, to the market.
Further, in their current form, regulatory sandboxes are small-scale programmes targeting relatively few firms that receive personalised advice, assistance, and monitoring. As a result, sandbox programmes put pressure on the time and resources of already over-burdened regulatory authorities. More generally, such programmes often do not scale well, limiting the potential market-wide benefits.
At the same time, digitally enabled innovations that operate across two or more traditional sectoral areas may not fit easily into the remit of regulatory sandboxes. For example, in countries where financial regulations are structured along traditional banking, pension, and insurance sectoral divisions and each has a regulatory sandbox, an emerging and innovative payment services model may not be eligible in any of them. Some countries are trying to overcome this challenge by involving more than one ministry or by developing an overarching strategy for regulatory sandboxes that cuts across sectors and ministries, such as Germany’s Regulatory Sandboxes Strategy (BMWi 2019).
Are sandboxes effective? The million-dollar question
There are some questions as to the efficacy of regulatory sandboxes. This partly stems from their newness. The first sandbox-like initiative was created in 2012, and the term ‘regulatory sandbox’ was not coined until 2015 (Jenik and Lauer 2017). Rigorous, cross-country and cross-sector evaluation of sandbox programmes is sparse, making it difficult to unambiguously determine that they are always the best way to approach rulemaking (Allen 2019).
One analysis of fintech sandboxes, for example, suggests that while sandboxes are one way to enhance communication between regulators and innovative firms, other approaches of structured experimentation include class waivers for eligible products, leniency for testing and piloting, and sandbox umbrellas – i.e. a public sector body supported by stakeholder groups that helps set up a fully licensed development platform, run in the public interest, so as to further innovation (Zetzsche et al. 2017). Others suggest that ‘innovation hubs’ may be more effective in some circumstances (Buckely et. al. 2019). All this points to the fact that while a useful tool, sandboxes are not a panacea, and policy experimentation can – and should – take many forms.
Allen, H (2019), “Regulatory sandboxes”, The George Washington Law Review 87(3): 580-643.
Armstrong, H, C Gorst, and J Rae (2019), “Renewing regulation: ‘Anticipatory regulation’ in an age of disruption”, Nesta.
Attrey, A, M Lesher, and C Lomax (2020), “The role of sandboxes in promoting flexibility and innovation in the digital age“, Going Digital Toolkit Policy Note, No. 2.
Australian National Transport Commission (2018), “Guidelines for trials of automated vehicles in Australia”, Report.
BMWi, (2019), Making Space for Innovation: The Handbook for Regulatory Sandboxes.
Buckely, R, D Arner, R Veidt, and D Zetzsche (2019), “Building fintech ecosystems: Regulatory sandboxes, innovation hubs and beyond”, Working Paper.
Carney, M (2017), “The promise of fintech – Something new under the sun?”, speech at the Bank of England, 25 January.
Comisión de Regulación de Comunicaciones de la República de Colombia (CRC) (2020), “Sandbox Regulatorio para la innovación en servicios de telecomunicaciones”.
European Banking Authority (2017), “Regulatory sandboxes: A proposal to EBA by the Banking Stakeholders Group”.
Information Commissioner’s Office of the United Kingdom (ICO) (2020), “The Guide to the Sandbox”, Report.
Jenik, I, and K Lauer (2017), “Regulatory sandboxes and financial inclusion”, CGAP Working Paper.
OECD (2019), Going Digital: Shaping Policies, Improving Lives, OECD Publishing.
OECD (2018a), OECD Regulatory Policy Outlook 2018, OECD Publishing.
OECD (2018b), Financial Markets, Insurance and Private Pensions: Digitalisation and Finance, OECD Publishing.
Office for Gas and Electricity Markets (2018), “What is a regulatory sandbox?”, Guidance Report.
Pelkmans, J, and A Renda (2014), “How can EU legislation enable and/or disable innovation?”, European Commission Paper.
Research Centre for Autonomous Road Vehicles (2018), “World’s largest test bed for self-driving cars to be opened in Korea”.
United States Food and Drug Administration (2018), “Digital Health Innovation Action Plan”, Report.
Zetzsche, D, R Buckley, D Arner, and J Barberis (2017), “Regulating a revolution: From regulatory sandboxes to smart regulation”, Fordham Journal of Corporate & Financial Law 23(1): 31-103.