Regulators can use simple foresight methods to test today’s decisions against tomorrow’s conditions.
In the 48 days before New Zealand’s Pike River coal mine exploded in November 2010, there were 21 reports of methane at explosive volumes. Twenty-nine men died. The Royal Commission found that the warnings were not heeded. The information existed, but no part of the system put it together or asked what it meant.
In a guest editorial for the Australia and New Zealand School of Government, my colleague Grant Pink and I described four sights a regulator needs: hindsight, insight, foresight, and oversight, or “HIFO.” Hindsight asks what was. Insight asks what is. Foresight asks what if. Oversight asks what matters now, across the whole system. Regulators are usually practiced at hindsight, through inquiries and investigations. Foresight is the sight they most often lack and the one they most often misunderstand.
A common mistake is to treat foresight as forecasting with more detail. Forecasting has its place: Philip Tetlock’s forecasting tournaments at the University of Pennsylvania show that people can learn to forecast well up to a few months into the future. For 2040, there is no reliable record of past events to calculate from, so a probability is little more than a guess. Bridget Hutter, in a discussion paper published by the Penn Program on Regulation, called the expectation that regulators anticipate and control risk “arguably unreasonable.” A regulator cannot predict its way to safety. It can prepare for a range of futures and notice early warnings about which one is arriving.
Noticing warnings starts with knowing where to look. The “Oxford scenario planning approach,” developed by Rafael Ramírez and Angela Wilkinson, separates two environments: An organization’s transactional environment includes the actors it deals with and can influence, and the contextual environment holds the factors it cannot, such as technology, demography, climate, geopolitics and markets. In the Oxford diagram, regulators appear as actors in a firm’s transactional environment. Turn it around. A regulator’s actors are regulated firms and duty holders, industry bodies, other agencies, ministers and the public. Most regulatory attention goes to actors. Most regulatory surprise comes from factors, which reach a regulator through changes in its actors, as when a firm adopts a new technology.
Stress-testing runs today’s regulations or policies, such as a licensing regime or a risk model, through three or four plausible futures. Mark where they fail, and decide what follows. The test is whether what you learn changes what you decide today. After Rolls-Royce adopted a scenarios-based approach in 2016, every investment proposal had to show how it would fare in each scenario or be rejected.
Provocations test the assumptions underneath a decision. The Copenhagen Institute for Futures Studies Seven Provocations method starts by asking a team what it assumes will not change, then pushes on that assumption in different ways. The first test, the “Flip,” inverts the assumption. The “Uncomfortable Truth” names what everyone is avoiding. At Pike River, the operating assumption was that methane was under control. The flip asks: What if it was not, and the readings meant exactly what they said? Some uncomfortable truths are named and still ignored. In 2015, New Zealand businesses were warned that their fragile supply chains were exposed to a global pandemic. Very little happened until COVID-19 arrived.
Gary Klein’s “pre-mortem” is the method I would put in every regulator’s hands. Before a decision is locked in, a team is told to imagine that two years have passed and its planned decision has failed badly. Each person writes down why. Klein cites research finding that imagining an event has already happened increases people’s ability to correctly identify reasons for future outcomes by 30 percent. A pre-mortem takes half an hour, and it makes dissent legitimate. The inspector who doubts the plan is explaining a failure that, in the exercise, has already happened.
Rafael Ramírez and Cynthia Selin argue that productive discomfort is a better test than plausibility. A future that feels obvious teaches little. One too strange gets dismissed. They suggest ranking risk factors by how uncomfortable they make people and how little is known about them, and report that doing so with mining executives produced a very different list of priorities. Discomfort, though, needs care. A recent paper in Futures by Alex Fergnani, Gerard P. Hodgkinson, and Gabriele Rizzo advises agreeing at the outset the terms of a scenario workshop, where a team tests a decision against imagined futures: its purpose, what it asks of people, how outputs will be used and how difficult reactions will be handled.
This matters most where technology outpaces rules. Cary Coglianese and Colton Crum argue that artificial intelligence calls for “leashes, not guardrails”: flexible regulatory instruments that keep people in control. A leash only works if someone watches where things are heading. That is the job of foresight, and the Organisation for Economic Co-operation and Development’s Regulatory Policy Outlook 2025 already urges governments to build this capability at both system and agency level.
Building this capability is a skills question. The United Kingdom’s Core Regulatory Skills Framework gives regulators a shared language for 20 core skills across four levels. Horizon scanning and scenario planning, the core of foresight, are named only once, as a skill for advanced practitioners anticipating the impacts of new technologies. Elsewhere the framework asks regulators to spot trends and anticipate challenges without naming a method. The framework reserves foresight for a few specialists and points it at one factor in the contextual environment. Neither the methane at Pike River nor the pandemic was a new technology.
A better design builds foresight at every level. An inspector names the assumption behind a site visit and asks what would overturn it. A practitioner runs a pre-mortem before a significant decision. An advanced practitioner scans the whole contextual environment and stress-tests settings against scenarios. A leader, and the board whose job is to question that leader, ask which future would find the agency least prepared.
So start small, with the next right move. Choose one decision due this quarter. Write down the assumption it rests on and flip it. Run a pre-mortem. Ask which contextual factor would most change the actors the decision depends on.
Māori have a whakataukī, or proverb, “ka mua, ka muri,” often rendered as “walking backwards into the future.” The past lies in front of us, where we can see it. Hindsight gives a regulator its footing. Foresight is the habit of checking over the shoulder before the next step. Regulatory agencies that do both are harder to surprise.
