The writer is chief executive of Nesta and chair of the Behavioural Insight Team
When Tony Blair was asked whether he had ever felt a sense of imposter syndrome, he looked puzzled. “What’s that?” he began, before declaring that, as prime minister, he never had the time to indulge in self-analysis. For political leaders, doubting your abilities is rarely something to own up to.
Yet, the opposite of imposter syndrome is something even more unflattering. The Dunning-Kruger effect refers to the cognitive bias whereby people who perform badly at a task do not realise their own incompetence. After the Conservative party’s recent experiments in political leadership, perhaps the phenomenon of overconfidence in poor performers should be renamed the Truss-Kwarteng effect?
When the stakes are highest, corporations and governments ought to subject ideas to the most profound scrutiny. But as the Chilcot inquiry into the Iraq war found, crises often bring out the worst in our decision-making processes.
So what can be done to counter overconfidence? The world I know best is Whitehall, so let’s start there, even if the lessons may be more widely applicable.
First, when creating new laws, taxes or spending programmes, leaders should commission a pre-mortem, where the idea is assumed to have failed, and the investigation tries to uncover the causes. Policies can also be tested by “red teams” who war-game how an opponent would defeat or nullify your approach.
Second, whenever an important decision is made, the supporting evidence used to inform it should, in future, be scrutinised by the public spending watchdog, the National Audit Office. Ministers and civil servants often suffer from confirmation bias — they filter out evidence that conflicts with their beliefs. If ministers were forced to publish the evidence supporting key decisions, their incentives would shift.
Third, every decision should be accompanied by a prediction of whether the initiative will achieve its intended result. In hindsight, everything tends to look obvious and intuitive. But by repeatedly being confronted with our optimism bias, we gradually recalibrate and learn to be more humble. Risk registers are written by few people and read by even fewer, so why not open up the forecasting process to people within or outside the organisation, and draw on the wisdom of crowds and superforecasters? The UK government has made a good start on this via its internal forecasting platform Cosmic Bazaar, which has collected tens of thousands of forecasts since 2020.
Finally, as hard as it is, leaders should embrace and communicate uncertainty. Instead of stating a single figure projecting inflation, carbon emissions or migration, why not publish a visualisation explaining the range and probability of different outcomes? When planning the net zero economy, there is no need to assume we can know now the precise mix of technologies to deploy for the next 25 years. Run hard at the changes that are likely under all scenarios, and then invest in a technology race to flush out the winners.
Increasingly, leaders may, of course, turn to artificial intelligence and predictive analytics to dampen the biases of human judgments. But even if the technical solutions exist to help us calibrate our decisions better, the political and media environment will always compel leaders to demonstrate unwavering conviction. Politicians are incentivised to overpromise, so when they inevitably underdeliver, they lose trust and credibility.
The best leaders, however, find ways around this. When conveying the dearth of evidence or uncertainty, they do so emphatically, not tentatively. They commit to doing whatever it takes to figure out a path forward, pledging to achieve results through intensive trial and error rather than a single grand project. They are zealous about ends, agnostic about means.
At his best, former US president Barack Obama embodied this approach, combining passion and perspective — what Max Weber described as the ideal virtues in a politician. We sorely need them in the turbulent Twenties. So here is a new year’s resolution for everyone in the political class: admit what you don’t know and so gain more trust in what you do.
