Skip to content
Menu

Working Paper No. 007: The UK public is overconfident

Baby boomers most likely age group to be wrong but think they’re right, finds UK behavioural survey

  • Working paper
  • 22nd Aug 2024

Gen Z and millennials are less likely to be overconfident than baby boomers, according to the first major survey on ‘calibration’ and overconfidence in the UK. 

A nationally representative survey of 2000 adults found that 81% of people are overconfident in their answers to a series of general knowledge questions – that is to say, they answer incorrectly but believe their response to be right – but there are notable differences between generations.

Baby boomers (60-78) are the most overconfident (84%) compared with 75% of Gen Zs (18-27) and 79% of millennials (28-43). The findings suggest many of us age into overconfidence, while younger people are more likely to ‘know what they don’t know.’

Do you think you’re well-calibrated? Find out in five minutes by taking our calibration quiz 

Take the quiz

Calibration is the difference between how confident you are and how correct you are. Being well-calibrated means you’re usually right when you say you’re sure about something. Poorly calibrated people are much more likely to be ‘confidently wrong’ – they get things wrong even when they’re sure they’re right about something. Overconfidence, a key theme of the UK’s Covid inquiry, can have a negative effect on outcomes including in policy, economics and crisis response.

The survey is part of ongoing research into cognitive bias and decision-making at BIT (The Behavioural Insights Team). The team gave 30 statements to the 2,000 adults. For each, the respondents had to say whether they thought it was true or false, and also state how confident they were in their answer – ranging from 50% (i.e. no idea) to 100% confidence. The questions were a mix of easy, moderate and difficult.

We found that 81% of respondents are overconfident – defined as having confidence that is at least 5 percentage points higher than actual correctness. Just 15% of respondents are well-calibrated – i.e. their perception of their own correctness was consistently accurate and rooted in reality. Around 3% were underconfident – people who are often right, but don’t back themselves. 

While age is correlated with overconfidence, the results found no significant association with gender in this group. In terms of political persuasion, right-leaning voters are around 10 percentage points more overconfident than left-of-centre voters, but this may be a function of voter demographics.

People were most overconfident about surprising facts. The survey asked if it’s true or false that Lagos is the capital of Nigeria (false – it’s actually Abuja). Although respondents were 85% confident they knew the right answer to this, only 30% got it right – a huge calibration gap of 55%.

Similarly, in terms of generational confidence – the team asked respondents if Jimmy Donalson is the real name of YouTuber, Mr Beast (it’s true). Among those under 25, 87% answered correctly – and confidence in their correctness was 81%. Understandably Mr Beast is less influential among older people – just 37% of over 55s gave the correct answer, yet 61% were confident in their correctness – a calibration gap of 24%.

Dr Mark Egan, Principal Research Advisor at BIT said:

“Being well-calibrated is a superpower that very few of us have. Well-calibrated people are highly prized in many fields. Their predictions are usually right and they’re cautious about what they don’t know.

Calibration is a simple thing to measure, but most people haven’t heard of it, even though it’s really consequential to our lives. We can probably all think of overconfident figures we’ve seen in politics – people who are full of certainty but often very wrong. For example, the COVID Inquiry repeatedly threw up testimonies about overconfident decision-making in government during the crisis. For most of us, the consequences of poor calibration in our day-to-day lives are minor, but overconfidence in fields like politics, business, and economics can have very significant negative effects.

Trying to be well-calibrated runs counter to what many of us are told and taught and exposed to through popular culture. Forget projecting certainty, it’s actually really good to be unsure and consciously so. Even though we seem to age into greater overconfidence, the good news is there’s solid evidence that it’s possible to become better calibrated. That’s important for teams, businesses, campaigners and any of us who need to make tough calls under pressure.”

Notes to editors

  1. For more information on the analysis or to speak to one of the experts involved, please contact Mark Byrne, BIT Press Office, on  07745 234 909 or mark.byrne@nesta.org.uk. Spokespeople are available for broadcast interviews. 

Calibration training at BIT

We’ve worked with hundreds of leadership teams to improve their understanding of cognitive bias and drive success in their business. Our two-hour sessions for business leaders can be delivered remotely or in person. Groups gain a practical, evidence-based understanding of methods to mitigate biases and improve decision-making. 

Please enquire by contacting training@bi.team

About the Behavioural Insights Team

BIT is a global research and innovation consultancy which combines a deep understanding of human behaviour with evidence-led problem-solving to improve people’s lives. We work with all levels of government, nonprofits and the private sector, applying behaviour science expertise with robust evaluation and data to help clients achieve their goals.

 

Authors