People can put an exact number on their feelings but we don't know how they do it - yet

People's assessment of their own feelings on a numerical scale has no objective basis but is a better predictor of major life changes such as moving house, leaving a life partner or changing jobs than is a whole basket of social and economic indicators, according to a new study by Warwick economist Professor Andrew Oswald and Dr. Caspar Kaiser of the Wellbeing Research Centre at the University of Oxford. Their work was published on the same day as a giant Office of National Statistics conference on the use of wellbeing data in UK policy making.

In The Scientific Value of Numerical Measures of Human FeelingsLink opens in a new window, published in The Proceedings of the National Academy of Sciences (PNAS), Oswald and Kaiser explore the connection between human feelings and actions, analysing survey data on the feelings and decisions of approximately 700,000 citizens followed over four decades in the United Kingdom, Germany, and Australia.

Oswald explains: "Large amounts of information are collected by governments, health bodies and employers from people on their feelings using a 0 - 10 or a 0 - 100 scale. But there are no objective units of measurement for happiness or job satisfaction. Are these surveys pointless? No - our analysis shows that 'feelings numbers' are remarkably valuable.

"While we cannot check a person's inner feelings in a direct way, we can look at people's subsequent actions and ask what they reveal about a person's state of mind.

"Our study searched for - and found - evidence of a reliable relationship between 'feelings integers' - the number people chose to describe their feelings - and what we called 'get-me-out-of-here actions' - major life changes such as moving house, changing jobs or leaving a life partner."

Oswald and Kaiser's analysis found that the 'single made-up feelings integer' had better predictive power than a combined group of standard economic and social variables, including household income, employment status, number of children, education, and homeownership status.

The analysis also found an inverse relationship between these feelings numbers and subsequent get-me-out-of-here actions in the domains of neighbourhoods, intimate partners, jobs, and hospital visits in the three countries from which data was drawn. The relationship between feelings and actions appears replicable and is close to linear.

Oswald and Kaiser hope that their findings will bridge a gap between economics and psychology in the weight and credibility attached to feelings data.

Oswald concludes: "Our analysis has demonstrated that the numbers humans provide in response to feelings survey questions do have strong predictive power over future actions. Somehow humans are choosing these answers in a systematic way as though they can sense within themselves a reliable, objective numerical scale. How they achieve this is not currently known and deserves further exploration."

3 October 2022

• The scientific value of numerical measures of human feelingsLink opens in a new window. Caspar Kaiser and Andrew J Oswald. PNAS 2022 Vol 119.