Methods reflect values: Evaluating the shortcomings of the average for measuring population well-being.

Abstract

As governments and institutions embrace subjective well-being as a policy outcome, aggregating well-being in a population has become commonplace. The default method used to aggregate population well-being is taking the arithmetic mean (average). However, using average well-being as a key performance indicator, while useful, can omit morally relevant information, like the extent of suffering and inequality. We examine three alternative methods for aggregating life satisfaction, grounded in the ethical theories of: prioritarianism (a weighted average that prioritizes improvements at the bottom of the scale), sufficientarianism (the proportion of respondents answering above a "suffering" threshold), and egalitarianism (the degree of inequality) and compare them to the average. Toward this end, we used nationally representative data from 3,035,971 participants across 148 countries drawn from the 2005 to 2022 Gallup World Poll and the 1981-2021 World Values Survey. We found that the distribution of life satisfaction deviated significantly from a normal distribution in all countries, suggesting that using the mean and standard deviation cannot adequately capture the full distribution. After re-ranking countries according to the degree of life satisfaction inequality, we found that 56 countries deviated by at least 20 ranks compared to their average life satisfaction rankings. Finally, we observed that 9%-46% of the time, increases in average well-being at the country level were accompanied by increasing suffering and inequality. Our findings show the downside of using the average and offer alternatives that are aligned with promoting equitable well-being growth. (PsycInfo Database Record (c) 2025 APA, all rights reserved).