What makes a happy country? A large computational search suggests a new model and mechanism for national satisfaction

To improve human wellbeing, we must understand what drives it. Life satisfaction data have been collected internationally since 2005, but without a rigorous model of what drives these outcomes, it has remained difficult to draw policy lessons. We fill this gap by aggregating life satisfaction data with nation-level data on an unprecedented number of potential drivers (n=5,533). We use the Lasso, a computational method for variable selection, to construct a simple but highly accurate model with low risk of omitted variables or human selection bias. We show this model has higher prediction accuracy than the current world standard from the United Nations, and the variables it selects are more consistent with studies from political science, social psychology, and other fields. It also selects many variables the current world standard does not, including LGBTQ+ acceptance, gender equity, and accessible political power. Unlike the current standard, which suggests GDP determines 40% of a nation’s satisfaction, our model estimates GDP’s independent contribution at 1.5%. We also find few of the variables hypothesized by recent Nobel-awarded economics work. Finally, we show that our model is more consistent with theories of anti-authoritarian psychology than with “institutions” as a driver of national outcomes. These findings allow policy makers and advocates to prioritize higher-impact goals. It also reveals that current actions in countries across the world directly oppose predictors of satisfaction. In the face of severe global crises, including failure to keep global warming below the “red line” of 1.5C, and a rise in authoritarianism, we need far greater clarity about the future we desire as a society, and how to achieve it. As theories compete on the world stage, often violently, our model offers a simple, empirically driven message to cut through the fog: if we are to improve universal human wellbeing, there is no option but to empower individuals who truly desire that improvement.