J.M. Rohrer, S.E. Wenz, Inappropriate causal assumptions underlie Killingsworth, Kahneman, and Mellers’ conclusions, Proc. Natl. Acad. Sci. U.S.A.121 (46) e2313712121, https://doi.org/10.1073/pnas.2313712121 (2024).
In their commentary on an article by Matthew A. Killingsworth, Daniel Kahneman, and Barbara Mellers, Julia M. Rohrer (Leipzig University) and Sebastian E. Wenz (GESIS) identify two major issues in the study by the late Nobel laureate and his colleagues:
First and foremost, Rohrer and Wenz argue that the causal research question and the causal language (“Can money buy happiness?”, “Do larger incomes make people happier?”) used in the paper are not backed up by the study design. Killingsworth, Kahneman, and Mellers (KKM) cannot identify the causal effects they appear to be after because they use cross-sectional observational data but do not control for any covariates to rule out alternative explanations for the estimated associations of income and happiness.
Secondly, Rohrer and Wenz point out that—in contrast to what KKM suggest in their paper—the type of Quantile Regression used by KKM does not estimate associations or effects for particular subgroups of people, such as the 15% least happy.
While some might consider this an example of academic nitpicking, understanding and avoiding both issues matter since both are likely to cause confusion among researchers, policymakers, and the general public with regard to the relation between income and happiness.