While there is now something of a consensus in the literature on the economics of happiness that income comparisons to others help determine subjective wellbeing, debate continues over the relative importance of own and reference-group income, in particular in research on the Easterlin paradox. The variety of results in this domain have produced some scepticism regarding happiness analysis, and in particular with respect to the measurement of reference-group income. We here use data from an original Internet survey in Japan to compare the results from happiness regressions to those from hypothetical-choice experiments. The trade-off between own and others' income (showing the importance of absolute and relative income) is similar in these two sets of results. This kind of validation of experienced utility via direct comparison with decision utility remains rare in this literature.