| Abstract: | This paper aims to explore the sensitivity of longevity-adjusted measures of income, which have become increasingly popular as indicators of basic standards of living. For that purpose, longevity-adjusted income measures are computed for post-war Japan under various sets of postulates, concerning the temporal horizon regarded as relevant for the measurement of welfare, the ethical treatment of age structures, the degree of endogeneity of longevity, the value of a statistical life (VSL) used in the calibration of preference parameters, and preference parameters themselves (for a given VSL). Pictures of Japan's development are significantly sensitive to those postulates, suggesting that longevity-adjusted income measures should be computed under not one—as is usually done—but several assumption sets, to account for the difficulty of solving the income/longevity weighting problem. Hence, this study casts new light on the trade-offs raised by the aggregation of economic and demographic achievements into a preferences-based composite indicator. |
