This study contributes to the literature on multidimensional poverty measurement by providing the first empirical application of the social welfare method for poverty aggregation using real-world data. We build on participatory research on the “hidden dimensions of poverty”, which engaged individuals with lived experience of poverty to identify dimensions of deprivation often overlooked in conventional frameworks–particularly social isolation. Using data from the Statistics on Resources and Living Conditions survey conducted by the French National Institute of Statistics and Economic Studies (Insee), we quantify several non-monetary poverty dimensions and estimate individuals’ preferences through life satisfaction regressions. Our findings show limited overlap between income-based and multidimensional measures of poverty, underscoring the value of a broader approach. Moreover, the social welfare method identifies a subgroup of poor individuals–typically older, urban, economically inactive, and in poor health–not captured by income-based or equal-weighted multidimensional indices. These individuals report significantly lower life satisfaction, despite higher income levels, suggesting that monetary resources alone fail to reflect their living conditions. Overall, this study supports the development of more inclusive and preference-sensitive poverty measures to better identify and address the multiple forms of deprivation experienced by vulnerable populations.
Previous studies show that the use of different data sources and methods can lead to substantial variation in estimates of income inequality. This study assesses how well sample-based inequality estimates represent income variation among the whole population and what the key reasons are for the discrepancies. We draw on a unique dataset for Estonia providing income information from administrative records for the whole population, from which the Estonian SILC is effectively sampled. We show that the value of the Gini coefficient obtained from the SILC sample data significantly differs from the one obtained for the whole population (0.32 vs. 0.39). A comparison of the datasets shows that the tails of the income distribution differ substantially as the mean disposable income of the first decile is 34
Empirical analyses of redistributive preferences have operationalized an individual’s demand for redistribution using a number of different metrics. These have included responses to questions about reducing income differences between rich and the poor, raising taxes on the rich, helping the poor, and the appropriate level of government welfare/safety net spending. Some studies have analyzed one of these, others have examined two or more, separately, or as summative scales. What has not yet been addressed is the fact that though reducing income differences and helping the poor are closely related in the context of redistribution, they are distinct, the former concentrating on narrowing gaps in the shares of total income across classes, the latter on raising the absolute level of material well-being of those at or near the bottom of the income distribution. In this paper, we take advantage of this distinct but related characterization and analyze jointly the public’s support for government reducing income inequality and reducing poverty using a bivariate ordered probit model. Estimation, using data derived from the 2021 General Social Survey, affords the opportunity to assess the extent of American public support for government taking major responsibility for reducing both income inequality and poverty. The major takeaway is that the prospect of amassing widespread public support for government taking major responsibility for reducing both income inequality and poverty in America is fairly dim. On the bright side, the prospect of amassing a majority to support government responsibility for reducing income inequality seems more promising.
We propose a simple and flexible framework to assess relative intergenerational mobility. The approach defines a dynasty as a parent–child pair, measuring achievement by each individual’s rank within their own generational outcome distribution, and mobility by the change in this rank across generations. This measure accommodates both continuous outcomes, such as potential earnings, and discrete or ordinal outcomes, such as education levels. It also allows for dominance characterizations (e.g., the relative progress made by women vs. men) consistent with social preferences over desirable mobility patterns. We apply the framework to Indonesia using long-panel data linking parents observed in 1993 to their children in 2014. Results show that a large share of the population escaped illiteracy—an instance of absolute mobility possibly driven by major education reforms. However, relative educational mobility was regressive, as dynasties from higher socio-economic backgrounds progressed faster. This pattern limited the overall progressivity of relative earnings mobility. Mobility in both education and potential earnings was markedly more favorable to women.
