What principles should inform poverty indices? Insights from a cross-country survey

1,392 undergraduate students from Bolivia, Brazil, France, Italy, Kenya, Laos, Switzerland and the UK are involved in a survey exploring the support for customary poverty measurement principles. Our study allows us to enhance the studies of Amiel and Cowell (Empir Econ 22:571–588, 1997; The distribution of welfare & household production. International perspectives, 1998) in a variety of directions. We find that the support for Weak Monotonicity crucially depends on whether the poor income rises or falls and the (generally low) agreement with Regressive Transfer is even lower if the recipient is lifted out of poverty as a consequence of the transfer. Further, the support for a certain poverty axiom can significantly differ according to the characteristics of the income distributions to be compared, as is the case for Population Replication and Poverty Growth when, respectively, the society to be replicated and the poor person added to society are very poor ones. Our results also suggest a certain concern for the poorest in society and cast doubts on the desirability of continuity at the poverty line. Finally, notable heterogeneity emerges across relevant subgroups. Poverty perceptions are significantly different for students living in high- and low-income countries: the latter more strongly support Weak Monotonicity and Regressive Transfer, unveiling a more pronounced sensitivity to lower levels of poor incomes. Interesting differences, though less marked, are also found between the views of economics students and those of their colleagues from other disciplines.