Poverty rates reveal how many people are poor, but not whether the same people remain poor. We review poverty and income mobility through a framework that links policy questions, time horizons, reference frames and data requirements, and that yields a data-tier-contingent reporting menu and a decision tree. An inventory of 218 economies identifies a qualifying household welfare panel in 88, with recent documented longitudinal fieldwork in 57. Three empirical illustrations then show the opportunities and the limits of the available methods. Published synthetic-panel validations yield small average deviations in selected joint transition shares, but do not establish a general success rate. Among 1,235 sampled subnational regions, about one in five crosses into a higher fixed global welfare class, which describes places rather than people. Adding noise to a Vietnamese panel changes measured entry and exit rates substantially while leaving the share poor in both years comparatively stable. Together the three illustrations distinguish what can be measured directly, what requires modeling assumptions, and what the available evidence cannot establish.