Stefan Dercon and Joseph Shapiro of University of Oxford review recent work on poverty dynamics derived from developing country panel studies which use consumption or income data. They find recent research identifying a considerable number of people moving in and out of poverty. The literature suggests that household endowments, including education and assets such as land or livestock, are important for mobility; so too are community characteristics such as roads and other infrastructure. For urban areas, evidence points to location and access to certain types of jobs as important. The authors also signal health shocks as not only important for short-term downturns in mobility but also for longer lasting effects that keep households trapped at lower levels of wellbeing over extended periods. In revisiting this literature, the authors conclude that many opportunities are missed with the analysis to move from correlations to providing causal evidence. They also call for more attention to underlying problems with the methods and datasets, including difficulties related to measurement error, attrition and limitations in tracing households that no longer reside in the original localities.