This paper critiques the last decade of research on the effects of high-skill emigration from developing countries, and proposes six new directions for fruitful research. The study singles out a core assumption underlying much of the recent literature, calling it the Lump of Learning model of human capital and development, and describes five ways that research has come to challenge that assumption. It assesses the usefulness of the Lump of Learning model in the face of accumulating evidence. The axioms of the Lump of Learning model have shaped research priorities in this literature, but many of those axioms do not have a clear empirical basis. Future research proceeding from established facts would set different priorities, and would devote more attention to measuring the effects of migration on skilled-migrant households, rigorously estimating human capital externalities, gathering microdata beyond censuses, and carefully considering optimal policy – among others. The recent literature has pursued a series of extensions to the Lump of Learning model. This study urges discarding the Lump of Learning model, pointing toward a new paradigm for research on skilled migration and development.