Emigration first increases before decreasing with economic development. This bell-shaped relationship between emigration and development was first hypothesized by the theory of the mobility transition (Zelinsky, 1971). Although several mechanisms have been proposed to explain the upward segment of the curve (the most common being the existence of financial constraints), they have not been examined in a systematic way. In this paper, we develop a novel migration accounting methodology and use it to quantify the main drivers of the mobility transition curve. Our analysis distinguishes between migration aspirations and realization rates of college-educated and less educated individuals at the bilateral level. Between one-third and one-half of the slope of the increasing segment is due to the changing skill composition of working-age populations, and another third is due to changing network size. The microeconomic channel (including financial incentives and constraints) only accounts for one fourth of the total effect in low-income countries, and for less than one fifth in lower-middle-income countries. Finally, our methodology sheds light on the microfoundations of migration decisions.