This dissertation lies at the intersection of regional development economics, demography and applied statistics. It aims to answer the broad question of how development interacts with space. The first study proposes an expansion of multidimensional poverty indices, by accounting for ordinal properties of its inputs, to better understand its depth across space. The second study examines how development influence out-migration likelihoods —the so-called migration hump— at a national scale using spatial interaction models. Lastly, the third document is concerned with the prediction for small area estimations of poverty in lieu of sparse or absent data, a common issue in developing settings.