The last ten years have seen an upsurge in research on intergenerational earnings and income mobility. A good deal has been learned about this topic, as witnessed in the survey by Solon (1999), the collection of papers published in the 2002 volume of the Journal of Economic Perspectives, as well as the chapters of this book. On the methodological side, empirical researchers are now much better equipped with data and techniques to avoid various biases that plagued earlier attempts to estimate parent–offspring and siblings correlations in long-run earnings. Further, the order of magnitude of father–son elasticities and brother correlations has been established with some degree of confidence in a number of countries. Nonetheless, it is almost an understatement to claim that much more needs to be learnt before analysts can claim to have uncovered the mechanisms that generate these relationships and how various policies could affect them.

There are different opinions about how research on intergenerational mobility should best proceed. A theorist could complain about the lack of a theoretical underpinning in much empirical research published since the early 1990s. A more empirically oriented researcher could argue that natural experiments generating useful exogenous information should be identified and used so that truly causal mechanisms can be identified. Most likely there is much room for both these approaches. The goal of the research reported in this chapter, however, is different and in a sense quite modest.