In the Hummer article, I really liked how they did the simulation to show how much out-migration and how many deaths would be needed to account for the difference in mortality and mortality ratios. It would be interesting to see a similar simulation applied to the Ro study to see how extreme the difference in urban/rural origin would have to be in order to influence obesity and migration likelihood (since they note the rural/urban issue as a potential limitation of the study).
In another intersection of Hummer and Ro, it would be interesting to see the obesity rates and other health metrics for the populations in Hummer to see if there are noticeable health differences among women that may then account for the differences in neonatal mortality. In Ro, women who were recent migrants to the US had the lowest rates of obesity, so that may contribute at least in part to some of the differences in outcomes in Hummer (maybe).
A last note on Hummer, I was surprised that they excluded certain races. I'm sure they had appropriate rationale for their study, but it would be great to see a follow-up that would include more races/ethnicities.