The CIS recently released a detailed study on welfare usage by immigrant households in the US using Census data. That immigrants are more likely to use welfare programs than natives are hardly constitutes a novel discovery, though it's rarely acknowledged in Congressional and popular media debates on the subject. This could easily be remedied by the institution of a selective immigration system with the well-being of current US citizens rather than of aspiring migrants in mind, but that's not the purpose of the post.
What grabbed my attention is just how predictable welfare usage by country of origin is. Table 4 of the CIS report shows welfare usage rates of immigrant households with children by householder's country of origin. Near the bottom of the list are affluent, functional nations like Great Britain and South Korea. Topping the list are places you don't want to be (for more than a week or two, anyway) like Mexico and the Dominican Republic.
Picking up on an old habit, I looked at mean national IQ of the 20 sending countries listed and welfare usage rates of immigrants from those countries who were residing in the US in 2009. The correlation is an inverse .57 (p = .01). That relationship is tempered by Indian immigrants, who constitute a conspicuous exception to the general rule (only those from Great Britain use fewer welfare programs than they do). Removing India from the regression ups the correlation to .64 (p = .00). The Indian demographic landscape is an incredibly complicated one, but it's clear that Indians living in the US are not representative of the country they were born in.