Fascinating analysis of Obama's supporters and detractors over at Kos. The Essence:

Percentage of naturalized citizens, e.g. immigrants.  Surprisingly, I did not find that Obama performed worse in states with large Latino populations.  Keep in mind that the difference in Obama's vote share with white voters and Latinos is no longer all that great; he's getting about 45% of the former, and 35% of the latter, and even these differences can be explained by the other variables in my model (for example, a relatively small percentage of Latino voters have college degrees).  However, I did find that Obama performed slightly worse in states with a higher percentage of foreign-born, but now naturalized citizens.  This distinction is important, because  neither the Latino population nor the Asian population are monolithic.  New Mexico, for example, has a huge number of Hispanics, but most of them have been here for generations.  This helps to explain how Obama could virtually tie Hillary in New Mexico, in spite of its population being more than 40% Hispanic.  New Jersey, on the other hand, has a rapidly-growing Latino population, and it consists mostly of recent immigrants.  So it is one's immigration experience, and not one's race, that appears to account for Hillary's stronger support with Hispanic and Asian voters. A zero-gen Hispanic voter is somewhat more likely to vote for Hillary -- and perhaps that is intuitive, because many of them either came to this country or became citizens when Bill Clinton was in power. However, I would guess that native-born Hispanics vote for Obama at nearly the same rates as white voters do, accounting for their other demographic characteristics.

A lot of the other conclusions are less surprising. Still, the detail with which they're rendered is a welcome respite from all McAnalysis on the cable news shows.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.