Damn Statistics


Last night we noted David Sirota’s “race chasm” argument for explaining where Barack Obama has and has not done well this year. In particular, David’s post contains a chart which appears to show Obama doing well in states with black populations of less than 5% and more than 15%. Between 5% and 15% is the ‘chasm’.

Meanwhile, Brendan Nyhan has posted a lengthy critique of his own which also includes copious graphs. Nyhan takes a number of steps to organize the data in more sophisticated ways. He orders the states by their black populations rather than by their national ranking of black populations. And he removes caucuses from the mix to control for whatever distorting effect there may be from this different method of holding the contest. Finally, he looks at the percentage of the white vote for Obama across the different states. And what he finds is not Sirota’s ‘chasm’ but a fairly straightforward “negative linear relationship between Obama’s white support and black population.” In other words, the more black voters in the state, the worse Obama does among white voters.

Looking over the evidence, however, I’m not sure this apparent disagreement is anything other than the same point expressed a different way. What I take to be implicit in Sirota’s argument is that racially polarized voting increases with the size of the black population in a given state. That leaves Obama winning a lot of states with few blacks. But once the black population gets into the high single digits, racialized voting kicks in and Obama then can’t get enough of the white population to win. Only when blacks approach 20% of the population does the black population get large enough to make up for and often overcome the increased white resistance to voting for Obama. (After all, I don’t think Sirota is saying that Mississippi and Louisiana are models of racial harmony.) Of course, these are Democratic primaries, not general elections. And that is the key. Because that means that in most cases the percentage of the black electorate is roughly double what it is in the general election.

Take that all together and Sirota’s numbers look just like Nyhan’s, only pushed through the strainer of a Democratic primary.