The White Female Vote in 2008
What if only white females voted in the 2008 presidential election?
This is the type of question social scientists and individuals like me love to explore, and which everybody else presumably finds quite boring. More fascinating still, there is actually a somewhat reliable answer to the question. This is because, in every state of the union, there are exit polls of the white female vote in 2008.
It turns out that if only white females voted in 2008, Senator John McCain would have won the popular vote 53% to 46%, taking a comfortable eight-point lead.
Senator Barack Obama, however, would be president. He would win a razor-thin, 273 to 265 majority in the electoral college:
This is quite a remarkable result. Mr. Obama loses by eight percentage points amongst white female voters – yet still wins the electoral college and becomes president. Imagine if Senator John McCain lost the real 2008 presidential election by the exact same popular margin and then magically won the electoral college.
This is a graph Nate Silver once compiled of the chances this would happen in the real electorate:
According to the analysis, a four-point margin in the popular vote translates into a one percent chance of losing the electoral college. Notice how the graph does not even go beyond a seven-point popular victory.
So how does Mr. Obama lose so badly amongst white females yet still become president?
Here is the answer:
As it turns out, white female supporters of Mr. McCain are distributed very inefficiently. They are packed in states the Republican is already winning, especially in the racially polarized Deep South – where Mr. McCain does so well it is quite amazing and sad.
The Democratic white female vote, while not as numerous, is far more efficiently placed. Democrats win white females where it matters – in thin but strategically located margins in enough states to win the electoral college.
This fact can be illustrated visually:
The map above constitutes the 2008 white female vote, except this time differentiated by margin of victory. Except in a few parts of New England, Democrats never win white females by margins greater than 20%.
Finally, this analysis also illustrates the continuing racial divide confronting the United States. More than a century after slavery and fifty years after Civil Rights, in too many parts of the country one can tell far too much – about voting habits or other behavior – just by looking at skin color.
P.S. For those interested, here is a full of table of the white female vote in 2008, by each state’s exit polling.