Greenberg is double counting the gaps, while I present them as just the difference between Romney's percent among one group and his percent among another. But the message is similar. The marriage gap has been bigger than the gender gap for a half dozen consecutive elections (at least):
President Obama won women by a 55 to 44 percent margin and lost men by a 45 to 52 percent margin, leaving an 18-point gender gap, up from 12 points in 2008. But this difference is dwarfed by the marriage gap—the margin between married and unmarried women. Married women supported the Republican candidate in 2012 by a comfortable six-point margin. It is Obama’s huge victory among unmarried women that delivered the women’s vote and with it, the White House. There is a 43-point difference in the margin between married women and unmarried women, a number which exceeds the gender gap by a factor of two.
This marriage gap is not new. Marital status has been driving how voters vote—and, whether they vote—for several election cycles. What is different about this cycle is that the Obama campaign was the first national campaign to recognize the importance of unmarried women and target advertising and messaging explicitly to these voters. This proved a smart investment.
In 2010, Democratic support among unmarried women dropped significantly, and Democrats lost the overall vote among women and the Congress. In 2012, Democratic support among unmarried women bounced back sharply.