Yes he did. Did you even read the article?
To DrScholl: The voters that were between 21-24 years of age voted heavily for Obama (over 70% I believe), but those that were between the ages of 18-20 voted comfortably for Romney (somewhere in the high 50's), so that's why the 18-24 group voted 60% for Obama overall. Why is this so hard to understand?
That would be the lowest turnout portion of the group, which means pollsters are less likely to get an entirely accurate sample. It makes little sense for the 18-20 to have voted so much more Republican than 21-24, since it goes against the trend of the overall exit poll by age.