So did you apparently. That's turnout, not demographics. Assuming same turnout in 2012 as 2008 is pretty dubious.
Still not getting it.
Even if turnout is lower, the raw numbers would still be the same or higher due to a much larger demographic.
And again, we're comparing 2000 to 2016. The 2000 election is infamous for historically low turnout, especially the youth bracket.