An issue has popped up a couple of times in the comments section of my 2012 Voter Turnout post so I wanted to be clear about a few things. First, I am most definitely NOT saying John McCain would have won in 2008 had the White voter turned out at 2004 levels. 2008 was Barack Obama’s year and all the votes mentioned in my write up simply would have made the Presidential race marginally closer and likely only flipping Indiana and North Carolina. I was space constrained (this was originally considered for print publication) so I knew my discussion of vote differentials in 2008 could easily be misconstrued that I was implying John McCain would have flipped many or all of those states mentioned. He would not have.
The thrust of my write up addresses demographic assumptions for 2012 and I use those vote differentials to show just how close many states were and how many votes are missing from one segment of the population (White Voters). Those voters are not evenly distributed across all 50 states just as they are not concentrated in the enumerated Battleground states. It is simply an illustration that small changes in registration (which are occurring), increases in enthusiasm (which is showing up in polls) and turnout (which remains to be seen) could have big impacts with the largest segment of voters (White voters). And when you see the vote differentials and how many voters we already know are missing, suddenly the gap between Obama’s 2008 win and today’s contest gets a lot closer.
Ohio is a perfect example. The aggregate vote totals between the two major parties in the two most recent Presidential contest is strikingly similar:
2004: 5,600,929 votes cast
2008: 5,607,879 votes cast
Per CNN elections website:
2004 Bush (2,859,764) — 51%
Kerry (2,741,165) — 49%
2008 Obama (2,933,388) — 52%
McCain (2,674,491) — 47%
In 2004 non-Whites made up 15% of the Ohio vote and in 2008 non-Whites made up 17% of the Ohio vote. But if the aggregate vote totals are the same, that means 2% of White voters in Ohio who voted in 2004 did not vote in 2008. If the White vote in 2004 totaled 4,760,790 (85% of the aggregate) and 2% stayed home, that’s 95,216 voters who are overwhelmingly likely Republican voters. Barack Obama’s entire margin was 258,897. Give 80% of the stay at home vote to Republicans (76,172) and you’re ~30% closer to flipping the state before you flip one 2008 Obama supporter. For discussion purposes, those figures all assume a static White population in Ohio which is true for neither Whites or non-Whites.
When you see President Obama campaigning in states he won by 14% in 2008 like Wisconsin, giving great evidence that as much as 10+ percent of his 2008 vote has flipped, in Ohio alone if Romney flips ~6% of Obama voters when combined with the missing White vote he erases Obama’s entire lead. This is before factoring in a potential for both increased White population, registration and turnout between 2004 and 2012 — the keys to any election day. Democrats rightly counter this also fails to incorporate increases in non-White population, registration and turnout since 2008. All true but Whites still make up 80+% of Ohio voting population so its a demographic where smaller percentage changes have meaningfully greater impact and the enthusiasm argument (i.e. actually showing up at the polls) overwhelmingly favors Republicans based on all polling up through today.