The 2012 Current Population Survey November supplement has now been released which enables one to examine voting patterns by different racial and ethnic groups and different age groups.
Youth turnout among those ages 18-24 dropped 7.3 percentage points from 2008 to 2012, to 41.2%. This contravenes a CIRCLE report that preliminary turnout in 2012 was closer to 50%. [These declines have undone a lot of the gains in youth turnout in the 2004 and 2008 presidential elections.]
If one breaks it down by racial origin, among all racial groups, voting rates for 18-24 year olds were down from 2008 to 2012, declining by 7.4% for non-Hispanic whites, 4.6% for Hispanics and 6.7% for blacks. This reverses patterns seen in the 2004 and 2008 presidential elections where the biggest voting increases came from 18-24 year olds. For example, from 2004 to 2008, Hispanic 18-24 year olds and black 18-24 year olds increased their voting rates by 5.8% and 8.3% respectively.
CIRCLE noted earlier that the National Exit Poll from November 2012 estimated that these young people represented 19% of the voters and that President Obama captured 60% of their votes compared to 37% for Romney. CIRCLE pointed out that if young people had stayed home or if Romney had captured half of the youth vote in Ohio, Florida, Pennsylvania and Virginia, he would have won the presidency. The youth vote was also decisive in Obama’s 2008 victory.
Among adults, non-white voter turnout increased in 2012: more of that story here. Blacks voted in 2012 at higher rates than whites for the first time since 1996. Bad news for Republicans is that, according to the CPS, the share of white votes declined in every one of the battleground states in 2012 (other than Iowa). As Nate Silver has noted (on 538 Blog), the share of non-white votes in battleground states makes a difference in changing the election results, unless one assumes that Republicans are suddenly going to start doing a better job of appealing to non-white groups.
Census reports that the overall turnout (youth and adults) was down somewhat in 2012 as well: from 63.8% in 2008 to 61.8% in 2012. Michael McDonald, using different measures that rely not on self-reports but on ballots cast by state finds a similar pattern: 61.6% of those eligible to vote voted in 2008 and that declined to 58.2% in 2012 (a decrease of 3.4 percentage points or a decline of 5.6%).
There was some speculation in November that Hurricane Sandy was to blame. While turnout declined substantially in Eastern states devastated by Hurricane Sandy, turnout also declined in most states nation-wide, making it dubious that one can lay the blame for lower turnout on Sandy.
For more on racial patterns in 2012 turnout and age-based patterns see the Census report on this topic.