Much has been made about voter turnout the past couple election cycles. In 2004, white evangelicals were seen as catapulting Pres. Bush to a second term while in 2008 young people were viewed as the movers in nominating and electing Pres. Obama to his first term in office. While I do not have official turn out numbers for 2012, just exit polls reported by the media, the US Census provides some very nice graphics on voter turnout from 1968 to 2008. There are several graphs but this post concentrates on age and race (black/white) differences over time.
It shows several interesting facts that might be surprising. First, people just vote much less than in the past. In 1968 and 1972 voter turnout was larger then today. Second, older people vote more than younger people both in 2008 and in past elections. Third, the difference in voter participation among young blacks and whites very small and from 1984 to the present and except for 1988; young blacks voted at the same or higher rates then young whites (note 2012 was probably similar to 2008 but I do not have hard data to show this). The graph below shows that older blacks, 45 and over, continue to have lower participation rates than older white people through 2008 but that has reduced significantly from 1968 (remember both groups still vote a higher rates than younger people). Given the trends there is reason to think this reduction will continue in the future.
Midterm elections are a different story, African American participation consistently drops below white participation by a few ‘more’ percentage points. Nationally midterms have always dropped in participation rates for all voters compared with the nearest presidential elections.
Voter participation is a difficult calculation as eligible voters must be over 18, not disqualified, and a citizen. To calculate this information precinct data is reported, along with surveys, and population estimates. We do know in 2012 blacks made up about 13 percent of all voters, Hispanics 10 percent, and Asian 3 percent. Exit polls themselves will not tell you voter participation rates with a high degree of accuracy or broken down by several categories like race and age, although estimates are possible. Also data is lacking for geographic breakdowns.