National exit polls have overstated the Hispanic share of the vote at least since 2000

Hispanics didn’t account for 9% of the 2008 electorate, but 7.4%. And in 2004, they weren’t 8%, but 6.0%. Way back in 2000, the exit poll claimed that Hispanics made up 7%, but the real number turned out to be 5.4%. On the other hand, from 2004 to 2008 the number of black voters has grown 15%. Blacks added another 10% to their vote total from 2008 to 2012. Nationally, 66.2% of eligible blacks voted compared to 64.1% of whites, 48.0% of Hispanics and 47.3% of Asians. In the crucial battleground state of Ohio, blacks achieved a voting rate of 71.7% compared to 61.9% for whites. Overall, the raw black vote total grew more than even the Hispanic vote from 2008 to 2012: an incremental 1.68 million for blacks versus 1.44 million for Hispanics, and a decline of 2.00 million for whites. Among blacks, older women were the most diligent at increasing their turnout in 2012. This growth in black turnout was particularly concentrated among those over age 65. Also, black women traditionally vote at significantly higher rates than black men and the black gender gap in turnout hit a new record in 2012. In contrast to the fervent black effort to re-elect Obama, whites were strikingly unmotivated by Romney. The total white vote dropped from 100 million in 2008 to 98 million in 2012 (down 2%). Only 64.1% of eligible whites voted in 2012, down from 66.1% in 2008 and 67.2% in the recent high-water mark year of 2004. This was the first time in the history of the Census survey that whites were not the highest-ranking group in terms of their rate of voting. Among Hispanics eligible to vote, gross numbers continued to rise — but the rate of those taking the trouble to vote dropped from 49.9% to 48.0%. The number of Hispanics who claimed to be eligible but didn’t bother to get to the polls soared from 9.8 million to 12.1 million. Naturalized immigrants made up only 7.0% of the 2012 electorate. Conversely, native sons and daughters comprised 93.0% of the vote.