After President Obama’s relatively easy reelection, analysts and commentators wondered whether his young and diverse coalition would outlive his presidency. Many believe, based mainly on their intuition, that 2008 and 2012 were the anomalous results of a historic candidacy. On the other hand, the country is getting more and more diverse with each passing year. Recently, one prominent demographer at the Brookings Institute used the exit polls to argue that Obama would have lost if turnout rates returned to ’04 levels. But his effort was misguided and premature.
Today, the Census released the November 2012 Current Population Survey (CPS) Voting and Registration Supplement, which is based on interviews with hundreds of thousands of residents. The CPS asks Americans whether they participated in the last election. The CPS is imperfect like any survey, but it is considered the gold standard for analyzing turnout. It demonstrates that, in the debate about the GOP’s future in an increasingly diverse America, both sides are right, to a certain extent. On the one hand, Obama’s historic candidacy led to historic black turnout. On the other hand, the Obama coalition is the product of irreversible demographic changes. If Republicans hope to win presidential elections, they will need to broaden their appeal—not just count on lower minority turnout in the post-Obama era.
Unsurprisingly, the CPS found that the 2012 electorate was more diverse than any in history. Whites represented just 73.7 percent of the electorate, down from 76.3 in 2008 and 79.2 percent in 2004. In comparison, the exit polls found that whites represented 72 percent of the electorate in 2012, compared to 74 percent in 2008 and 77 percent in 2004. For the first time, the CPS found that black turnout rates exceed white turnout rates, with 66.2 percent of voting age blacks turning out, compared to 64.1 percent of whites. Many expected that black turnout would decline in 2012, but the CPS actually found that black turnout was even higher in 2012 than it was in 2008, increasing from 64.7 to 66.2 percent.
Contrary to the pre-election expectations of many commentators, the CPS found that non-Hispanic white (henceforth white) and Hispanic turnout rates declined: Hispanic turnout fell from 49.9 percent to 48 percent, while white turnout declined from 66.1 to 64.1 percent of voting age citizens. Even so, the growing Hispanic share of the population allowed the Hispanic share of the electorate to increase by a full point, from 7.4 to 8.4 percent.
At the state level, the CPS often contradicted the exit polls—sometimes substantially. In every battleground state, the CPS found a more white electorate than the exit polls. In Colorado, Pennsylvania, and Ohio, the difference was 4.4 to 6.8 points. The CPS estimate is more plausible: The exit polls routinely found that the electorate was significantly more diverse than the voting eligible population and therefore suggested that minority turnout rates were implausibly in excess of white turnout rates, even in states with large Hispanic populations. In contrast, the CPS found that the electorate was whiter than the voting eligible population in the battleground states—often by a significant margin.
The CPS confirmed that there were “missing white voters” in 2012, as Sean Trende described in a post-election piece. However, the CPS found fewer than Trende expected. In Ohio, the CPS found about 250,000 white voters were “missing.” Nationally, the figure was 3 million. If those voters are allocated according to the exit polls, Romney gains a net 40,000 votes in Ohio and 600,000 nationally—not nearly enough to overcome Obama’s 5 million vote edge nationally, or 160,000 vote advantage in Ohio. In the other battleground states, the “missing white voter” phenomenon is less noticeable. Indeed, white turnout even increased in Iowa and Colorado. Obama’s strength in those two states calls into question whether the missing white voters were more Republican than white voters more generally. Obama’s tendency to do best in polls of registered voters, rather than likely voters, adds further doubt.
Even though the CPS found a far whiter electorate than the exit polls, the white share of the electorate still declined in every battleground state except Iowa. The declines were usually modest, but far greater in North Carolina and Nevada, where the white share of the electorate plunged by 5.8 and 5.5 points, respectively. In North Carolina, the CPS finding is probably wrong. The CPS, after all, is still a survey—subject to response rate issues, and a margin of error.1
The big problem for Republicans is that the primary cause of the declining white share of the electorate is demographic change, not high black turnout. Demographic changes have been so significant that the Obama coalition would survive a return to ’04 turnout rates. If minorities had turned out in 2012 at 2004 levels, whites would have represented 74.8 percent of the electorate—1 point higher than the actual 2012 electorate, but 4.4 points lower than 2004’s electorate, which was 79.2 percent white. If all voters turned out at ’04 levels—which increases white turnout and decreases minority turnout—whites would represent 75.6 percent of the electorate. That’s nearly 2 points higher than 2012, but it’s still even more diverse than the supposedly historic 2008 electorate, when whites represented 76.2 percent of the vote.
A return to ’04 turnout wouldn’t have cost Obama the presidency, either. If the exit poll survey results are plugged into the CPS, and adjusted for the discrepancies between the exit poll’s electorate, the CPS, and the actual results, then Obama would have won the election by 2.6 points with ’04 minority turnout rates and 1.9 points if white and minority turnout returned to 2004. In the battleground states, the consequences of declining minority turnout are felt unevenly. In states like Colorado or Iowa, with small black populations, a return to ’04 turnout does little to the president’s standing, since neither state has a significant African American population. On the other hand, Obama suffers 4 point losses in North Carolina and Virginia, making Virginia a true toss-up and pushing North Carolina solidly into the Republican column.
But the problem for Republicans is that the white share of the voting eligible population is likely to decline even further over the next four years. What’s causing the decline? Today’s 15-18 year olds are only about 58 percent white. As they enter the electorate and older whites depart, the non-white share of the voting eligible population rises. This prediction is not subject to great uncertainty. These 15-18 year olds are alive, they’re counted in the Census, and, unless they die, they’re going to be eligible to vote in 2016. If the non-white share of the voting eligible population declines by another 2 points, as expected, then the 2016 electorate will about as diverse as it was in 2012, even if turnout rates return to 2004 levels. The Obama coalition is not going away, even if elevated minority turnout rates are gone for good.
The biggest mistake that Republicans made in 2012 was assuming that 2008 was a special, one-time product of a historic candidate. That was superficially appealing and maybe even “felt” right, but the CPS said that the 2008 turnout wasn’t as unique as the huge crowds and palpable enthusiasm made it seem. The GOP should not delude itself into believing that taking Obama off of the ballot will return them to the White House, even if black turnout rates should be expected to decline in 2016. Demographic change, not turnout, is the primary force driving the declining white share of the electorate, and the GOP will need to adapt.
In 2008, the CPS showed that black turnout rates surged by 33 percent in Virginia, but just 6 percent in North Carolina—even though black voter registration increased by 21 percent in the Tar Heel State. The county-level results join Virginia, the voter registration numbers, and logic in suggesting that the big increase in black turnout came in 2008. For instance, in Bertie County, the North Carolina county where African Americans represent the largest share of the population, turnout increased by 22 percent in 2008, then 4 percent in 2012. Perhaps catching-up to its old error, the CPS in 2012 shows North Carolina black turnout increasing by 17 percent.