Now that the ads, robocalls and dust have settled on South Carolina’s GOP primary, one of the nice things (for a politics data geek like me) is the chance to analyze election returns and exit polls. Everyone fixates (of course) on the immediate “who won & lost,” but beyond that, it helps to understand “why” someone won and lost.
For many, the term “exit poll” raises skepticism about what these figures really tell us. Most everyone remembers the Florida fiasco of the 2000 presidential election, and that experience certainly left a bad taste for many in regard to the reliability of exit polls. But this year, Florida may help settle what has been a truly unsettling GOP nomination race.
Exit polls provide a snapshot sample of who showed up to vote in an election. First, a little background: exit polls are conducted at predetermined polling sites, with a worker outside the polling location when voters leave.
Seeking a random sample, the worker asks voters at specific intervals (say, every 10th voter) if they would answer a series of questions on the front and back of a piece of paper. Voters who agree fill out the questionnaire and place it in a box. The worker then calls in the results three times during the day, where the information is tabulated. Adjustments are made to ensure the sample approximates the population of the area. This was my experience in 2002 when I worked as an exit pollster in Georgia.
One valid concern nowadays is the fact that so many voters are taking advantage of early voting. In North Carolina, for example, early votes in 2008’s presidential election made up 57 percent of all the votes cast (Even though John McCain lost NC, he easily received the majority of votes cast on Election Day). This makes it more difficult for the exit polling company to gather its information, but the exit poll firm calls voters who cast their ballots early and includes them in their sample.
The main benefit that we get from exit polls is not necessarily predicting who wins, but why a candidate won, who made up the electorate, and what issues were important to those voters. So with South Carolina the first true Republican base state voting in the nomination process (Iowa and New Hampshire are classic swing states), who showed up in the South Carolina primary?
Interestingly, among the 2,381 respondents, men were the majority among Saturday’s electorate, 51 percent to 49 percent for women. This may sound normal to some, but in 2008’s primary election, men made up 56 percent. In the 2008 general election, women were in the majority at 54 percent of the electorate.
In the 2012 primary, South Carolina men went for fellow Southerner Gingrich 42 percent to 26 percent for Romney, while women went 38 percent for Newt to 29 percent for Romney. White males (in Saturday’s SC GOP primary, whites were 98 percent of the electorate) are core voters for the GOP, especially Southern white males.
Conservatives, who typically are 40 percent of South Carolina’s general electorate, were 68 percent of Saturday’s primary electorate, and Gingrich won them convincingly, 45 percent to Romney’s 24 percent.
The conventional wisdom held that Gingrich would suffer at the hands of two key voting groups in the SC primary: Married voters and born-again/evangelical voters, due to the three marriages and the Nightline interview of his second wife. But married voters, who were three-fourths of the electorate, went 41 percent for Newt, compared to 28 percent for Mitt. And born-again/evangelical voters—those who believe in “redemption” and were two-thirds of the primary electorate—broke in favor of Newt by a 2-to-1 margin over Mitt.
An interesting question posed before the primary was, “how would Tea Party supporters break?” With Governor Nikki Haley’s endorsement, many believed that Romney would benefit from that support. But Tea Party supporters went 45 percent for Newt, compared to only 25 percent for Mitt. And those who said they approved of Gov. Haley went 42 percent for Gingrich, while Romney got 30 percent of their vote. One has to question whether there really is a leader of the Tea Party, something advocates have argued for a long time that there isn’t, and whether endorsements truly influence a voter’s decision.
One question that seems to indicate the power of the two debates leading into the primary was, “When did you decide who you would vote for?” For those who decide in the days heading into the primary (53 percent of the electorate), Newt’s performances at the debate, especially Thursday night’s CNN debate, must have had a dramatic impact: 44 percent of those deciding within the “last few days” of the campaign broke for Newt, while Romney garnered 22 percent of the last-minute deciders.
So what does all this tell us about the Republican base in a key red state? For Mitt Romney to become the GOP nominee, losing critical party constituents is a significant warning sign. White males, Tea Party supporters, born-again/evangelical voters—all of these constituents are critical components to the modern GOP party. Granted, these groups have a smaller impact within a general election ( for example, evangelicals made up 40 percent of South Carolina’s electorate in November 2008), but failing to capture these crucial base supporters in a nomination fight means that if Mitt Romney becomes the GOP nominee, he will have serious work to do among those core supporters.
Florida’s primary on January 31 will be a real test of whether Romney — who received 25 percent of Iowa’s caucus vote and 28 percent of South Carolina’s primary vote — can craft a winning coalition among core Republican voters he needs to secure the Republican nomination.