Political scientists John Sides and Lynn Vavreck are the authors of "The Gamble," the only book about the 2012 presidential election formally endorsed by Ezra Klein. Sides blogs for The Washington Post over at The Monkey Cage.
Our book about the 2012 election is full of what Wonkblog likes best: graphs, and more graphs. Here are 10 of our favorites, and a bit about what each graph tells us about the 2012 election and politics right now. If you like campaigns (and graphs), then this book is for you.
1) Republicans liked Romney. Really!
In December 2011, YouGov interviewed over 14,000 Republicans who said they were likely to vote in the primary. Of all the candidates, they had the most favorable feelings toward Romney. He wasn’t every Republican’s first choice, but he was viewed positively by enough of them to become the nominee. This is one reason we called Romney “inevitable” even before the New Hampshire primary.
2) Conservative Republicans liked Romney too.
One of the myths of 2012 was that most Republicans wanted “anybody but Romney” — with opposition to Romney concentrated among the conservative wing of the party. In fact, in December 2011, views of Romney were more favorable among conservative Republicans than liberal or moderate Republicans. And views of Romney were not much different regardless of Republicans’ views of abortion or the tea party.
3) Republican primary voters were not much divided by ideology.
Part of the “anybody but Romney” idea was that Republican voters were strongly divided by ideology. But when we mapped the economic and social views of the supporters of the major Republican candidates, we found them clustered tightly together — with the exception of Huntsman supporters, who looked more like Obama supporters overall. This validates what the National Review’s Ramesh Ponnuru wrote in late 2011:
The Republican party now features a remarkable degree of programmatic consensus.
4) Romney appealed to the mainstream of the party.
One version of the “anybody but Romney” idea suggested that Romney only appealed to about 25 percent of the party, with the rest comprised of extremists likened to a “Molotov Party.” In fact, Romney appealed to the majority of the party.
This graph shows that Romney got more support among the party's largest groups — for example, those who were not “born again” and those who thought that abortion should be legal “always” or “sometimes.” Santorum got more support among smaller groups within the party, like those who believed that abortion should always be illegal.
The “anybody but Romney” myth suggested that Romney appealed to a minority of the party. In reality, it was the other Republican candidates whose appeal was more niche. Despite the growing conservatism of the Republican Party, it has nominated a relative moderate from the field of its presidential primary candidates in every election since 1988.
5) The economic fundamentals favored Obama.
This graph compares the incumbent party’s percent of the major-party vote and how much gross domestic product was growing in the first two quarters of the election year. It shows that the outcome in 2012 was no surprise. Incumbents running amidst even modest economic growth have been more likely to win. This is one reason why the Wonkblog election forecast was accurate, even in April. A combination of other forecasting models also showed that Obama was the favorite. Alas, this fact was often misunderstood by commentators.
6) Party loyalty is really powerful.
So this one is a table, not a graph. It compares voters’ preferences in December 2011 to their preferences when they were reinterviewed after the election — almost 11 months later. It shows that the vast majority of people who supported Obama or Romney in December 2011 also reported voting for them. Democrats and Republicans were similarly loyal.
The campaign was integral in helping to increase and solidify these partisan loyalties, but, as this table illustrates, it was very difficult to convince people to defect and vote for the other party’s candidate. It also shows that voters who were undecided or supported another candidate broke nearly evenly (40 percent-41 percent) between the two candidates. Party loyalty is one reason why presidential general election campaigning isn’t really full of the “game-changers” that pundits pine for.
7) Most groups of voters move in similar fashion from election to election.
Before the election, there were any number of stories about whether Obama or Romney was “winning” or “losing” or “in trouble with” some particular group of voters. But for the most part, groups of voters aren’t moving in idiosyncratic ways between any two elections. They are all trending in the same direction: They supported Obama at a slightly lower rate in 2012 than they did in 2008.
This shows how much elections are driven by broader factors like the economy, which was not as favorable to Obama in 2012 as in 2008. And it also shows that elections are driven less by the idiosyncratic interests or agendas of groups of voters.
8) Obama’s “gifts” didn’t amount to much.
After the election, Romney suggested that “gifts” that the Obama administration had given Democratic constituencies — “especially the African-American community, the Hispanic community, and young people” — had won him the election.
Regarding the Hispanic community, Romney cited “the amnesty for children of illegals.” On its face, this seems plausible because the exit poll suggested Hispanics did shift toward Obama between 2008 and 2012. The graph above shows Obama’s approval rating among whites, blacks, and Hispanics. There is no notable trend around Obama’s June 15 “amnesty.” Instead, Latinos were affected more by the Democratic National Convention, illustrating how the conventions rally the party faithful. In "The Gamble," we provide even more evidence that “gifts” like the auto bailout and contraception didn’t mobilize key constituencies like the Rust Belt working class and women.
9) It was hard for Obama or Romney to out-campaign the other.
This graph shows the balance of advertising — measured in gross rating points (GRPs) — in all media markets during the fall of 2012. It also shows the effects of advertising imbalances — that is, what kind of imbalance we would estimate to shift the polls in Obama’s or Romney’s favor. Pay the closest attention to the average across all the media markets in the battleground states, which is tracked with the gray line. It shows you that, on average, neither candidate had a large or consistent advantage in advertising.
Romney had a small lead for most of August, Obama had a small lead for most of September, and then Romney had a small lead from mid-October to the end. We show that these small advantages did not amount to much in terms of durable shifts in the polls.
The broader point is that both candidates and their allies had roughly equal amounts of money, which meant that their campaigns were often canceling each other out. We show that Obama’s five-point margin of victory was likely not due to any advantages in the air war or the ground game.
10) Romney did not lose because he was perceived as too conservative.
In fact, as I previously noted on Wonkblog, he was perceived as closer to the average voter than was Obama. Here is what we say in the book:
The weekend before the election, 52% of voters placed themselves closer to Romney than Obama, while 38% placed themselves closer to Obama. The rest saw themselves as equally close to both. Romney also had this advantage, as he did throughout the campaign, among potential groups of swing voters, such as true independents or voters who as of December 2011 did not prefer either Obama or Romney.
This also complicates any interpretation of the election as a mandate for Obama. He seemed to win in spite of how his political beliefs were perceived, not because of them.
For more, please check out "The Gamble."