The Crowdwire

You’ve found The Crowdwire, an effort to understand how social media are changing politics. Beginning in late summer 2012, we analyzed the social response to the U.S. presidential election, including the conventions, debates and other key events. The results are shown here, exactly as they were published during the campaign season

The Crowdwire was created by Bluefin Labs, a technology company that grew out of research conducted at the MIT Media Lab. Bluefin developed sophisticated language-based tools to make sense of the social response to televised content. We used those same tools to learn what people were saying about the presidential candidates and the issues.

To read more about our goals, see “Choosing a President in the Social Age,” an op-ed by Deb Roy, co-founder of Bluefin Labs, and William Powers, director of The Crowdwire.

To check out our work, just scroll down or browse the archive. In addition, all of our subject tags are listed below. Click on any one to see all the posts on that topic in chronological order.

We recommend two tags in particular: 1) The Three P’s, an experiment in parsing the new political conversation; and 2) Lessons, in which we share what we learned.

The Crowdwire was a non-commercial, non-partisan project. Our goal was to make a contribution to the public discourse about politics and policy. We tracked public comments only. Since the Twitter conversation is predominantly public, and Facebook largely private, most of our data came from Twitter.

Thanks for stopping by.

andrew heyward    ann romney    audience groups    benghazi    bluefin labs    biden     black voters   boston globe    brands    candy crowley     corrections    crowdwire backstory     crowdwire live election day exit poll    dan siegel    david quammen    debates    depth    economy    elizabeth warren    eliza newlin carney    first-time voters    foreign policy    47%    geo-location of tweeters    healthcare    human factor    jennifer hollett    jim lehrer    journalism    lessons    listeners    machine learning    martha raddatz    media coverage    men    michael franz    michelle obama    national geographic    neil swidey    obama    partisans    political ads    post-election   predictions   Q and A   retweets    romney    russell  stevens    ryan    seniors    sentiment    slogans    super pacs   super pac app    superstorm sandy    stefania barina    the president    the three p’s    2012 democratic convention    2012 republican convention   viral hashtags    visualizations    wesleyan university    women    youth vote

Lesson #7: For the future of social media, the important question is not how large the conversation is, but how meaningful. The challenge is understanding what all those voices are saying. Through human-machine collaboration, we’re learning to do just that.

Pictured above is a giant sequoia in California known as The President. One of the tallest trees in the world, it’s at least 3,200 years old. To put that number in perspective, when democracy was first emerging in ancient Greece, this tree had already been around for six centuries.

We’re telling you this because the tree is a nice metaphor for this project. The President was featured on the cover of of National Geographic’s December issue, just after a much younger president won his second term in the White House. And this sequoia has a number of things in common with the subject we’ve been exploring, the social-media conversation that’s reshaping democracy:

• Like the tree, the social conversation is enormous. There are now half a billion new Tweets appearing every day. Within that virtual agora, an immense discussion of politics and public policy blossomed during the election.

• The tree is “still growing quickly,” scientists report, and so is the social discourse.

• In both cases, advanced digital technologies are making a colossal phenomenon more comprehensible, by helping us humans process vast amounts of information, Big Data-style.

This last point was central to The Crowdwire’s work. We took the sophisticated, language-based tools developed by Bluefin Labs to make sense of the social-media response to televised content – shows, news events and ads – and applied them to the presidential election. Every analysis we published was based on our team’s work with these tools.

LEAF CALCULUS

This brings us back to the tree. One evening after Election Day, some of us went out for an impromptu drink, to celebrate the end of our experiment. Someone brought along the full-length photo of The President shown above and asked the group to play a guessing game. The question: How many leaves does this tree have?

A few hints were provided: 1) The tree is 247 feet tall; 2) The answer is equal to the total number of views that YouTube videos mentioning the two presidential candidates had during the election.

One of our colleagues studied the photo for a moment and started scribbling numbers on a slip of paper. “It’s all about the packing fraction,” he said. “Once you know that, it’s easy.”

His answer was 111.35 million leaves. The other entries ranged from 17 million to 7 billion, but nobody came close to the correct number: 2 billion. The winning guess, 1.25 billion, was off by 775 million leaves.

Though it was just a game, it points to where Big Data, social media and politics are headed:

1) The Power of Tools. It would be nice to think that you could just look at a tree and know everything about it. But for most of the time The President has been alive, some very basic questions about it were beyond the capacity of humans to answer. Modern tools have now solved many of these problems. National Geographic writer David Quammen explains that scientists measured its branches and burls, took core samples and gathered other data. “Then they fed the numbers through mathematical models informed by additional data from other giant sequoias” and determined, among other things, its age and leaf count.

We took an analogous approach to measure the size and growth of the social response to the election, using different tools. Without them, we would have been guessing in the dark, just as we did with the tree. With them, we were able to make the massive, chaotic trove of language that social media offer up each day more coherent and, we hope, useful.

2) The Human Factor. Machines are very good at certain kinds of work, such as mathematical calculations. But to understand something as complex as a giant tree or global conversation, you have to go deeper than numbers. Our best work was a product of the processing power of computers combined with the creative thinking that’s unique to humans.

At one point, for example, we realized that we’d learned a lot about what people were saying on Twitter and Facebook about the election, but had no idea if they were listening to each other. Since a true conversation involves talking and listening, we wondered how much of the latter was occurring. Our computers never thought to ask this question. But when we did, they helped us answer it.

Another lesson we learned over and over was the power of direct human observation. The President houses a large ecosystem of plants, insects, birds and animals that spend most of their lives there. In short, the tree is alive in more ways than one. The scientists didn’t learn this from the software. They physically explored the inner world of the tree and saw the creatures for themselves.

Likewise, we took a hands-on approach to our subject. No matter how good your algorithms, never blindly trust the machines. We worked in close collaboration with the computers, actively monitoring their work and often digging into the data ourselves. Immersed in the details and nuances of the conversation, we noticed many things they missed.

3) The Quest for Meaning. So we need the tools and the tools need us. But where are we going together? If we’re smart about it, toward a better understanding of ourselves.

The Crowdwire answered dozens of compelling questions, but many more are crying out for attention. For example: 

• Will computers become more skillful at parsing the subtleties and ambiguities of human language?

• What was it exactly about the presidential debates that drew such large TV audiences and so much social commentary?

• The election is over, but the issues that drove it still need society’s attention and best thinking. Can we keep this lively public discourse going? Might it offer a new model for journalism?

GO DEEP, AIM HIGH

Social media haven’t been around for three millennia; they’re brand new. But the questions discussed on these platforms during the election - questions about the human struggle and how we might live together in peace and prosperity - are as old as the tree.

The best conversations are not necessarily the largest ones. They’re the meaningful ones. If we can learn to understand this conversation in all its depth, we might just change the world.

Tree photos by Michael “Nick” Nichols for National Geographic

Lesson #6: Despite complaints about the stupidity of modern politics, the 2012 election conversation was remarkably serious. In social media, substantial policy issues were often at the center of the discussion, especially in response to the debates.

There’s a widespread belief that our political process is broken. In a new Gallup poll, more than three out of four Americans said politics is doing serious harm to the country.

Some blame the politicians for narrow-minded partisanship. Others cite the news media for treating politics as entertainment. This is especially true during elections, when journalists tend to focus on who’s up and who’s down in the polls - the so-called Horse Race - rather than the issues that really matter.

Is the American political conversation on a relentless downward slide? The recent rancor in Washington about fiscal policy would suggest it is.

But during the 2012 election we identified a striking counter-trend, not in Washington but on the social platforms. When the candidates had serious exchanges about the issues, the social response was enormous and dominated by policy questions.

This flies in the face of everything we’ve been told about the decline of political discourse and the alleged triviality of social media commentary. Yet based on extensive data we gathered in the final month of the election, the conclusion was inescapable.

A NEW KIND OF ANALYSIS

It emerged from the The Three P’s, a Crowdwire project that aimed to parse the social conversation in a new way. In the debut post, we asked:

In an age when every citizen with access to technology can join the conversation, is our political process changing? Does the horse race still predominate over the issues?

To find out, we used sophisticated language-based analytic tools that broke down the social response to the election into three categories: 1) Politics: the campaign itself, the horse race; 2) Policy: the issues, from the economy to foreign policy to education; and 3) Personality: the candidates as people - their life stories, human qualities and quirks.

We determined what statistical proportion of the conversation - a metric called “share of voice” - was devoted to each of the P’s, and tracked how the balance fluctuated over time. We checked in often, particularly around each of the debates, and were surprised by what we found.

After decades of growing media focus on pure horse-race politics, it appears the public - the slice of it using social media, anyway - has a stronger interest in policy questions than journalists often assume.

Going into Debate #1, for instance, Politics had the greatest share-of-voice in the election conversation. But during that very issue-oriented contest, dominated by unemployment and other economic issues, the balance shifted noticeably toward Policy.

And the effect persisted. Two days later, when the government released an important new jobs report, Policy’s share-of-voice spiked, overtaking Politics. Why? Because the new jobs number was meaningful within the context of the debate, and social-media responded.

BEYOND BINDERS

So it went through the final weeks of the election. When Obama and Romney had serious policy exchanges, mainly in the debates, the Tweets and public Facebook posts exploded with policy-focused comments. Even as the mainstream media focused almost exclusively on viral phrases like #bindersfullofwomen that made good headlines, the sprawling social crowd was talking about the issues.

Here’s our final graph showing the Three P’s fluctuating in influence from early October until the week after the election. Notice that, with the exception of the Vice Presidential debate (analyzed here), Policy spiked either during or just after each debate.

What can we learn from this? We asked Russell Stevens, a friend of the Crowdwire who conducted an election-media study at Harvard’s Kennedy School of Government in the 1990s that was an inspiration for The Three P’s project. He said:

I think these results reflect human nature. Voters care more, deep down, about issues than the horse race. Perhaps the fact that the social response tended more toward substance is a signal for the press to dig deeper on the Policy ‘P’ next time around.

MACHINE REVELATIONS

When the election was over we kept the tool running, just to see what happened. The resignation of the CIA director, Gen. David Petraeus, and the resulting scandal immediately took over the headlines. We programmed this in as a Policy issue, meaning all Petreaus mentions would be classified under that heading.

In fact, many social commenters saw the Petraeus imbroglio as political, in the sense that it was mainly about damage control, not a real threat to national security. The tool picked up on this in the language used in Tweets and public Facebook posts, and it was reflected in the data.

Depending on the terms used, any comment could potentially be assigned to more than one category. So although we directed all Petraeus mentions to Policy, if any of these comments also contained language we’d previously classified under Politics, they would contribute to the share-of-voice for both P’s.

As the section of the graph for the Petraeus period shows, after a brief in increase in Policy’s share-of-voice, Politics soared.

In retrospect, the Petraeus story was largely a political firestorm. We didn’t see this coming, but the tool had our backs. This is a reminder that human expectations can be wrong in many directions. If you assume elections are all about polls and political gamesmanship, the public might surprise you by talking policy. Conversely, a seeming policy crisis will sometimes turn out to be a largely political one, or a mix of the two.

Though humans invent smart machines and control them, the machines can amaze us with revelations of their own.

Lesson #5: Super-PAC-funded ads did not swing the 2012 election but could be a potent force in the future. Energetic individuals and organizations are working hard to make the new world of campaign finance more transparent.

At the start of the campaign season, there were widespread fears about super PACs influencing the election. The worst scenarios, which saw these new fundraising organizations essentially taking over politics, were not realized. It’s generally agreed that $500M in super PAC-funded ads didn’t decide the election.

We’ve noted that as social media increasingly frame the election conversation, TV ads seem to be declining in influence. But that could easily change as the business of politics responds to the new media landscape and advertising itself evolves.

So campaign finance needs to be closely watched, and there’s every indication it will be. Concern about super PACs has reignited a longstanding movement to shine a brighter light on political money and the ads it supports. We talked to three leaders in the field, each coming at this challenge from a different direction.

The Journalist

“Super PACs have injected multi-million-dollar contributions from wealthy CEOs, billionaires, and corporate titans into the [political] system, something we haven’t seen since before the post-Watergate reforms of the 1970s.” -Eliza Newlin Carney

In 2010, two Supreme Court rulings gave independent, politically active groups the green light to collect unlimited contributions from any source. Washington journalist Carney, who has been covering campaign finance for years, immediately knew that these PACs would take off and should be aggressively covered. But how to make this complex business comprehensible to the lay person?

“I started thinking about how I could write about them without making readers’ eyes glaze over,” Carney told us. “The technical, legal term for these organizations is ‘independent expenditure-only Political Action Committee’ – not something you want to write ten times in a story.”

She started calling them “super PACs” – and the phrase took off.

“Election laws are complicated, even turgid, so it can be challenging to write about them without getting bogged down. I liked ‘super’ because it was short, sharp and punchy, and seemed to capture the notion that these groups could do a lot of things that conventional PACs could not.”

In a few short years, super PACs have indeed transformed the world of political finance. They’ve forced candidates, parties, and their outside allies to raise and spend money much earlier, dramatically front-loading the campaign.

However, she notes, “their greatest impact may be in the congressional and state legislative races. That’s where a last-minute seven-figure ad buy can really tilt the outcome.”

Carney believes super PACs pose a serious threat to American politics. “The two trends unleashed by Citizens United – unrestricted spending and secret money – make for a scandal waiting to happen.”

Thus, she and many other journalists are watching the field closely. “This used to be a beat that I covered alongside a relatively small handful of other reporters. Now I’ve got plenty of company, and the political money stories coming out of major newspapers and investigative nonprofits are breaking new ground.”

Even Stephen Colbert has taken on this subject, using humor to explain the arcane, shell-game aspects of this world.

Carney, too, increased the public’s understanding of these organizations by coining the phrase that everyone now uses to talk about them. Though modest about this achievement, she says, “if the term has helped average Americans understand our political system, I’m glad ‘super PAC’ has become part of the national vocabulary.”

The Academic

“The 2012 presidential race had more ads than any previous election.” -Michael Franz

Franz knows political advertising. His organization, the Wesleyan Media Project, has been tracking campaign commercials since the late 1990s. But in these post-Citizens United days, the stakes are even higher.

“With outside groups increasingly prevalent in federal elections, it is imperative that voters know to what extent wealthy groups and donors are bankrolling persuasion efforts on television,” Franz said in a Crowdwire Q & A.

As an academic research organization, the Wesleyan Media Project addresses this challenge without any partisan or ideological agenda, by providing voters with information about who is advertising and at what levels.

Using weekly data from Kantar Media/CMAG, the project codes each TV ad on nearly 100 dimensions including tone, sentiment, issues covered, and the sponsor’s identity.

“By appending our coding to the frequency of ad airings, we can then see what kinds of ads are airing where, by whom, and at what levels.”

The 2012 election has kept Franz and his colleagues busy. It’s “broken lots of records with respect to political ads. Both candidates, especially Obama, have aired an incredible number of negative ads. And outside groups, especially in support of Romney, have been extremely active. We never expected to see so many congressional candidates, and a major presidential candidate, so dependent on outside spending.”

Data from the Wesleyan Media Project have allowed political scientists to develop and test sophisticated theories about how ads work in convincing citizens to vote, and to see if ads have any effects on turnout or knowledge of key issues.

“It seems pretty clear that ads “work” in the sense that a great number of ads can move votes in your direction. But because many campaigns wind up with an equal number of ads for each side, the net effect is often zero.”

That might help explain the ads’ underwhelming influence in this presidential election.

Meanwhile, Franz believes commercials have an upside. He notes political science has reached consensus that ads do not depress voter turnout; they might even stimulate interest in the election and in some cases, raise turnout.

“Ads are attempts to speak to all voters, and they can sometimes be the only place for a Republican voter to hear what the Democratic candidate has to say, and vice versa.”

The Entrepreneur

“With your TV screaming political messages at you during every commercial break, there’s a tendency to tune out. We were inspired to create something that cuts through the noise, because voters shouldn’t have that hopeless “forget it” feeling.” -Dan Siegel

Entrepreneur Siegel and his colleague Jennifer Hollett wanted to help ordinary Americans make sense of the money game while bringing more transparency to the 2012 presidential election. Their idea: A free mobile application called the Super PAC App.

“Who is behind these ads?” Siegel asked. “What are they saying, and are the claims based on facts?”

Until now, finding answers required hours of homework and probably a trip to town hall. But the Super PAC App uses audio-fingering printing technology to tag political commercials on TV or the web so users can instantly access news outlets and independent fact-checking organizations that have investigated the claims in question.

“Super PAC App won’t tell you if what the ad is claiming is ‘true’ or ‘false,’ but it will connect you with information to make that assessment yourself,” Siegel said in our interview.

The application also allows users to provide feedback on how they received the ad – if they loved it, thought it was Fair, Fishy, or a Fail. The result: a sizable new database of political advertising with crowd-sourced ratings.

New Directions

These trailblazing efforts demonstrate that, even in a rapidly changing politico-media environment, a handful of smart, energetic people can substantially increase understanding of how money influences elections.

What next? It’s hard to say, but as long as money and politics are intertwined, there will be a need for watch-dogs, and the more, the better. Future initiatives might use social media to give the public a larger role in opening up this world for closer inspection.

Lesson #4: The business of politics is still rooted in TV, but in a new way. The election conversation is migrating to the place where TV and social media intersect, and it’s changing how candidates sell themselves and how voters vote.

In one of our early posts, we noted that presidential campaigns have long behaved like commercial brands. Since the 1960s, the business of politics has involved transmitting campaign messages to a passive mass audience, largely through TV ads. The most effective ads, such as Ronald Reagan’s 1984 “It’s Morning Again in America,” had a major impact on elections.

But like commercial brands, political brands are being transformed by social media. And TV’s role is shifting. In this election, the sheer quantity and cost of the TV ads aired by the two campaigns and their super PAC supporters drew a lot of attention. Journalists and others were justifiably concerned about big money taking over politics (we did Q & A’s with several forward-thinking experts in this area and will sum up what we learned in a separate post).

But to the surprise of many, the super PAC ads were not very influential. And the ads aired by the campaigns themselves had minor impact. Can you think of a single TV commercial that was decisive in this election?

What’s going on here? Is TV dead as a political force?

Far from it. TV remains at the center of American politics, even as the influence of political ads seems to be declining. It’s a paradox that the mainstream media haven’t wrapped their arms around, as they scramble to make sense of how social media are affecting TV. The standard view is that TV is fading into the past and social networks are the future, with the latter soon to replace the former.

But when it comes to the media, this old-supplants-new storyline rarely pans out. In the years after World War II, for instance, radio was widely thought to be doomed, made obsolete by a revolutionary new gadget called television. Yet more than sixty years later, radio is still thriving alongside TV.

Something similar is happening now with TV, which continues to draw large audiences in the social age, especially around elections. What’s new is that viewers are joining in with their smartphones, speaking up and effectively becoming participants.

If we’ve learned anything in this project, it’s that politics and TV are recombining in a new way, and social media are the catalyst. This shift has enormous implications for politics generally, and in particular for the business of elections.

Key points:

• TV as Narrative. The main narrative of presidential politics still unfolds on TV through: 1) Mass-audience televised events such as conventions and debates, and 2) News coverage of the candidates and their campaigns. 

• Debates as Pop Culture. The presidential debates weren’t just a major election event, they were a pop-culture landmark. The first Obama-Romney debate drew the largest debate audience in 32 years, nearly 70 million viewers. Very few televised events today are

  

  watched by that many people. This year’s Grammy Awards and Academy Awards each pulled in just under 40 million viewers. Could it really be that more Americans are interested in watching two politicians argue health-care and tax policy than in Hollywood extravaganzas featuring the likes of George Clooney and Adele? Yup.

• Social Response Grows. The social response to televised election events grew rapidly this year. During the GOP primary season, 500,000 comments about a debate was impressive. By October, the presidential debates were pulling in an average of more than 10 million tweets and public Facebook posts each. We discovered that the debates prompted many viewers to post their first-ever tweets about a televised event. As social media transform TV from a passive experience into an active one, politics on TV becomes more compelling.

• Political Brands Thrive in Social. Presidential candidates must compete with commercial brands for the public’s attention. And the social conversation - where both politics and consumerism are major topics - is a handy way of measuring how well the political brands do. We tracked how Brand Obama and Brand Romney performed on Twitter and Facebook versus 600 of the largest consumer brands. Focusing on the sheer volume of discussion as a measure of attention, we found that from early September on, Obama and Romney were respectively the #1 and #2 most talked-about brands in the marketplace, beating Starbucks, Google and the iPhone.

• Hashtags = Crowd-Sourced Slogans. Traditional approaches to political branding are losing their effectiveness as the action moves to the TV-social nexus. We showed that, in this election, TV ads didn’t spark anything close to the social-media interest that the

  

  debates did. In another analysis, we examined the campaigns’ efforts to launch popular slogans via TV ads. None of their slogans caught fire. Meanwhile, every debate produced viral hashtags such as #bigbird and #bindersfullofwomen. These phrases, essentially crowd-sourced slogans, shaped the election conversation much as traditional slogans used to do.

• Authentic Participation. Why did the debates trump ads in terms of public interest and influence? We hazarded a theory: They’re less packaged and slick, more in-depth, spontaneous and real than political ads. Americans tend to view TV ads as entertainment and art form, as the annual the Super Bowl ads ritual makes clear. But ever since presidential campaigns starting using commercial-style TV ads, there’s been a widely shared discomfort and weariness with political ads. Though everyone knows politics long ago adopted commercial methods, and candidates must use ads to make the sale, many voters view campaign ads as noxious. The complaints typically cite the quantity, negativity and deceptiveness of the ads. In accepting his party’s nomination in September, Obama himself said, “If you’re sick of hearing me approve this message, believe me, so am I.” The debates - live, nonpartisan, unfiltered, largely unscripted political contests - don’t have these problems. And now that they’re interactive, debates offer voters a chance to join the conversation. Grab the hashtag of the hour, tweet out your take and suddenly you’re a player. With this combination of authenticity and participation, it’s no wonder millions tuned in and shared their views.

In sum, elections are no longer driven solely by TV, but by the symbiosis of TV and social media. As millions of voters engage in this new dynamic, it will transform the business of politics.

Since launching The Crowdwire, we’ve received some nice coverage from the news media, most of it citing our analysis of specific Election 2012 events.

But we’ve always viewed this as an experiment with implications beyond this election, and even beyond politics. Our ultimate aim has been to help advance understanding of the huge, vibrant conversation that’s emerging in social media and reshaping society.

Now that the election is over, journalists and others have begun to take a longer view of The Crowdwire, and we want to share two examples from the last week.

The Boston Globe Magazine’s most recent cover story is an in-depth look at Bluefin Labs, sponsor and home of this project. At one point in the piece, writer Neil Swidey

If you don’t know the Bluefin story, or even if you do, it’s a terrific read.

The other piece is a thoughtful review of our efforts published by the Research Centre of Digital Ethnography in Milan, Italy. Among all the analysis of social media data produced during the very recent U.S. elections, “those proposed by The Crowdwire project appeared to us absolutely remarkable,” writes Stefania Barina.

Ms. Barina kindly translated her post into English for us and you can download the translation here.

Beyond tracking the public response to Obama and Romney, she writes, The Crowdwire “has a larger goal: it aims to investigate the deeper dynamics of interaction between social media and politics, redefining models of conversation and participation, measuring the ‘pulse of the political process’ and the changes taking place throughout the system.”

That’s exactly what we’re about and we’re grateful people are noticing.

Lesson #3: Social media add depth and texture to our understanding of the public mind.

In this and all other presidential elections, polls are an obsession. And why not? Everyone wants to know where the horse race stands at any given moment, so we follow the numbers.

Though the numbers are just statistics, they’re driven by real people and their views. And the weakness of the horse-race conversation is how little it reveals about what voters are actually thinking and feeling about the candidates and the issues.

It’s good to know who’s ahead in the swing states, but it’s much better to know why.

This is where the social conversation adds serious value. By revealing exactly how millions of people are responding to the campaign as it unfolds, social data make poll numbers more meaningful. Two examples from The Crowdwire’s analysis of this election:

• Women and the Economy. As we noted in the previous post, women spoke up more then men about certain election events, including the Democratic National Convention and all three presidential  debates. Since female voters were considered decisive in this election, this evidence of engagement was significant

  

  all by itself. But it doesn’t tell you what women were thinking about.  News reports about women in this election tended to focus on abortion, often leaving an impression that females vote on reproductive issues alone. Curious what else mattered to women in this election, we looked at how they responded to a major international issue - Benghazi - versus how they responded to economic questions. Did one of these issues get more traction with women? At one point when both were prominent in the news, we showed that women were more engaged by the economy. Later,  when pay-equity issues and Benghazi both came up in the second presidential debate, we again found that women spoke up more about the economic question. In an election that was centered on the economy, women wound up voting for Obama 55% to 44%.

• Seniors Tweet Healthcare. Seniors are not considered a major force in social media, yet they’re speaking up on Twitter and Facebook in growing numbers. Since older voters were an important factor in this election, we wondered if we could detect what issues they were tracking most closely. Answer: healthcare. When Paul Ryan was booed at an AARP conference for remarks he made about seniors and Medicare, we reported that this elicited a significant response in social media. In the same post, we showed that statements about healthcare by Obama and Romney on 60 Minutes, a TV show with a distinctly older viewership, also drove a notable social-media bump. Even seniors can be viewed more clearly through the social lens.

Bottom line for future elections: If you just want to follow the polls, call up the latest numbers. If you want to understand them, turn to social media.

Lesson #2: Social media can provide useful clues about where a big election is heading.

“Are you going to predict the winner?”

When we told friends and family that The Crowdwire’s mission was to track and analyze the social-media response to the election, many asked us that question.

No, we didn’t make any predictions. As we’ve often noted here, the social-media population is not a perfect mirror of the voting public. And in an election decided by a handful of swing states, geography is very important. But in analyzing Twitter data, geo-location is a challenge (we took a stab at it here).

Earlier in the campaign, USA Today attempted to predict the results of the Super Tuesday GOP primaries, using social media. The paper forecasted which states would go to Romney, Gingrich, Santorum and Ron Paul. They were 50% right, coin-toss territory.

So we stayed out of the prediction business. But we discovered that, with the right tools, one can piece together a lot of useful clues about where an election is headed.

The clues we

In short, this experiment was about the substance of the election, which in the end is what really matters. Some of the clues we unearthed:

• Overall, Romney had a tough time in social media. After clinching the nomination, he had a bumpy summer, including a string of foreign gaffes that gave rise to #romneyshambles and other negative viral hashtags.

• The response to the Republican National Convention was dominated by anti-Romney comments. We found that people with anti-Romney views (identified by their previous social remarks) tweeted and posted publicly on Facebook more frequently than the average social user by a factor of three. 

• Meanwhile, the strongest response to the Democratic National Convention came from people supporting Obama. In short, Obama seemed to be better at rallying his own supporters to speak up.

• While men participated more than women in the conversation about the GOP primary race and the RNC, with the DNC it was the reverse. More women spoke up about the Democratic gathering and also about the debates. In a tight election in which female voters were the most sought-after demographic, this was good news for Obama.

• Beginning with the DNC, we established that African-American voters, another important Obama demographic, were watching and commenting on key election events in large numbers. Since these voters were also considered key to Obama’s victory, this signal that they were engaged by the election was another clue that he might be on the way to victory.

In the next post: more insights from the social data confirmed by the election.

When we started our Three P’s project, the aim was to break down the election conversation in a way that made it more understandable and meaningful. Beginning in early October, we checked in regularly on the data to see which of the three P’s - Politics, Policy or Personality - was pulling in the most commentary or “share of voice.” 

Throughout this experiment, Politics, the horse-race aspect of the campaign, tended to lead the way. But as the graph shows, Policy was also strong throughout, and there

With the election over, we were curious to see what happened next. Would Politics sink to the bottom of the graph? Would Policy dominate as Obama turned his attention back to governing?

In both cases, the answer was no. As the graph shows, in the days immediately after the election, Politics remained dominant. The only exception was the brief period from November 8th to 9th when Policy overtook Politics. This occurred as purely political chatter about the election began to decline while two Policy subjects - the ongoing response to Superstorm Sandy and the Fiscal Cliff - took center stage.

Then, late on the 9th, the resignation of General Petraeus seized the Twitterverse. We updated our tool to capture discussion of the Petraeus story, which we classified as a Policy topic. Even so, a good portion of that conversation was inherently political from the start. When political terms such as “election” were used in remarks about Petraeus, those tweets and posts were automatically assigned to both categories. And there were more Tweets and public Facebook posts about Petraeus as Politics than Petreaus as Policy dilemma for the administration and Congress.

In addition, there was a great deal of commentary about political subjects unrelated to the scandal, including the vote count in Florida and post-mortem analysis of the election. As a result, going into this week, Politics was still on top.

Lesson #1:  Social media have become the pulse of the political process.

Are social media changing politics? Influencing our choice of leaders and how they govern? Making democracy more democratic?

These questions have driven The Crowdwire’s analysis of the 2012 presidential election. Now it’s over and we’re reflecting on what we’ve learned. This is the first in a series of posts.

Take a high-stakes race for the most powerful political office in the world. Add a series of debates on primetime TV. Give the audience tools that let them publicly share their reactions in real time… and something remarkable happens: Millions of voices come alive.

It caught on quickly. Last winter, if one of the GOP debates drew 500,000 tweets and public Facebook posts, it was amazing. But the closer we got to Election Day, and the more the race looked like a dead heat, the more the social zone lit up.

Last month, the three presidential debates drew enormous TV audiences - nearly 70 million viewers for the first one alone - and an average of more than 10 million tweets and posts each. The second debate was a bigger social-media event than this year’s Super Bowl.

That one fact arguably tells us more about the political future than Tuesday’s vote count. It says that even in this entertainment-obsessed time, politics and policy - the hard questions about what kind of society this should be - still have the power to engage. Indeed, it’s the combination of the old tool that’s entertained us for decades (TV) and the new one (social media) that’s enabling this fresh wave of political engagement.

It happened because politics and social media, Twitter in particular, are made for each other. The constantly shifting drama of the campaign, a narrative that unfolds mostly on TV, has long been a one-way transaction. They debated, we watched.

Not anymore. Anyone with a digital connection now has the power to talk back to these events, on the same platform where the candidates, the journalists and the rest of the influencer class are sharing their messages and keeping up with the story. In a sense, everyone becomes an influencer.

And, like a finger pressing the carotid artery of the body politic, Twitter picks up the pulse.

This wasn’t the first election with social technologies. But it was the first one in which they played a leading role. Since the last election, there’s been a crucial shift in the TV-social dynamic. Four years ago, television was in the driver’s seat and the Twittersphere was riding in back. This time, the newcomer was up front.

Twitter birthed the viral hashtags that gave this election its pop-culture currency - #emptychair, #bindersfullofwomen, #horsesandbayonets and many others. But in a larger way, it often framed the whole conversation, by telling the mainstream news outlets what the story was.

For instance: How, in the course of the first debate, did the political journalists quickly realize that Obama was bombing? First, they saw it with their own eyes. But just as important, they got that essential 21st-century confirmation: Twitter said so.

Nick Bilton of the New York Times nicely captured the role reversal in an election-night tweet:

This was one of 28.5 million tweets and public Facebook comments posted as the returns came in, or roughly one for every four votes cast. Together with the conventions and the debates, that made this election one of the most commented-on events of any kind in the history of social media.

It’s not a long history. And the size of a conversation - how many people spoke up - says nothing about its quality or value. From the start, The Crowdwire has aimed to go deeper than mere numbers, and in forthcoming posts we’ll sum up what we found.

[Photo: David Silver, 2007]

Note: This post was published on Election Day when our Live Exit Poll was running here in real time. The poll was turned on at 12:01 am that day and turned off at 12 midnight, shortly after the presidential-race results were known. Though no longer live, it stands as a snapshot of a very big day in politics and social media.

Running now and throughout Election Day at the top of this page is The Crowdwire’s exclusive Live Social Media Exit Poll. It shows the number of people who are sharing the news that they voted - either by tweet or public Facebook post - and, if they reveal it, how they voted.

The numbers are updated every five minutes. The gray section at the top shows the overall number of people reporting they voted. This reflects everyone who reports voting today by saying “I voted,” #voted, #justvoted and many other similar expressions. The tally of those who also included an indication of how they voted - either for Obama or Romney - is shown in the red and blue sections.*

Hover over the graph and it will show the numbers for any 5-minute period.

This is obviously not a mirror of the voting public, or a reflection of how the public at large is voting today. Rather, it reflects the particular slice of the public who are social-media users. Imagine that a new state has just been added to the union. In terms of population, it’s a big California-sized state, but it has no geographic location. Its residents are famous for having a lot to say about politics and other subjects. Our exit poll is a live dispatch from the state called Social Media where, as the graph shows, the voting is now underway!

*If our “vote count” were based on pure mentions of Obama and Romney, it would capture people who happen to mention the candidate they oppose and other irrelevant mentions. The tools we used aim to filter out as many such comments as possible. We are monitoring the results through the day to ensure that the data reflect actual voting tendencies as reported in tweets and public Facebook posts.

Joe Biden bigger than the iPad? Paul Ryan close to overtaking Starbucks? Conventional wisdom says that in the U.S., elections take a back seat to consumerism. But not lately.

The Crowdwire launched with a look at Obama and Romney as brands. Presidential candidates are not typical brands, and far more is at stake in this election than in the battle between, say, McDonald’s and Burger King.

Yet branding is essential to every presidential campaign, because in order to win the people’s vote you first have to get their attention. As we noted, that’s not easy:

“In their efforts to get our attention and win us over, the candidates are competing not just with each other, but with the teeming thousands of brands that populate everyday life.”

To make the sale, Romney and Obama have used the most sophisticated tools of the marketplace, including the more than $1 billion worth of television ads that filled the airwaves in the crucial swing states.

How have they done in the brand war? Social media helped us answer this question. Growing numbers of people go on social platforms regularly to talk about every imaginable subject, including both consumer brands and political ones.

So, using analytical tools developed by Bluefin Labs, we tracked how Brand Romney and Brand Obama have performed, in terms of social-media awareness, against about 600 of the largest consumer brands.

Last summer, before the party conventions and the debates, we were surprised to see the political brands already dominating. In our first analysis, Obama was at the top of the list, followed by the iPhone in second place and Romney in third.

In September, on the heels of the GOP convention, we ran the data again and discovered that Romney had moved up to second place.

Now, on the eve of the election, the top two brands are the same: Obama first, followed by Romney. But since the last update, the presidential contenders have pulled far away from the competition. The iPhone drew less than half as many social mentions as either candidate during the three-month period covered - which included the release of the iPhone 5.

Meanwhile, there are two new arrivals in the Top Ten: Paul Ryan and Joe Biden. The vice presidential candidates are, in their own right, now top brands. This is due, at least in part, to the vice presidential debate, which sparked a great deal of social discussion.

More broadly, Biden and Ryan are important players in an election that has riveted the social audience. Even the iPad couldn’t compete with that.

To read our caveats about political brands, please see the first post in this series.

There was no Election 2012 debate last week, but social media behaved as if there had been one. What could possibly have the same effect as those widely watched debates?

Hurricane Sandy, that’s what. Even before it slammed into the East Coast, the super-storm was reshaping the election conversation, putting a greater emphasis on Policy.

As we’ve shown in our previous Three P’s posts, all the debates (three presidential and one vice presidential) bumped Politics - the so-called “horse race” - down among the three topics contending for dominance.

The presidential debates were such substantial conversations, each drove Policy to the top of the list (in the first debate, the Policy spike happened after a two-day delay, as we explained in our post). The vice-presidential contest produced a spike for Personality, due to high-volume commentary about Joe Biden’s debating style and behavioral tics.

As the graph shows, in all four cases, the shift was clear but temporary. Eventually, after a few days, Politics resumed its dominance.

As Sandy was approaching the East Coast of the U.S., there was a burst of tweets about its possible election implications. Many commenters speculated about the possibility that the storm would be politicized. These remarks were classified under Politics by the analytic tools that help us track the Three P’s.

But once Sandy hit, Obama’s response to it, along with discussion of Romney’s positions on crisis management and the Federal Emergency Management Agency, drove a spike in Policy that you can see on the graph. That Policy spike peaked the day after the storm and then began to decline.

Almost inevitably, as Election Day approached, chatter about polls and swing states began to grow. This is reflected in the rise of the Politics line at the far right end of the graph.

Now, as Americans prepare to cast their votes and the mainstream media focus again on the race itself, will Politics resume the top spot? After the votes are in, we’ll take another look and report back.

Social media are transforming politics in surprising ways. The new tools aren’t supplanting the old one that has defined and delivered our elections for decades - television. They’ve moved in beside TV and are altering how we use it.

For a better understanding of this shifting landscape, we chatted with Andrew Heyward, the former President of CBS News (1996-2005), who now advises media and technology companies including Bluefin Labs, home of The Crowdwire. Straddling two eras,

Let’s start with the role of television in this campaign. How has it changed from a decade ago?

The role of TV is very similar to ten years ago. What’s changed is that TV is no longer the only place where important impressions about the campaign are being created. Television pretends to be a conversation, but it’s mostly a conversation between insiders, with the occasional insertion of a vox populi - and that’s often canned.

What’s great about the current environment is that it’s what the Founding Fathers had in mind, except at an inconceivable scale. Social media allow anyone to join the conversation. It’s had a wonderful democratizing influence, freeing people to express their opinions outside the gatekeepers of TV or newspapers. At the same time, it’s a fantastic tool for sociologists and political analysts to observe the campaign as it evolves.

And social media have a symbiotic relationship with TV.  It’s TV that exposes the candidates and issues to the public. Then social media allow ordinary citizens to talk back to the TV and to talk amongst themselves.

You mentioned that TV’s basic function hasn’t changed all that much. What else hasn’t changed? 

The campaigns are essentially packaged for TV, as they have been for decades. The candidates themselves are carefully groomed for TV consumption, and so is almost everything they say. They’re now like ultra-sophisticated media companies using multiple technologies to reach voters. But television is still the most powerful by far.

What did you think of the debates, not just as electoral events but as TV?  Were they good TV?

I think they were fine, adequate. The producers are hemmed in by the fact that the campaigns negotiate almost every aspect of the programs - the format, the setting and so on. Yes, they’re live on television, but they’re not anywhere near as freewheeling as the word ‘debate’ would suggest.

If Romney wins, that first debate is going to take on historical importance at the level of Nixon and Kennedy in 1960. The race changed significantly as a result. President Obama had an opportunity there to solidify a victory, as opposed to putting a victory in doubt.

One other noteworthy aspect of the debates, and you pointed this out on The Crowdwire, was that a significant portion of the comments about the debates were devoted to policy.

Were you surprised at the size of the TV audience - the fact that nearly 70 million watched the first one? 

I was pleasantly surprised but not terribly surprised. There are very significant differences between the candidates. I think people are sick of commercials, canned appearances and pundits, and the chance to actually see the candidates themselves on the stage is very powerful and appealing.

More generally, do you see downsides to this media transformation?

I think there are several. One aspect of the symbiotic relationship between social media and the commentariat in general, is that the reaction is now so swift and large, it’s actually contributing to wilder swings of the pendulum than we had before. The scrutiny that every single utterance gets, and the sheer amount of analysis and commentary, has the potential to turn hiccups into bellows and roars.

Arguably, the negative effect for the Obama campaign from the first debate was a result as much of the reaction to the debate as to the debate itself. Because not everyone watched it. If you didn’t watch it and only heard about it, then I think you might get an even more extreme impression of one-sidedness - Obama took a drubbing - than if you had watched it. There’s a kind of echo effect that might be distorted

Another possible negative, related to the first, is we don’t really know yet the degree to which the Twitter-using population accurately reflects the population at large. Yet because it’s so tempting to analyze the data and because in fact there are very interesting observations to be made, there’s always a risk that the Twitter-using population has a disproportionate influence on people’s perception of how the race is going.

A final potential downside is the possibility that social media can be gamed or manipulated. But again, when you analyze social media as an indicator of trends, you just have to be aware there are these potential pitfalls.

How have social tools changed the practice of political journalism?

We’ve already talked about a couple of them. Twitter has become a significant source of not just sentiment but content. Journalists, as users of Twitter, are also publishers and content creators themselves. This is a very effective way for journalists and analysts to share their views in a pithy and instantaneous form. It’s a very valuable way for those who cover the race and think about the race to share content.

The upside here - and also the downside - is that Twitter forces you to condense your insights and observations to 140-character units. You could argue that that imposes a wonderful discipline or that it creates a forced superficiality.

It’s also an enormously effective way to distribute longer-form content through links. That’s new and important. It’s a significant form of distribution and content- awareness-building, and not just for the short-form stuff.  

During debates and other big, televised campaign events, do you keep one eye on Twitter and/or Facebook in real-time?

I don’t happen to do that, because I really like the idea of focusing and forming my own impressions. But I see why that’s an appealing way to watch these events for a lot of people. And certainly on election night I will be at least a two- and probably a three-screen person. That’s the way the world is going.

Among the people and organizations you follow on Twitter, do you have a favorite? 

No, to me what’s fascinating is the overall phenomenon rather than the particular insights of one or two individuals. Right now, the whole is more interesting than the parts.

We won’t know until after Election Day what impact TV ads had on the outcome. But there’s no question that ads are changing the nature of politics, thanks to huge increases in fundraising levels and the rise of super PACs. The question is how they’re changing it, and it’s not easy to answer.

Enter the Wesleyan Media Project, a continuation of the University of Wisconsin Advertising Project. From 1998-2008, the Wisconsin project studied how candidates, political parties, and special interest groups communicated with voters. Now based at Wesleyan University, the project is still run by three of the original graduate students from Wisconsin, now teaching at other universities.

We interviewed co-director Michael Franz about the project’s methods and findings.

What is the basic mission of the Wesleyan Media Project?

Second, we archive the data and eventually make it available to scholars for use in their research. Together with the Wisconsin Advertising Project, we have an archive of political ads that stretches back to 1996 and accounts for over 14 million airings.

How does your ad tracking system work?

We buy the raw data from a commercial firm called Kantar Media/CMAG, which has ad trackers in all 210 media markets across the country. These trackers capture and record ads every time they air on local broadcast stations (i.e., local ABC, NBC, CBS, and Fox affiliates) or on national cable.

We get the data on the ad airings weekly, and each individual ad in video form. We then code each ad on nearly 100 dimensions, such as its tone, the issues covered in the ad, and the identity of the sponsor. We even code for whether the ad is likely making an emotional appeal such as humor or fear. By appending our coding to the frequency of ad airings, we can then see what kinds of ads are airing where, by whom, and at what levels.

How does your coding system work? Do you do it by hand?

We code these by human hand, and the tone categories are essentially objective. If an ad only mentions an opponent, it’s a negative ad; if it only mentions a supported candidate, it’s positive; and if it mentions more than one candidate, it’s a contrast ad. Our inter-coder reliability on this measure is very high. For the emotions, we do a lot of training with coders about how to identify those appeals, and as expected, our reliability there is lower. We get strong agreement across coders for emotions like anger, fear, and humor. Pride and enthusiasm are a bit harder to detect or code consistently. We are looking at ways to automate this, and some political science research is arguing for an automated approach over a human coder.

What are your long-term goals for the project? What have you accomplished so far?

Our long-terms goals are to continue adding to the archive in subsequent elections, and also disseminate our results to the broader public. With outside groups increasingly prevalent in federal elections, it is imperative that voters know to what extent wealthy groups and donors are bankrolling persuasion efforts on television. We can speak to that—without any partisan or ideological goal on our part—by simply providing voters with information on who is advertising and at what levels.

To date, scholars have used the ad data from the project in dozens of scholarly articles, book chapters, and books. The data have allowed political scientists to develop and test sophisticated theories about how ads work in convincing voters to vote, and to see if ads have any effects on turnout or knowledge of key issues. Prior to this, the data for testing these effects were indirect and incomplete.

Do you have any hard results to share from this research? Do certain types of ads work better than others?

Great question, and there is no easy answer. It seems pretty clear that ads “work” in the sense that a great number of ads can move votes in your direction. But because many campaigns wind up with an equal number of ads for each side, the net effect is often zero. Travis Ridout and I have a book on this, The Persuasive Power of Campaign Ads, that uses data from the 2000, 2004, 2006, and 2008 elections. And political science has generally reached consensus that ads do not depress turnout; they might even stimulate interest in the election and raise turnout in small amounts. As to what “type” of ad works, the work continues.

What trends have you seen with political ads for the 2012 election?

This presidential race has had more ads than any previous election; both candidates, especially Obama, have aired an incredible number of negative ads; and outside groups, especially in support of Romney, have been extremely active. Even in congressional races, the level of spending from outside groups has been very high. This has been a very interesting election, one that has broken lots of records with respect to political ads.

How does the political ad landscape in swing states compare to other areas of the country?

In the presidential race, it’s simple. If you live in a battleground state, you are probably sick of hearing candidates approve their message by now. If you don’t live in a battleground state, you are being ignored. The only exception is one particularly interesting feature of media markets across the country—they don’t always line up with state boundaries. So some voters in non-battleground states (like Maine or Massachusetts) see ads because of their proximity to battleground states (like New Hampshire, which is covered in part by the Portland, ME and Boston media markets).

Of course, House and Senate races are a lot broader geographically, so voters in lots of places across the country are seeing ads in competitive congressional races.

What is the most surprising thing you’ve learned so far in the 2012 election through this project?

The level of outside spending is astonishingly high. 44% of all pro-Romney ads in October and 33% of all Senate Republican ads in October have come from outside groups. Obama, by contrast, has aired in October about 86% of the ads that support his candidacy. But in comparison to 2008, the level of pro-Obama ads coming from outside groups has risen from about 4% to about 14%. We never expected to see so many congressional candidates, and a major presidential candidate, so dependent on outside spending.

Have you seen anything noteworthy in the Massachusetts Senate race, where the candidates agreed not to allow outside groups to advertise?

That’s what’s interesting. The Brown-Warren pact set up ground rules for trying to prevent outside groups from airing television ads. That’s seemed to work so far. Of course, there are loopholes in the agreement, and outside spending is fairly aggressive in that race, just off the air. We’ve been surprised, though, that the two candidates were able to work out an agreement that made it difficult for groups to air ads during the race. It could conceivably be a model for future races, though it requires both candidates to sign on—and that’s the sticking point.

Does any part of your project involve speaking with the general voting public about their experience with TV ads? If so, what have you found?

We love talking about our work to the general voting public. We generally find that voters express exhaustion at the level of ads on television. Our response, though, is a bit of a defense of such ads. Political ads are still one of the only ways for candidates to speak to a wide spectrum of the public at the same time. Increasingly, voters are consuming less news through television and in print form, and they are getting their news in much more partisan places (especially online). Ads are attempts to speak to all voters, and they can sometimes be the only place for a Republican voter to hear what the Democratic candidate has to say, and vice versa.

Are there any non-political applications for what you’re doing?

One particularly neat application would be to look at the level and intensity of ads for pharmaceuticals. Anyone who watches the nightly national news understands that ads for prescription drugs are almost as ubiquitous as ads for candidates. It would be interesting to compare the frequency of such ads to the demographics of the programs on which they air; and even to see if sales of drugs rise and fall with the frequency of ads.

The last time we checked in on The Three P’s, the final presidential debate had just taken place. That very issue-oriented discussion gave Policy a nice spike in social media, but we weren’t sure how long it would last.

As the graph above shows, it lasted about a day, an emerging pattern. Politics - the horse race - tends to have the greatest share of the social conversation. But occasionally, when either Policy or Personality gets a boost from a specific event such as a debate, it takes the lead for about one day. Then it’s back to Politics, as usual.

Our latest analysis also brings a surprise. For the four-week period shown in the graph, we isolated the social comments of people we’ve previously identified as either strongly Pro-Obama or Pro-Romney (classified as such based on their own previous public comments about the candidates). These are large groups, with tens of thousands in each one.

For these relatively partisan people, we found that during this period Policy had the greatest share of the commentary. As the chart below shows, partisans gave Policy a 48% share of the conversation. Yet among all election commenters, Policy had just a 42% share, while Politics was on top at 44%.

In other words, the very citizens one might expect to be the most political of all, and therefore the most interested in the horse race, turned out to be even more engaged by Policy.

How to explain this? The 6% gap between the two groups isn’t huge, but something has to be driving it. We’re just not sure what. We know partisans tuned in heavily to the debates, which were very issue-driven, and this presumably shaped their comments. Still, it doesn’t explain why they would talk about Policy more than everyone else responding to the same televised exchanges.

If a better explanation emerges as we continue tracking The Three Ps, we’ll let you know.

We’ve been documenting the explosion of the election conversation over the last two months. The three presidential debates alone sparked more than 30 million tweets and public Facebook comments.

That’s a lot of talking. But is the social crowd doing more than just talking? After all, if this is a true conversation, there has to be some listening involved. Are all those tweeters also listening to each other?

In fact, many people only listen in social media. Twitter says 40 percent of its users never tweet. How does Twitter know this? They see these folks signing on and not commenting. They must be going on Twitter to read what others are saying. But determining how much and what they’re reading is a challenge. It’s easy to sign on to a network, then get distracted and leave it open on your screen behind other tasks. Some of the silent 40 percent might not listen very much.

Then there’s the majority that does tweet. How many of them also take the time to read and consider others’ expressions?

There is one way that people reveal they’ve read others’ comments: retweets, or the forwarding of others’ tweets to one’s own followers. Most retweets are preceded by at least a few seconds of reading (the exceptions would be tweets that are blindly retweeted in order, say, to boost someone’s exposure on Twitter). Thus, retweets offer a basic glimpse into how much people on Twitter are taking in others’ points of view.

So we set out to analyze retweet behavior during the debates. And since women in key states are the most sought-after voters in this election, we thought it would be interesting to see if listening habits vary by gender.

As the chart shows, a significant percentage of those who tweeted and posted about the debates also retweeted the remarks of others. The portion who did so ranged from 41% (Debate 2) to 52% (Debate 1). It makes sense that the debate that had the greatest impact on the race would also have the most retweeters. Astounded by what was unfolding, some viewers would have been eager to see if others shared their perceptions, and in certain cases, to share what they’d read. A few of the most widely retweeted comments from the last debate are shown here.

But the real surprise is the gender breakdown: Women retweeted at significantly higher rates than men. In the first debate, there was an 18 percentage-point difference between the two.

Are women just better listeners? Or is there something about this election that’s making them listen more? Perhaps those undecided female swing voters who have gotten so much attention lately are just trying to make up their minds.

Who are Romney’s supporters? Who are Obama’s? How do they live? Where do they shop? What do they like to eat and drink?

These questions intrigued us, and we didn’t know the answers. So we spent the last few months combing tweets and public Facebook posts looking for people who were clearly pro-Obama or pro-Romney. We found a lot of them, tens of thousands in each group. Then, using Bluefin Labs tools, we analyzed their past comments and learned a great deal about their tastes and lifestyles. 

To tell the story, we got together with the talented folks at Now This News, who made a terrific animated video showing the results. It debuted today on The Atlantic’s website, accompanied by a smart explanatory piece by Kasia Cieplak-Mayr von Baldegg.

Enjoy.

In our previous Three P’s post, we reported that Policy has been having a nice run. Hours later, the final presidential debate took Policy’s share of the election conversation to a new high. Note the spike at the far right end of the chart.

Will this debate resonate in the coming days and keep Policy on top? We’ll take another look soon and let you know.

Meanwhile, to get a flavor of what drove this policy-fest, here are the words most frequently used in tweets and posts about the debate, listed in order of frequency. Is the social sphere just for frivolous chatter? You decide:

romney, obama, debate, mitt, president, policy, foreign, vote, america, people, back, israel, china, presidential, debates, years, iran, talking, plan, mr, question, bin, laden, bayonets, military, governor, horses, jobs, barack, teachers, bob, foot, point, voting, left, war, moderator, fact, potus, stomp, country

It would be easy to call presidential debate No. 3 a social-media flop. Too easy. Its 8 million tweets and public Facebook posts look small in comparison to the previous two debates, but this conversation was very large compared to nearly everything else. 

Just last month, the 2.5 million tweets and posts sparked by the Democratic National Convention made headlines. Today, it’s a blip. That’s how quickly the social conversation about politics is growing. Collectively, the debates have been a landmark in that growth. Key Facts:

      • Making the Lists: The final debate was the third most commented-on political telecast ever, behind the other two debates, and the seventh most commented-on telecast of any kind (more stats from Bluefin Labs here).

      • Policy Over Sports: This debate was competing for viewers with two major sports events, Monday Night Football and Game 7 of baseball’s National League playoffs. Though sports is a magnet for social comments, the two sports conversations were miniscule in comparison to the debate. The election face-off accounted for 79% percent of all social commentary about the evening’s primetime TV. A foreign-policy debate over sports? Further confirmation of social media’s wonkish tendencies.

       • Seriously: Like the previous debates, this one was mostly about serious policy matters, and the social response appeared to mirror this. Even the viral phrases were weightier. After Big Bird and binders full of women, #horsesandbayonets is positively brainy. For the exact breakdown of how politics, policy and personality fared in the debate response, watch for our Three P’s update.

       • More Romney Mentions: In all three debates, Romney was mentioned more often that Obama. This does not imply that he won over more viewers. Social mentions can go either way.

      • Women: In all three debates, a majority of the comments came from females. This was unexpected, given that males dominated the social response to the earlier GOP primary debates and the Republican National Convention. With women widely viewed as the deciding factor in this election, it’s significant that they’re now tuning in and speaking up more than men, if only marginally.