The Hybrid Media System

by Andrew Chadwick

  The Camp Obama Assemblage

  A significant chunk of the central campaign’s “strategically sound” resources for volunteers came in the form of Camp Obama. These gatherings were held across the key swing states during 2007 and 2008 (for some examples of the typical program of each event from early in the campaign see Camp Obama, 2007). Camp Obama combined strict campaign goals with looser, though still focused, principles of community organizing that had been important in forming Obama’s political identity in Chicago during the 1980s and 1990s. Campaign goals included targeting, registering, and turning out key voter groups, such as Latinos in Denver or African Americans in North Carolina. The community organizing principles were injected by Marshall Ganz, the renowned veteran of 1960s community movements, who acted as an adviser to the Obama campaign and who personally led several of the Camps.

  The Camp Obama gatherings are best understood as unusual assemblages of organizational rules and capacities, mobile technologies, and goal-oriented coordinative behaviors. They integrated people, local community buildings, training manuals, and ideas about experiential learning. They also structured opportunities for personal self-discovery and the formation of emotional bonds among activists.

  Consider the camp held in Denver, Colorado, a couple of months before election day (see the wonderful ethnographic account in Alexander, 2010: 45–59). Around two hundred Americans of Hispanic origin from across the state converged in a community center in a working-class neighborhood. They were joined by twenty of the Obama campaign’s full-time salaried field organizers, around a dozen state organizers, and a similar number of national staff who were representatives from campaign headquarters in Chicago. The goal was to register twenty thousand new Hispanic voters in the twenty-six days before the state registration deadline. The means to do so was hundreds of small local teams, each of whose job it was to call, canvass, and complete registration forms for Hispanic voters in their neighborhoods. The national organizers distributed a detailed grid of tasks to the local field staff, but these were not handed to the wider gathering of volunteers for fear it would give the impression that orders were simply being handed down from the top; which was, of course, precisely what was happening.

  There followed a day of workshops and breakout sessions in which the principles of community organizing were spelled out and personal stories were shared about what the campaign meant to each volunteer. As the leader of the sessions explained, organizing was not about “lecturing” people, but about sharing stories and quickly building trusting relationships with strangers. Written on the chalkboards at these meetings was a series of binary opposites that summarized Ganz’s model in shorthand terms. Instead of allocating “tasks,” participants were told that they should encourage “stories.” Don’t ask “what?,” they were told; instead ask “why?” or “how?” But even in this intensely face-to-face context, older media logics intervened: motivational videos of Obama’s most powerful speeches were shown to the group on television before they were handed their goals.

  Canvassing was organized in shifts from 9:00 a.m. to 9:00 p.m. each day. Individuals signed up for shifts and agreed to meet face to face every week for the remainder of the campaign. Role-playing exercises provided opportunities for the newly educated to put these principles into practice before the watchful eye of the camp leader and the audience. The teams were instructed to give themselves distinctive names, to develop their own group poems or songs, and then to perform these in turn in front of the whole gathering. Despite the emphasis on the principles of indirectness and trusted relationships in community organizing, the Camp Obama teams were instructed to call up to fifty friends and family members on their cell phones right there and then and invite them to a house party for Obama. The field organizers told the volunteers that the parties should follow a common format: the inspirational Obama video should be shown, and this should be followed by group conversation and voter sign-ups. Each Camp Obama group then reported its phone contact statistics back to a meeting of the whole audience in the community center hall. Following a break for dinner, the meeting reassembled for a speech by a leader from the national campaign’s Latinos for Obama division. The leader ended by reminding the group about the twenty-thousand newly registered Hispanic voters the Colorado campaign must acquire. Finally, a graduation ceremony was held and volunteers were awarded diplomas and T-shirts.

  Crucially, and beyond the scope of Alexander’s ethnography, the practices of the physical Camp Obama gatherings were transcoded into the online part of the campaign assemblage. A few days later, the national campaign launched “Neighbor-to-Neighbor,” an online phone-banking platform produced by political consultants Blue State Digital. This enshrined in digital form many of the principles of community organizing at Camp Obama (BarackObamadotcom, 2008d). When a supporter joined an Obama group on Facebook, for example, they were soon contacted by a campaign volunteer asking for a phone number. Information from a person’s profile was used to spark conversation as a prelude to a request for volunteer engagement (Cornfield, 2010: 224). Camp Obama and Neighbor-to-Neighbor thus constituted the hybrid integration of real-space and online community organizing.

  Digital Media and Control

  The more it progressed, the more Obama’s campaign was animated by an impulse of control. As Daniel Kreiss has argued, it combined new digital technologies with “the bureaucratic objects that were part of the imaginary of the nineteenth century: the social organization, management structures, large-scale coordination, and meticulous planning” (Kreiss, 2012). This was an older organizational logic combined with a newer media logic of control. What Kreiss terms “computational management” enabled a shrewd awareness of the comparative opportunity costs and calculable returns associated with different types of campaign activity. During the primaries, the campaign’s leadership used its polling and sign-up data combined with its estimated television and other paid media costs to devise “cost-per-delegate” formulas that assessed the utility of campaigning in one state versus another. After all, the primary was eventually decided, not by simple victories based on the share of the vote in each contest, but by the numbers of delegate seats each candidate had managed to accrue in each state.

  The general election was similarly amenable to this newer media logic of control. Share of electoral college votes were what mattered in that context, but in 2008 there was also an unprecedented amount of early voting, sparked in large part by the Obama campaign’s massive and continually calibrated education program encouraging young and sporadic voters to vote early. A third (34 percent) of the electorate cast their votes before election day, an increase of 14 percent on 2004 and 20 percent on 2000 (Kenski, et al., 2010: 255). Eleven of the fifteen key battleground states had some form of no-fault absentee voting or early voting. And these early voters broke decisively for the Democrat: of the 6 percent who cast ballots three weeks before election day, 62 percent supported Obama but only 37 percent supported McCain. Swing voters who voted early were also more likely to vote Obama. And those who received e-mails from the Obama campaign were significantly more likely to vote early. Receiving an e-mail from McCain made no difference (Kenski, et al., 2010: 260–262).

  It was the extensive face-to-face ground campaign combined with targeted e-mails that created such a powerful advantage in the struggle to get people to vote early (Kaye, 2009: 17). This advantage was felt most keenly during late October, when McCain started to make inroads on tax policy. By that stage, large numbers of voters had already declared their preferences and the majority of these had voted for Obama. The Democrats thus exploited a huge financial advantage at ground level to run what was a effectively a rolling program of get-out-the-vote mobilization across the last two months of the campaign.

  Higher quality data than had previously been available to Democratic campaigns proved remarkably useful for allocating technological and human resources. The Democrats had learned from the Republicans’ approach in 2004, when George W. Bush’s
campaign chief Karl Rove had orchestrated the “72-hour strategy,” a massive word-of-mouth get-out-the-vote drive that saw Bush supporters contact their friends and neighbors in the final hours before election day. By 2008, the Democrats had their first truly national voter file (Finney, 2009). This was built between 2004 and 2007, in large part by party chair Howard Dean and his team of assistants, including Joe Rospars. Dean aggregated the state-level party databases, provided a unified mode of access to the data, and offered it to all Democratic candidates during the 2008 race. Drawing upon three microtargeting, demographic, and supporter modeling databases—Votebuilder, Catalist, and Strategic Telemetry—the Democrats were much better equipped than they had been at any time since this type of tool made its debut in the 1990s (Kreiss, 2012). Rospars and the new media division capitalized on their previous experience with Dean, but by 2008 they had the resources to construct a set of robust practices for effective online campaigning. They focused on building direct relationships with individual voters and avoided having Obama’s schedule dictated by appearances at events organized by local Democratic Party committees for the benefit of the party faithful. These were not the kinds of electors to which Obama needed to appeal to defeat the Republicans. This created time for the campaign to create its own local events, many of which did not require donations as a condition of entry. Not only could these events attract those who would feel uncomfortable paying a donation to attend a political meeting, but also, more importantly, the campaign could use the information gathered through the sign-in process to build data profiles of those in attendance. These profiles were then compared against demographic targets that needed to be met if Obama was to have a chance of expanding the electorate in the key states (Plouffe, 2010: 71). The model proved its worth.

  In addition to using e-mail as a means of spurring donations and volunteering, the war room issued several much more detailed text and video messages during the course of the campaign. Initially these were aimed at professional journalists as a means of convincing them of the viability of Obama’s candidacy. But it soon became obvious that there was an appetite for them among the volunteer base and online blogger networks. These memos contained a mixture of stimulating rhetoric, historical facts from previous elections, snippets of internal polling data, surprisingly candid projections of the race ahead, and grainy video clips from inside the office (the slickly produced ones were criticized by supporters). They served to educate volunteers about the leadership’s strategy and they generated interest by pulling back the curtain on what were supposedly the inner workings of campaign headquarters. Just as importantly, they softened the impact on donors and volunteers of journalists’ narratives, whose skepticism about Obama’s challenge to Clinton continued until very late in what was already an unusually long primary season. These e-mails also provided a counterpoint to the “poll-of-polls” information that now features in withering detail on websites like RealClearPolitics.com and Pollster.com. These sites emerged as important reference points for broadcast, newspaper, and online media players during the 2008 cycle. Providing credible counter-analysis and detailed projections complete with concrete numbers was an effective way for the campaign to try to reduce uncertainty and therefore the thresholds to participation for potential supporters. For example, as voting day loomed, the campaign sent out an extraordinarily detailed e-mail presentation in which they cataloged how it was going to spend all of its gigantic $38 million budget in Florida (Plouffe, 2010: 330). The risk that they would be spilling secrets to their opponents was perceived to be offset by the gains in commitment and trust that would come from a policy of transparency, even though it was made equally transparent that this was not an opportunity for the Florida grassroots to voice their opinions on precisely how the campaign’s money should be spent.

  As the primary campaign progressed, it became evident that even when he was losing in the headline-grabbing vote share contest, Obama was still picking up large numbers of state delegate votes: the real hard currency of Democratic primaries. The leadership’s analytical e-mails then became part of a strategy to educate elite journalists about the need to focus on the details when reporting on primary results. The campaign tried to “shift the national narrative from raw votes to the question of delegates” to address the “electability” criticisms that were putting doubts in the minds of the superdelegates and the Democratic party leadership (Plouffe, 2010: 157, 200). Here was a case of using the internet, supposedly a medium primarily for mass mobilization and the ground campaign, to target messages to journalists and political elites using a newly invented genre: the detailed field report e-mail.

  Yet the complexity of the delegate rules, which was further vexed by the potentially free-floating nature of the Democratic party superdelegates, seems to have defeated even the most experienced older media reporters, necessitating more traditional methods of intervention. Frustrated at the New York Times’ miscounting of Obama’s true support among delegates, the war room made direct contact with the paper’s highly respected political reporter, Adam Nagourney. They suggested to Nagourney that learning from the campaign about the delegate race math might pay off in the form of a scoop for the reporter: Nagourney could get one step ahead of the press pack. Nagourney duly agreed to a conference call involving Obama’s team, the paper’s political editor, and its polling unit. During the call, Plouffe explained the intricacies of the delegate system and how Obama was pulling ahead of Clinton where it mattered, in numbers of “pledged delegates.” He argued that the superdelegates would feel duty-bound to follow the lead given by the pledged delegates. Following the meeting, the Times agreed to start presenting the delegate counts differently. But more importantly, a few days later Nagourney published a big story that did much to force home the point that Obama had been picking up delegates even in those states where he had ostensibly “lost” to Clinton (Plouffe, 2010: 185–186).

  This intervention—a good illustration of the enduring relations of interdependence that exist among campaigns and professional journalists—had an important educative effect, not just on public opinion but also on political elites in Washington. The message was reinforced inside the campaign through a flurry of e-mail memos. Nagourney’s New York Times article was a genuine turning point in the race. It increased the pressure on the Clinton campaign by reframing the story around her need to “catch up” with Obama. It paved the way for a decisive television news intervention by another journalist in early June when ABC News’ Charles Gibson was the first to declare that Obama had secured the nomination for president and that it was now impossible for Clinton to win enough support (ABC News, 2008). Meeting with Nagourney was a classic moment of older media logic for the Obama campaign, but one that was first made possible by newer media logic—the e-mail field reports.

  Outcomes

  What were the outcomes of this combination of older and newer media logics? Pinpointing campaign effects is a minefield, and is not my main purpose here. But we can identify some areas where the strategy of symphonic consonance is highly likely to have made a difference to the outcome of the 2008 presidential contest.

  Obama won 52.9 percent of the popular vote in the general election. Second only to Lyndon B. Johnson’s 1964 landslide, this was the largest Democratic majority of the modern era, and well ahead of McCain’s 45.7 percent. Of the states won by Bush in 2004, Obama took back a total of nine: Colorado, Florida, Indiana, Iowa, New Mexico, Nevada, North Carolina, Ohio, and Virginia. Ohio, Nevada, Florida, and Colorado had not been in Democratic hands since the Bill Clinton era. The victories in North Carolina, Virginia, and Indiana were genuinely surprising. North Carolina had been Republican since 1980, but Virginia and Indiana had been solidly Republican since 1968. The national swing was 9.6 percent, and in some states, such as Indiana, Nevada, Michigan, and Pennsylvania, the swings were large by historical standards. Turnout was 61.6 percent, a significant increase of 1.5 percent on 2004 and a revival of participation levels last seen in the 1960s
, though turnout was perhaps not as significant as expected given Obama’s ground campaign. The headline turnout figure conceals a more complex situation, however. The increase in turnout for the Democrats, up ten million votes from 2004, was undoubtedly real. But the overall increase in turnout was tempered by a two million-vote decline in turnout for the Republicans (Cook, 2010).

  Young voter turnout (18–29-year-olds) increased by about 6 percent in absolute terms, though, again, this is not as impressive as it seems if it is considered as a proportion of the overall electorate (Sabato, 2010: 46). Still, 66 percent of 18–29-year-olds voted for Obama, against only 32 percent for McCain (Barr, 2009: 114). Seventeen percent of self-identifying conservatives voted for Obama, and 85 percent of African American conservatives did so. Seventeen percent of those who voted for Bush in 2004 switched to Obama, but only eight percent of those who voted for Kerry in 2004 switched to McCain (Kenski, et al., 2010: 22). Given that the goals of the campaign were to target these constituencies and expand the electorate, these were successes.

  The Obama campaign eventually raised an unprecedented $750 million, more than twice the amount raised by George W. Bush in 2004 (Toner, 2010: 149). About $200 million out of a total of $414 million for the primaries alone was raised online (Toner, 2010: 149). McCain managed to raise $221 million during the primaries but he was constrained to just $85 million in the general election due to his necessary acceptance of federal funding. This meant that McCain was outspent by a ratio of almost 2.5 to 1. Obama had a total of six thousand employees working for him by the end of his campaign, 95 percent of whom were under thirty years old (Plouffe, 2010: 370).

  By the summer of 2008, Obama’s ground advantage was clearly established. In key states, Obama’s field offices far outnumbered those of McCain. For example, in Ohio the ratio was 33 to 9; in Virginia it was 27 to 7 (Abramowitz, 2010: 103). Overall, the Obama campaign and the Democratic party had around 770 field offices and spent $56 million on paid staff. McCain’s field offices numbered 370 and his staff expenditure just $22 million (Toner, 2010: 154). The results speak for themselves: the Democrats increased their registered supporters by 2.9 million, compared with the Republicans’ increase of 1.5 million (Kenski, et al., 2010: 19).