The GOP’s Broken Machine
It was just past dawn on Election Day, and already the whale was dead in the water. Project Orca, the not-so-secret high-tech weapon of Mitt Romney’s 2012 campaign, was drowning in the deluge of thousands of attempted log-ins from volunteers across the country. The program was designed to be a 21st-century update to old-school poll watching, driven by more than 30,000 smartphone-equipped volunteers. The digital equivalent of scratching thousands of voters’ names off of a gigantic list, Orca was intended to give the Romney campaign staff up-to-the-minute insight into who had and hadn’t made it to the polls, allowing them to target their robocalling more precisely and avoid pestering those who had already voted. Instead, Orca became one last failure from a campaign that never managed to get the little things right because it couldn’t get the biggest thing right.
Orca was a near-total operational failure. The program had been kept under wraps to its detriment, never fully stress-tested for the crunch of Election Day traffic. Volunteers who had tuned in to cheerleading conference calls with campaign staff never received their information packets on how to use it, and simple coding failures meant many people couldn’t access the application at all. Calls to the relevant help lines went unanswered. And the checklist given to volunteers had somehow failed to remind them to bring the credentials they needed to monitor the polls legally, though it did tell them, twice, to bring a chair.
For Republicans depressed by the election results, Orca was a symbol of their despondency at the Romney campaign’s surprising failure to even come close to matching the get-out-the-vote efforts of the 2008 McCain campaign. Romney aides stepped up to defend the program, telling National Review’s Katrina Trinko that Orca “has no relation to the outcome,” adding, “we achieved in a large part what we set out to do in the swing states in terms of our electorate.”
Given that fewer than half a million votes in four key swing states ended up being the difference between victory and defeat for Romney, this comment will provide little solace to disappointed Republicans. But in so far as it goes, it happens to be true. The Romney campaign raised Republican turnout in seven of their eight target battleground states, largely hitting the marks they thought they needed to win the 2012 election. Its assumptions simply turned out to be woefully inadequate, a failure of strategy and operation that left the campaign ill-equipped to battle President Barack Obama’s tech-savvy get-out-the-vote machine. Campaign staff acknowledged that from the beginning they assumed they could not match Obama’s well-built machine toe-to-toe, so they didn’t try. In retrospect, Romney needed to match Obama’s technology advantage in order to have a real chance at victory. But even if he had, it may not have been enough.
From the start of the general-election season, Romney’s national campaign manager, Matt Rhoades, repeated to his Boston staff a mantra inspired by the legendary order given at nearby Bunker Hill: “Don’t fire until you see the whites of their eyes.” In the last month of the campaign, much of the staff thought they’d done just that. “We’re f—ing gonna win this thing,” Romney pollster Neil Newhouse announced boldly in a staff meeting. His latest numbers, taken in the wake of Romney’s strong debate performances, showed the ticket with a strong lead among independent voters. The campaign’s internal polling assumed election turnout would favor the Democrats by only two or three percentage points—and in such a scenario, the performance among independents had Romney’s circle buzzing about victory.
Actually, it was more of a murmur. Obama’s Chicago team was housed in a gleaming spire, its offices packed with wall-to-wall whiteboards and its chief technology officer culled from the irony-clad hipster site Threadless. It was a place bursting with the youthful verve of a tech startup. Romney’s Boston digs were more subdued, with little office conversation and an almost eerie quiet during the day. Romney’s was also a house divided, in the literal sense: Staffers were in two buildings, the political and digital staff separated in a fashion that led to more than a little frustration for both sides during the course of the campaign. While digital efforts were the primary focus of the Obama campaign from the beginning, with data miners and tech gurus culled from Silicon Valley, they were a relatively late addition to the Romney effort. Its digital operation was staffed after the rest of the campaign, with an operation that seemed remarkably inefficient for a campaign that was supposed to do things with the rigor of Romney’s research-intensive firm, Bain Capital. There were plenty of people working on the digital side, but tasks were poorly assigned and hampered by restrictive approval processes. Romney’s staff was politically diverse and more used to the world of business than politics—some had never worked on a political campaign before. Frustration set in, then boredom, then Facebook-browsing. The quiet was deafening.
For digital staffers who recognized they were playing catch-up with the Obama machine that had never stopped building after 2008, the contrasts were infuriating. Where the Obama campaign’s content and emails were tailored to the interests of individually targeted demographic communities based on topics of interest and other data-mined priorities, Romney’s campaign didn’t even make distinctions between whether someone had given $5 or $500, or whether the name came to the database through a petition about health care or energy policy.
The campaign was also fiercely hierarchical, to the surprise of some longtime Romney staffers who found their ideas for innovation shunted aside by senior staff and consultants who were unapproachable and unresponsive. Ideas for how to defend the candidate more effectively, activate volunteers on key issues, or push back harder on false attacks were met with a repeated response: “Don’t get in trouble, don’t rock the boat.”
In this, the campaign was fulfilling the risk-averse approach of Stuart Stevens, the man at the top of the campaign who built his career out of getting relatively bland politicians elected by keeping them on script and from making gaffes. Stevens is a strategist who wins by playing small ball—shrinking information about policy to the bare minimum, sticking to inoffensive bullet points, and blanketing the airwaves with ads that blend together into one long reiteration of believing in America. After the debates, even the more skeptical campaign aides felt Stevens’s strategy might work. Romney was hitting his stride at just the right time, they felt. The polls showed his “likability” gap with the president shrinking steadily. The images coming out of the campaign’s closing weeks were simply beautiful—Romney and vice-presidential candidate Paul Ryan speaking to tens of thousands of supporters at Red Rocks in Colorado, in Loudoun County in Virginia, in West Chester in Ohio. The events were Norman Rockwell Americana for The North Face set, sweeping panoramas of patriotism and pride—contrasted with the relative diminution of President Obama’s shrinking rallies, which had lower attendance and uglier backdrops.
But the visual contrast hid the important difference of purpose. While Romney was packing out fields and stadiums in suburbia, the Obama campaign was holding early-voting rallies downtown—come to see the celebrities, then go around the corner to vote.
Almost within minutes of Romney’s concession, Orca came to symbolize a campaign that mismanaged the operational details. But that ignores the truly bothersome part: Even had Orca worked, it would have been a waste of time. The approach Orca was trying to improve upon has become an anachronism.
Poll-watching is a tactic as old as Tammany Hall. It amounts to a group of volunteers sitting at a precinct crossing off names as they get ballots, a few more volunteers translating that data, and the rest working telephones to get missing voters to the polls. While such operations are still deployed by some consultants, more innovative consultants dislike the practice. They view it as a massive opportunity cost, with time on election day spent tracking votes your candidate already has instead of getting people to the polls. They consider it better to maximize the get-out-the-vote effort, deploying volunteers to walk door-to-door in key precincts rather than have them sitting at a poll location unable to increase vote totals.
Orca was designed to make robocalls work, reaching people who hadn’t voted. But it’s not as if the Romney campaign lacked the resources to use actual humans to call as many people as possible on Election Day asking them to vote—and if they already had, telling them to call their friends and get them to vote, too. Had the 30,000-plus Orca volunteers been deployed to target Republican districts door-to-door and by phone, it’s possible they could’ve made a much more meaningful difference in battleground states.
Instead of taking this approach, Romney’s campaign wasted money and man-hours on a project that ended up giving it a half-day’s worth of incomplete data from their target counties. Adding insult to injury, much of the Orca data they pulled into their election-night headquarters in Boston—before the system crashed completely—indicated the wrong result, because the campaign’s calculations about the necessary turnout targets were built into their analysis. According to one Romney aide, the assumption that it would be a narrow turnout advantage of a few percentage points doomed them from the start. A campaign that shot for just getting by was stunned by a Democratic turnout advantage that swept their data aside.
In the end, Obama’s team played small ball better. In an electoral race fought over slivers of the population in a handful of states with incredibly narrow ground, Obama’s team drilled down with specific messages tailored to each of its target interest groups, each designed to caricature Romney as a wealthy out-of-touch technocrat who doesn’t care about anyone who doesn’t look like him. Beginning in the spring, every step was built around this “destroy Romney” narrative, with flash points on issues prioritized by their constituents. The White House declared war over contraception with the Catholic Church and trumpeted the life of a dependent avatar as a stand-in for single women. The administration went to battle over student loans to activate college kids and picked fights over immigration to make Republicans look anti-Hispanic. The campaign trumpeted the auto bailout to unions and defended Big Bird to appeal to suburban moms. And the technological network built around this narrative served as a defensive weapon, too. Even in the wake of the campaign’s worst day, following the first debate, the Obama team was crafting issue-specific video clips and shareable content for their interest groups, counting on their vast network of digital supporters to minimize the damage.
The Obama campaign’s antagonism was specifically targeted, too, with brutal attacks on Romney’s record at Bain Capital and his remarks on the 47 percent. While the negative attacks may have appeared to be an appeal to the liberal base, Ross Douthat of the New York Times pointed out that the real aim of these ads was to depress working-class whites, particularly in Ohio. If that was the intent, it worked: The drop in the white vote, particularly in economically downtrodden Southeast Ohio, exceeded Obama’s margin of victory.
On offense or defense, the overall narrative was simple: Mitt Romney didn’t have your back, and Obama did.
Thanks in part to the candidate and in part to the campaign, Romney never developed anything approaching this straightforward narrative. The pragmatism and competence that were Romney’s best assets as a CEO seemed anachronistic in an era built on using key points of controversy to activate your base. Romney’s mild-mannered critique of Obama—that the president was a good man in over his head—required voters to trust Romney to do better, if only by default. Again and again, Romney expressed his belief in America—but he could never articulate why people should believe in him.
The irony of the 2012 contest is that a candidate who ran primarily on competence stood atop a campaign that spent so much time unwittingly undercutting him on that point. This is not a problem unique to Romney, however. Since the legendary get-out-the-vote operation of the 2004 Bush campaign that increased his vote total 22 percent from 2000, Republicans have been focused on building massive databases of people. Orca was an approach consistent with this assumption. But that model is dated. The party needs to become a party of technology and data mining, recognizing that precincts are now social, not just geographic.
The Saturday before the first debate, Rhoades stood next to Romney and predicted that the next week was going to amount to an amazing turnaround in the race. The time had come for the whites of the enemies’ eyes. As a campaign mantra, it is an odd one—an order delivered on defense, not offense, one that concedes the moment of attack to the opposition. But in retrospect, it is a fitting description for the Romney plan from the beginning: conceding the momentum, sticking to your enclosed position, and counting on making up the gap at the end with tricks like Orca.
Campaign aides I’ve talked to wonder if things would have gone differently if Romney himself had been on site at the Boston headquarters—if he would have seen the flaws in the defensive strategy, the last-minute cramming before the test, the risk of being so risk-averse. But it’s likely he would have made the mistake of trusting in a campaign that wagered it all on beautiful ads and beautiful toys. Each day for the last three weeks of the election, one aide told me, Rhoades would repeat to his staff a phrase borrowed from David Mamet’s Glengarry Glen Ross, “What do winners do? They close.”
And, the aide said, “We didn’t.”