The Victory Lab (36 page)

“Jeez, Karl, I don’t know how many of those are out there,” Shaw told him.

“Well,” replied Rove, “let’s expand the definition to being nonhostile.”

Shaw ended up identifying eight academics who met the standard, including Gimpel, and Rove called them his “Team B.” (The political professionals were Team A.) When the Academic Advisory Council gathered in Washington in late 2001 to preview the next year’s midterm elections, Gimpel arrived with a list of industrial midwestern counties where he said Bush had suffered in 2000 because younger Republicans had turned out at dismal rates. A renewed sense of civic interest after September 11 could offer a window to register and mobilize young conservatives, Gimpel suggested, a project best accomplished by mining his county data to identify
precincts with the best combination of young populations and Republican performance. It was in many ways a classic Gimpel subject: his specialty was political geography, with a focus on the spatial dimension of campaigns. His work with the RNC gave Gimpel access to the party’s national voter database, making him one of the rare political scientists comfortable with the same tools that campaign operatives used.

None of the four academics who had signed on with Perry expected that they would do much but repeat many of the straightforward voter contact experiments that Gerber and Green had run elsewhere, only this time within a big partisan candidate campaign. Running another test of phones and mail didn’t excite them much intellectually but it could help remove the cloud of skepticism that had shadowed Gerber and Green since their first experiment: a persistent critique that lessons learned from a nonpartisan New Haven civics project had little to do with the scrum of real politics. “In some instances, they made conclusions that phones weren’t effective based on one call, but they didn’t test multiple calls,” says Jeff Butzke, Perry’s phone vendor. “I’d agree that one call probably wasn’t that effective. More calls, especially on election day, are more important.”

Carney’s doubts about campaign spending extended far beyond phones and mail, however. In 2002, Perry’s opponent, a South Texas businessman named Tony Sanchez, spent $76 million, most of it his own money, including what Perry’s staff estimated was $40 million in ads attacking their candidate. (Perry had moved into the governor’s mansion in January 2001, upon Bush’s election as president.) Perry had spent $28 million and still beat Sanchez by almost twenty points. That disconnect between dollars and votes, along with his natural cynicism about the industry that thrived on it all, drew Carney uncomfortably toward the philosophical. “Dave could sit there and get reflective sometimes,” says Luis Saenz, who was political director of the 2002 campaign. “ ‘Does any of this fucking matter?’ ”

That became, to the political scientists’ surprise, their research agenda for the 2006 election cycle. What had long been off-limits to experimenters as “real politics” was now on the table: Carney was ready to
test anything the academics could figure out how to randomize, from lawn signs to television ads. He liked Gimpel’s idea of geocoding fund-raising invitations to learn the optimal radius for drawing people to events. In addition, there were a few things Perry’s advisers were itching to learn. Their candidate was pushing for a more manageable travel schedule in 2006 than he’d had in his previous statewide races; Perry told Carney he wanted to sleep in his own bed at night. Could the professors come up with a system for maximizing the effectiveness of Perry’s personal visits in a way that would allow him to avoid the long itineraries that kept him away from Austin for days at a time?

Perry’s advisers placed only one major restriction on Gerber, Green, Gimpel, and Shaw. They couldn’t publicly discuss their work with the campaign until it was over. The outside consultants and vendors feared press coverage that could embarrass them, but Carney wasn’t much concerned with their feelings. He didn’t want to see leaks that could reveal the campaign’s strategic decisions or unmask differences within the leadership. Other advisers worried that the experiments could make it look as though they were trivializing the work in front of them. “I don’t think we would want any of this out there—that we were just messing around with professors as though we didn’t care about the race,” says Saenz, who became campaign manager. “The reason we were doing this is because we were dead serious about the race.” In exchange for signing to a nondisclosure agreement, the professors were granted what mattered most in a line of work where “publish or perish” is a vital maxim. After November 2006, they would be free to use any findings from the campaign’s work in a journal article or book.

The academics happily agreed to these terms, amazed by their good fortune. They made plans to convene in Chicago to plot specific experiments they wanted to present to the Perry campaign. The four academics sat on the sidelines of the Midwest Political Science Association annual conference, arguing how to exploit an opportunity that had never before been extended to a political scientist. For years, mass media, the dominant form of political communication but the one whose effects were hardest to
isolate, had loomed as the holy grail of campaign experiments, and now they were being invited to treat a television advertising budget as their laboratory.

IN THE SPRING OF 2005
, Carney stood in an empty Mexican restaurant called the Oasis on Austin’s Lake Travis, in front of what amounted to Perry’s extended political family, and prepared to introduce a quartet of strangers. A crowd of about thirty people were seated, auditorium-style—a group that included the core campaign staff for 2006; a retinue of outside consultants and vendors; Perry and his wife, Anita; and their kitchen cabinet, many of them former aides.

Those who worked for Perry considered him “a dream client,” as Baselice put it: eager to make fund-raising phone calls, willing to defer to his advisers’ guidance and stick to a script when put in front of him. He also showed his personal gratitude to those who served him. After 2002, to congratulate them on his victory, Perry invited members of his core team (including Carney and his wife) to join him climbing Guadalupe Peak, the highest point in Texas. A little over a year later, he rounded up many of the same advisers (including Carney again) for a retreat in the Bahamas, an outing that was nominally devoted to school finance policy but was more hands-on in its treatment of scuba-related issues. (
The trip triggered an ethics investigation when it was revealed that the expenses had been paid by political supporters. Perry was eventually cleared of wrongdoing.) “The joke is we’re like the mafia—you only think you get out,” says Deirdre Delisi, who managed Perry’s 2002 campaign before becoming his gubernatorial chief of staff. “Perry inspires a lot of loyalty in terms of people who have worked with him and stayed whether they’re officially in it or not.”

The campaign was preparing for what it expected to be a difficult primary against Carole Keeton Strayhorn, the state comptroller and former Austin mayor. The rivalry was already intense between Perry’s world and Strayhorn’s, and early in the day Carney had set the tone by claiming,
“We’re going to rip her leg off and beat her over the head with it!” The statement jarred Bill Noble, who had expected to be attending more of a planning session than a pep rally, one where there would be an open exchange of ideas about strategy but little need to fire up Perry’s loyalists. Then he saw how Carney’s bluster was received by four strangers in their midst. “I just looked over at them, and there was this look of shock,” says Noble. “It turns out it was for effect.”

Carney appeared intent on unsettling everyone at the Oasis restaurant that day: the group of academics he introduced as “our four eggheads” and the political hands they were encountering for the first time. The daylong retreats were a ritual for Perry campaigns—and many in attendance were veterans of past retreats Carney had organized before the 1998 and 2002 races—but this was the first time he had prepared a syllabus. Carney had ordered fifty-five copies of
Get Out the Vote!
and everyone who was to attend the retreat received one with orders to read it.

Perry would likely begin campaigning in the fall and launch his ad campaign in January, setting up before the March 2006 primary a two-month sprint during which the eggheads could run their experiments. There would then be eight months for them to process the results and analyze the findings, leaving Carney enough time to apply their conclusions to Perry’s general-election tactics. “When you’re spending twenty-five million dollars on an election and you can save two percent, that’s a lot of money. You’ll have more money to spend on something that works,” says Carney, who thought he could boost Perry’s summertime fund-raising power by showing off his rigorous experimental regime to “assure donors that we’re using their money as best as possible—spend it different, spend less of it.”

Don Green began presenting the research that he had done with Alan Gerber over the years, rigorously itemizing all the things that campaigns did that he believed he had proven to be a waste of money. Green did this while facing a room filled with people who had gotten rich off these practices and had been looking ahead to the next Perry campaign as another big payday. Carney likened Green’s talk to “going into the Catholic
church telling everyone that Mary wasn’t a virgin, and Jesus really wasn’t her son.” Carney delighted in the face-off he had manufactured, the awkward pitting of academics against professionals—with millions of dollars, control of the country’s second-largest state, and claims of intellectual supremacy all at stake. “Carney is a very confrontational person. That’s how he drags out the best product in people,” says Deirdre Delisi. “What he’s trying to do is force creative thinking.”

There was little evidence of creative thinking at the Oasis that day. In fact, it became apparent pretty quickly to Carney that few of the attendees had read the book closely. Those who might have didn’t look any more persuaded by the experimental methods or conclusions Gerber and Green had reached. The consultants replied as Carney expected they would—in what he called “total denial,” boasting of pieces of mail or phone campaigns that had proven decisive in past elections. One vendor, making the case for robocalls, recalled a legendary voter registration program that featured a recording of a little girl’s voice reminding people to sign up.

“Great,” Carney said, wrapping up the presentation. “One of you is right. Either the eggheads are right or you’re right. We’re going to prove it out, and plan our campaign and allow these guys to develop experiments for everything we do.”

Whenever the eggheads walked into Perry’s Brazos Street headquarters, Gimpel thought, they were seen as the “internal affairs bureau of the police department,” a watchdog monitoring the campaign’s outside vendors. “They don’t want people looking over their shoulders and checking up on them,” he says. When the eggheads were out of earshot, the vendors mused ironically about the prospect that the Yalies were lefty moles. “Let’s hope to God there’s not an e-mail going out to the DNC,” they would joke.

More seriously, randomizing tactics as significant as media buys and candidate travel potentially set Carney up for charges of self-sabotage. Even though the academics had planned their media study in such a way as to make sure it didn’t dangerously undermine Perry’s prospects, the whole point of randomization was for the campaign to do something that
it might not have done for strategic reasons. Nonetheless, research objectives were likely to introduce new inefficiencies into the management of the campaign. What if, as part of a candidate travel experiment, the randomizer assigned Perry to do back-to-back visits the same afternoon in El Paso, in the far west corner of the state, and Shreveport, Louisiana, whose media market straddled Texas’s eastern border? “I just wanted to make sure it didn’t get in the way of the campaign,” says Ted Delisi, who was the campaign’s mail vendor (and married to Perry’s chief of staff). “Sometimes you’ve got to get in the car and stomp your foot on the gas and not wonder how the engine works.”