Getting to know you: Perry team targets voters
Professors chronicle effort, try to determine how to best drive turnout
Tuesday, July 25, 2006
By CHRISTY HOPPE / The Dallas Morning News

AUSTIN – The car you drive, the magazines you read, the catalogs you buy from, the house and
neighborhood where you live and your voting pattern have been gathered by Gov. Rick Perry's
campaign – and others. It helps them know which button to push to win your vote.

Such consumer-based, micro-targeting helped President Bush's campaign identify and attract 10 million
new voters in 2004. Mr. Perry's team is adapting the Bush model to the state level, and their work could
show up at a mailbox, telephone, doorbell or even iPod near you.

The campaign is looking especially for marginal Republicans and independents who can be persuaded to
go to the polls. The desire to boost voter turnout prompted Mr. Perry 18 months ago to open his
campaign to four political scientists – including two Yale University professors.

"We fondly refer to them as our eggheads," political director Dave Carney said.

The professors' work, which will evolve into a paper or a book after the election, delves into what kind
of contact coupled with which message will motivate which supporters to the polls.

So how much does the campaign know about individual voters?

"We know a lot," Mr. Carney said. "We have a lot of data. We know how to segment the votes. We have
a good handle on who's going to vote. And we know really what the results are going to be if we can
execute our campaign plan."

Multiple data sources

Some of the data, such as whether one has voted, is available in public records. Other information comes
from vast commercial databases that track consumer preferences and habits.

The professors and their studies are not involved with the campaign's message. "But they're coming up
with ideas to see if they can impact it and make it more positive," Mr. Carney said.

In a five-way campaign with a lot of variables that could tip the outcome, some hard science could help
the Perry camp maximize voter turnout. And the professors' chronicle of the effort could help raise the
governor's profile among political experts just as another national campaign is getting under way.

What's happening in Texas is radiating through about 16 other states, said Matthew Dowd, chief
campaign strategist for the 2004 Bush campaign.

Mr. Dowd was one of the pioneers who meshed consumer habits, public data, and information on
lifestyles and applied it to a different kind of consumer – the American voter. The technology has
allowed campaigns to determine whether a person is likely to tilt Democrat or Republican and – coupled
with polling on issues – to know what topic is likely to propel them into a voting booth.

All of it still requires the right candidate and the right message, he said. But then, it's about persuasion
and motivation.

"This is just using technology to be able to do that more precisely," Mr. Dowd said. "So if you wanted to
send someone to a neighborhood, as opposed to having to walk down the street and hit 35 doors, you
could say, 'Go to these three doors.' "

The formula is complex. Mr. Dowd, who is not involved with the Perry campaign, said that in 2004, the
Bush campaign had amassed about 182 pieces of information on any given voter.

"Whether they voted, their age, sex, what kind of neighborhood, what kind of car they drive ... whether
[they buy] fishing gear or [shop from] Victoria Secret catalogs," Mr. Dowd said.

Likely supporters were then identified, based on information gathered in the previous election, and the
campaign targeted those that, without the right push, might not have voted.

It's still high science, but it's now more widely available.

"The technology and efficiency of doing it finally caught up so that campaigns can afford it," Mr. Dowd
said.

Democrat Chris Bell's campaign, which has less than one-tenth the money that Mr. Perry has raised, is
using some of the technology.

"We're definitely looking at the data," spokesman Jason Stanford said. "I could shove a Bible out the
window, but it's not going to convert any people. You have to find a way to deliver your message."

Mr. Carney always has been intrigued by the science of politics. On a plane trip to Austin two years ago,
he read a book by Yale professors Donald Green and Alan Gerber, titled Get Out The Vote: How to
Increase Voter Turnout. By the time he landed in Texas, he had e-mailed Dr. Green asking if he wanted
to work with a real campaign "and actually help us test our methods in real time."

Dr. Green, who had worked mostly with nonpartisan campaigns and voter initiatives, immediately took
up the challenge. "It's great that a campaign of this magnitude has been willing to let us do these
studies," he said, likening the work to a "fifth-grade science experiment."

Primary tests

They used the March primaries as a laboratory. Along with University of Maryland professor James
Gimpel and University of Texas professor Daron Shaw, the four began testing the campaign's methods
of reaching voters.

For instance, if the campaign identified 1,000 homes to mail flyers with a certain message, Dr. Green's
group randomly eliminated 500 homes from the list. And then they watched how both groups voted in
the primary, trying to quantify what effect the message and delivery method had.

They did the same for personal appearances, TV, radio ads and automated, pre-recorded phone calls.
Mr. Carney was hesitant to offer many details because the information gleaned will be used in the
campaign, but he pointed to a major question he wanted answered: whether a candidate must personally
visit towns to win attention and votes, or whether a TV interview by satellite might suffice.

The campaign's goal is to get a few minutes on local TV news to address community issues. But it is less
expensive, more time-efficient and easier to give interviews by satellite to local stations, he said.

"So we did a number of tests along that line," Mr. Carney said. "We determined there was a tremendous
correlation between Perry in a city versus him being on the news."

The ripple effect of Mr. Perry shaking hands, meeting people, saying hello to volunteers and appearing
in a local setting had a huge effect, settling the question once and for all, Mr. Carney said.

Many long-held assumptions about how to reach voters have been debunked, he said. And that hasn't
been easy to take for media experts and direct-mail specialists who believe in their work.

That first meeting with the professors and the campaign staff was a little uneasy, Mr. Carney recalled.
"It would be like Dan Brown coming in with The Da Vinci Code before the College of Cardinals," he
said.

But Mr. Carney said he believes their work will help refine politics and voter turnout. And, he said, by
helping to identify how the campaign can use its money to have the biggest effect, "we've saved untold
hundreds of thousands already."

He said the danger to Mr. Perry is that any misstep will be documented in academic libraries nationwide.
But, he acknowledged, a successful how-to book appearing in a year – when Mr. Perry's name could be
bandied about as a potential vice presidential candidate – might not be bad.

Dr. Green said working with a large campaign in a big state is generating valuable information that
could apply to selling a candidate or a Big Mac.

"We're very interested in how many dollars it takes to produce a sale," he said – what works and what
doesn't.

Of course, Dr. Green said, he recognizes the healthy skepticism between the art and the science of
politics – from the cold numbers to the passionate believers.

"They look at us with a certain amount of amusement," Dr. Green said. "And we're an order of
magnitude nerdier than the other people in the room."