By Michael D. Shear
Washington Post Staff Writer
Sunday, August 28, 2005
RICHMOND -- The Democratic and Republican candidates for Virginia governor are using vast databases of demographic information and the consumer habits of state voters to create precision advertising that rivals the best of Madison Avenue.
Armed with sophisticated new polling techniques refined by both national parties in the 2004 presidential election, Republican Jerry W. Kilgore and Democrat Timothy M. Kaine have compiled detailed profiles based on how old voters are, what houses they live in, what newspapers they read, what restaurants they eat at and how much money they make.
Do Kilgore's voters read biographies, watch foreign flicks and shop at Nordstrom? Will two-parent households in Chantilly making $50,000 a year vote for Kaine? Which voters can be lured to the polls by messages about abortion, immigration or taxes?
With the click of a computer mouse, both campaigns will soon know the answer to those kinds of questions. The goal: direct mail letters and voter drives designed not for the masses, but instead for tiny, well-defined slices of the state's voting population.
It's called microtargeting, and in a close election like the one shaping up in Virginia, political experts say it could be decisive.
"Our message stays the same no matter who we talk to. But different issues, or different pieces, are going to be more important to some people," said Mo Elliethee, Kaine's communications director. "Microtargeting really gives campaigns the ability to speak directly to voters in ways they haven't been able to before."
Marketers, law enforcement agencies, security officials, background screeners and now politicians are tapping into the same commercially available databases with the goal of rating people with laserlike accuracy.
"Private industry spends billions of dollars on this targeting. The companies think it works. Why wouldn't it work for politicians?" said Peter Swire, a law professor at Ohio State University and an expert on privacy and technology.
But the use of such precise research in political campaigns raises concerns about the demise of a shared public discourse on the important issues of the day.
"One worry is that candidates will talk out of a thousand sides of their mouth," Swire said. "What if they have a thousand messages that are different for every tiny part of the population? It's much harder to get accountability for what candidates say."
Operatives in both campaigns are reluctant to provide details of their microtargeting efforts. Both sides refuse to say much about their methods, fearing that political countermeasures could be effectively deployed.
Kilgore campaign aides, for example, have divided Virginia voters into 34 categories, but they won't divulge what they are. Democrats declined to reveal the "scores" they give voters to determine how likely they are to be supporters.
Campaign strategists in both parties acknowledged that they are making Virginia history by marrying computers, polling and information culled from census data or consumer lists publicly available for purchase.
"It's a great resource for us in a state like Virginia because there's no party registration," said a GOP strategist involved in the microtargeting effort for Kilgore. "We don't know, necessarily, who our Republicans are."
The new techniques are rewriting the campaign playbook for both state parties.
Previous campaigns used data about precincts to target their money and time. Republicans, for example, might have ignored a largely African American precinct where Democratic candidates historically win with 80 percent or more of the vote. With microtargeting, however, Kilgore's campaign can target the precinct's smaller number of conservative African Americans.
For Kaine, the new technique is being used in part to pry Democratic voters out of the largely Republican exurbs in Northern Virginia, Richmond and Hampton Roads.
Kilgore's effort is run by TargetPoint consulting, the same firm that pioneered the technique for the Bush campaigns in 2000 and 2004. This summer, the firm conducted its microtargeting poll for Kilgore, paid for by the Republican National Committee.
The RNC would not provide a copy of the poll questions or the results, but a Virginia voter who was surveyed took detailed notes about the questions. Kilgore campaign aides said the notes were largely accurate.
Respondents were asked to express their opinions about programs Kilgore has been advocating and about attacks he has been making against Kaine. Strategists said the questions allow the campaign to find "anger points" that are important to small segments of the voting population.
One question asked about "anti-death penalty advocates comparing the death penalty to concentration camps." Kilgore has criticized Kaine for a 1987 comment in which he mentioned Virginia's death penalty in the same sentence with the Soviet gulag.
Another poll question asked about "politicians who increase taxes to cover what they claim is a huge budget deficit only to announce a few weeks later that it is really a surplus." That is similar to an attack on Kaine's support for tax increases during the 2004 budget fight.
Republican strategists said the poll results are being merged with consumer data to give Kilgore clues about how to woo certain kinds of voters. The 34 groups have their own consumer and demographic profiles. Those in Group 1 are strong Kilgore supporters. Those in Group 34 are strong Kaine supporters. Those in Groups 15, 16, and 17 are considered swing voters.
The data could allow Kilgore to communicate directly with rural Democrats turned off by Kaine's position on abortion. Or they could send mailings directly to "casual Republicans," who vote only in presidential elections.
"The RNC is committed to ensuring Virginia's next governor is one that represents the state's values and will deploy the latest campaign technologies, which will help identify and turn out voters of all persuasions on behalf of Jerry Kilgore," said Danny Diaz, a spokesman for the RNC.
Kaine's program is slightly different, according to strategists familiar with it.
Instead of relying on consumer data, Kaine's pollsters have used the results from a 10,000-person poll, which is designed to elicit beliefs and preferences in small groupings of voters, together with detailed census data to generate a score for every voter.
The poll was conducted by Kaine's longtime pollster, Peter Brodnitz, and statistical models were designed by the firm Copernicus Analytics.
A Democratic strategist involved in Kaine's microtargeting said voters' scores will determine how much campaign mail they receive -- if any -- and what kind of pro-Kaine political messages the mailings contain.
"We use the score to decide whether that person is worth the effort," he said, speaking on the condition of anonymity because of the campaign's desire to keep the strategy secret. "The index tells us who we should be targeting."