Micro-targeting voters may be key to election

Change in strategy for Democrats
But this year, Podhorzer said, Democrats will be looking at the entire voter file of people who voted in 2004, but not in 2002, and then try to figure out, regardless of whether those voters live in heavily Democratic precincts, which of them were Kerry voters and how to get them to the polls.

Using the list of self-identified Democratic voters or registered Democrats who have voted in recent elections, the party or an outside group will conduct a large survey of perhaps 3,000 voters, asking them a series of questions, either on issues, such as gasoline prices, or on values, such as privacy or the importance of religious faith.

Values-based micro-targeting, as the GOP did in 2004, would present voters with agree-or-disagree propositions such as, “I feel more comfortable with political leaders who have a strong faith in God.”

The survey might also probe for “anger points” — what drives voters crazy: “How angry are you that Democrats are trying to repeal tax cuts?”

Using consumer data and Census data on income, education levels and other factors, the party will then group the poll respondents into “clusters.”

Seeking 'traditional marriage Democrats'
A cluster called “traditional marriage Democrats,” for instance, would include Democratic voters who could be swayed with traditional marriage messages.

“That’s exactly what happened in West Virginia,” said Cabrera, where Republican micro-targeting on the marriage issue helped President Bush defeat Kerry in 2004.

Driving through southern Minnesota a few weeks before Election Day 2004, right outside the town of St. Peter, Minn., in bellwether Nicollet County, a huge billboard read: “Want Gay Marriage? Vote Democrat this November.”

“They knew exactly where to put that billboard based on micro-targeting,” surmised Cabrera.

Once the cluster of voters is grouped together by attitudes, then researchers can find out what else such people have in common and can design a targeted message to reach them and people akin to them in lifestyle and ideology.

So in theory, at least, Democratic Senate candidate Bob Casey’s campaign in Pennsylvania can target a libertarian, Wal Mart-shopping young mother with two kids ages 8 and 10 in public school, with a custom-tailored message on public education, rather than wasting its time trying to appeal to her on the issue of job losses due to Chinese imports.

“There are very few campaigns that actually realize that level of advanced sophistication,” said Podhorzer.

“We’ve moved from precinct-level data, which will show you a precinct and it shows up as either ‘red’ (Republican) or ‘blue’ (Democratic), to Census-level data, where you can actually see veins of ‘red’ or ‘blue’ running through these precincts,” Cabrera said.

“You know exactly which neighborhoods to skip, which doors you really have to knock on; we can go in the Republican neighborhoods now — places we'd have skipped before because they were predominantly ‘red’ — and just turn out a section of that precinct because we know that’s where the Democrats are,” he said.

The final blitzkrieg
By October, Podhorzer said, the campaign should be trying to communicate with targeted voters through television, mail, robo-calls (frequently with recorded voices of famous people such as Bill Clinton) — “everything they can to break through the din.”

Politics   Click here to share your thoughts on political matters, including the war in Iraq, the mortgage crisis, campaign fundraising, immigration, the economy, NAFTA, and the presidential campaign.

The GOTV blitzkrieg intensifies the Saturday before the election, with volunteers walking through neighborhoods with lists of addresses of targeted supporters, dropping literature and reminding people where their polling place is.

“We literally try to come up with a likely propensity for someone to vote Republican,” said Anuzis. “You personally might have a 62 percent propensity to vote Republican, while your next-door neighbor may be 72 percent, and the guy living across the street might have a 42 percent propensity. And yet you all live on the same block. So now we have a targeted list and we would more likely go to the 72 percent first, the 62 percent person second, and the 42 percent person third.”

And the side that plays this GOTV game more effectively will likely hold the majority in the new Congress.