Friday, May 30, 2008

DNC blunts GOP Micro Targeting Lead

By: David Paul Kuhn, Politico
http://www.politico.com/news/stories/0508/10573.html
May 24, 2008 07:10 AM EST

After years of struggling to catch up to the Republican Party’s sophisticated micro-targeting efforts, the Democratic National Committee appears to have come close to parity.

The DNC has now reorganized its data banks into one centralized file that goes a long way toward neutralizing the GOP’s advantage in drilling down and identifying crucial constituencies of voters.

In the past two presidential cycles, the Republican national voter file allowed the party to more efficiently locate, communicate with and galvanize voters. Democrats, by comparison, relied on a disjointed compilation of national and state party data files that varied widely in quality. To boot, said one DNC analyst, many of their files would vanish after each election year.

For Democrats, the shift to one “solid voter file” is “transformative,” said Ben Self, the DNC’s director of technology.

“Whether it is micro-targeting, regular targeting, neighbor-to-neighbor knocking on doors, or volunteers making calls,” Self said, “all these vital campaign activities are built on a national voter file and were not available in 2000.”

Or, in 2004, when Republicans took full advantage of their advances in technology. In Florida alone, as President Bush sought reelection, consumer data enabled the GOP to locate regular churchgoers who were not Republican, as well as identify a group of Hispanic mothers particularly supportive of the No Child Left Behind law.

The DNC and Republican National Committee now each boast one vast data coffer that merges traditional voter statistics on gender, geography, or party identification with consumer and census data.

Hal Malchow, one of the pioneers of micro-targeting, emphasizes five key characteristics in which consumer data points are most helpful: deciphering ethnicity and race, church attendance, marital status and gender, geography, and gun ownership. Democratic strategists have also found that determining the size of a household and whether someone rents his home also prove to be significant indicators of partisan leanings.

Each party compiles hundreds of these data points on one voter, mines the data for patterns and labels a voter on a 1 to 100 scale to measure likely support. Perhaps more important, the DNC cross-references phone numbers and addresses for each individual.

Mastering the art of micro-targeting is increasingly important to the parties as the more traditional forms of communication become less and less effective in campaigns.

Political advertising, for example, is becoming less salient because digital video recorders permit voters to skip commercials. The Yankee Group, a technology research and consulting firm, estimates that between 20 percent and 27 percent of American homes have a DVR.

According to the most recent report by the Pew Internet and American Life Project, 15 percent of Americans get most of their campaign news online.

And cell phones have made conventional phone banks less efficient, since it is more difficult to locate a cell number and since users commonly screen calls.

Recently, the DNC began a “neighbor to neighbor” program that “geo-codes” volunteers — a process that assigns geographic identifiers to them — to increase the likelihood that neighbors will canvass neighbors.

In short, a volunteer signs up. The 25 nearest neighbors who pique the DNC’s interest are then mapped out for the volunteer. The DNC also offers a script to use during canvassing as volunteers go door to door, asking their neighbors the degree of their Democratic support or their support for John McCain. The volunteers asks about their neighbors' top issue interests. The aim is to return and later target each person with a specific script based on their previously identified concerns.

Volunteers are ranked locally for their effectiveness and rewarded with invitations to intraparty conference calls or meetings. They are also encouraged to forward invitations by e-mail to friends or family, mimicking the viral success of social networking websites.

The program, which debuted in Kansas in late April, was expanded to Virginia. The DNC plans to gradually roll out the program nationally by mid-summer.

Some Republicans question whether the DNC could catch up in the span of just one presidential cycle.

“Speaking frankly, knowing how difficult a project this is, how many mistakes we’ve made over the years, there is actually no way they will be able to have caught up in such a short period of time,” said a top RNC analyst who spoke on the condition of anonymity.

The party continues to build the DNC’s voter file with some assistance from the Obama and Clinton campaigns, which have been offloading data to the DNC file.

That, in itself, is an accomplishment for a party that brought its voter file in-house for the first time in the 2006 midterm elections. That year, Democrats conducted a pilot program using the data in six states, including Montana, where Jon Tester unseated Republican Sen. Conrad Burns.

The data set allowed the party to narrowly predict possible Tester supporters. Democrats found that smaller households and unmarried voters were “predictive” of Tester support as were those with advanced degrees. Voters from more rural areas or with higher travel times were more likely to back Burns.

Burns voters were also more likely to participate in outdoor activities, gardening, investing, to be religious contributors and to have an interest in golf. Tester voters were more interested in health products and more likely to have a retail credit card.

The Montana pilot project expanded the party’s likely voter universe by some 15,000 residents. In an internal poll following the race, the DNC found that 85 percent of their top quartile of modeled Tester supporters voted for him.

“A lot of the consumer data helps at the margins,” said Keith Goodman, the director of special projects in the DNC’s political department. And, as Goodman notes, many elections are decided in the margins .

No comments: