Welcome to the Age of Consumer Politics
News: Both parties are aggressively applying the techniques of database marketing to electoral politics -- whether people like it or not.
September 9, 2004
|
|
You know the candidates. And they know a lot more about you than you might expect. This year, the Democratic and Republican parties, together with outside groups, are digging through voter rolls, census records, and consumer data for every possible scrap of information on American voters. The goal is simple: find out what people are like –- especially what they care about -- and pitch them your candidate accordingly. If it sounds like commercial marketing, that’s because, for better or worse, it is like commercial marketing.
Campaigners say so far they’ve collected information on hundreds of millions of voters –- as many as 500 data points per person, from age, race, occupation, and voting history through the car they drive and the magazines they read. The data then gets fed into a database -- the Democrats’ Datamart, or "Demzilla," reportedly has 158 million voter profiles in it; the GOP’s Voter Vault 165 million – and analyzed to produce voter profiles. Campaigns then gear their political “sell” to the profiles.
Say, for example, that education appears as a swing issue for you; in that case you might expect to have a teacher knocking on your door with a message about the negative –- or positive -– effects of No-Child-Left-Behind. Similarly, if it turns out you’re an almost certain Republican voter, Democratic canvassers won’t spend time and resources trying to win you over. Both parties have been attempting to profile voters since the early 1980s, but this is the first time they’ve done it in such a hi-tech, aggressive manner, and it’s just the beginning. "I think databases on the electorate are the key resources for political parties in the 21st century," says Michael Cornfield, research director of the Institute for Politics, Democracy, and the Internet at George Washington University. "The parties are obsessed with them because they want to reach every potential person whose going to vote for them."
The Democratic National Committee, for example, has teamed up with Emily’s List -- a political action committee dedicated to electing pro-choice Democratic women –- which has created a program for predicting how voters listed in Datamart are likely to vote. By plugging this data into their algorithm, Emily’s List campaigners say they can place voters into "lifestyle clusters." Three clusters are likely Democrats, two are hard-core Republicans, and four are considered swingable – "senior health care/social security voters," "bystanders," "young, economically pressured" and "education first."
What makes you an "education first" voter? According to Emily’s list spokeswoman Ramona Oliver, that depends on 13 variables including your age, occupation, income, number of children, gender, voting history, internet access, type of car, average number of long distance telephone calls, and type of dwelling. She kept the last three to herself –- trade secrets. If you’re an ‘Education First’ voter there’s a 68 percent chance that your top voting issue is education (hence the name), and a greater than 50 percent chance that you favor public schools over vouchers, are for gun control, abortion rights, and affirmative action. It’s also likely that you support the war in Iraq and currently prefer Bush to Kerry. After conducting more than 3,000 telephone interviews to check Emily’s List’s calculations, Oliver claims that the model’s predictions are valid. The ability to locate potential swing voters is a powerful development because, as Oliver points out, "it saves time, money and aggravation." Now says Oliver, canvassers armed with this information can "talk to people about the things they care about, rather than some random cookie-cutter pitch." For example, she explains, "I know there are about 1.7 million "education first" voters in Michigan and I can tell you what house they live in. So now, I can pull the list of those people [from Datamart] and go to their door with an education message.”
Not everyone is so sanguine about political data-mining. Some observers worry that the new techniques are invasive and unregulated, and may ultimately turn off the very voters they are supposed to energize. The parties are adopting data collection and marketing techniques most voters say they despise. "It’s traditional marketing," notes Beth Givens of the San Diego-based Privacy Rights Clearing House. "But that may not be what’s best for a healthy democracy."
And some critics suggest that the parties have gone too far already, arguing that most people are not aware of how much information the parties have collected on them, much less how it’s being used. "Overall, I don’t think most people would opt into systems like this if they were given the choice to do so," says Lauren Weinstein, co-founder of People for Internet Responsibility. It is not the information gathering per se that bothers Weinstein but rather the extent to which data is being consolidated in one central repository.
Kim Alexander of the California Voter Foundation, who has spent the last four years researching voter privacy in the digital age, has her doubts as well. "This area of public policy has been very much hidden from the public’s understanding," she says. "I know the precinct walker standing at my door knows who else lives in my household, when we were born, what party we’re affiliated with, how many elections we’ve voted in -– I know that because I’ve worked in elections and campaigns. But most voters don’t realize that the stranger standing at your door has a lot of personal information about them. When voters do discover how much data the campaigns have access to about them, there is often a sense of betrayal."
Few laws regulate the use of voter databases, and critics argue that those regulations that are in place -– mostly related to the use of state voter roles –- do not effectively control how the parties collect their information or what they may do with it. A recent Voter Privacy Task Force convened by the California Secretary of State -– the only task force of its kind –- considered the methods to detect misuse of voter data “lacking” and voiced concern that personal information contained in voter files, when aggregated with other information, could make people more vulnerable to identity theft.
"It’s really tricky [to control]," says Alexander, "because the people who are the biggest users [of these databases] and get the most benefit from the data, are the very politicians who regulate the use of the data." So far, she notes, state and local governments have not tried to "ensure that the data is not being re-used and abused as it makes its way down the campaign food-chain." Abuse could range from selling information contained in the database to a third party, to allowing it to influence federal hiring decisions – thus far there have been no documented cases of either.
Aside from privacy concerns, critics of the new techniques point to a broader risk to our democratic system. As Alexander and Keith Mills point out in their report Voter Privacy in the Digital Age, the idea of targeting political messages to specific groups of voters –and hence not to others – "exclusionary and inappropriate."
"The expectation is that the campaigns will inform the voters," says Alexander. "But as campaigns have become more sophisticated in their technology, they’ve also become better and better at targeting those who they want to reach and target out those that they do not want to reach." In other words, people who aren't voting now –- people who tend to be younger, less wealthy, less educated and more transient according to Alexander -– won’t be inspired to do so in the future, thus contributing to the decline in voter turnout.
"You’re not creating opportunities for people to engage in the debate because you are so enveloping them within a target market and target messaging," says Christopher Hunter, a former policy analyst at Netelection.org and an expert on privacy and on-line politics. In addition, he points out, targeting encourages polarization. "This is part of a larger trend of people only getting information from one side or the other. The end result is a very hyper-segmented public, where the people you’ve identified as swing Democrats are only going to get mailings from Democrats, or a particular point of view. You create little bubbles around people."
To be sure, polarization is a long-term trend in American politics. This year the parties are concentrating time and money in swing states, leaving large portions of the "non-swing" American public to learn about the issues and candidates on their own. Almost 60 percent of Americans live in areas where no presidential campaign adds were aired between March and June this year, according to a study released in July by Nielsen Monitor-Plus and the University of Wisconsin Advertising Project. Compare this to the average number of times Bush and Kerry adds ran in the 17 swing states over the same time period -- 393 and 405 per electoral vote respectively.
When asked, spokesmen from both parties were quick to emphasize their respective privacy policies and highlight the upside to targeting voters. "We have very high security measures on this [the Voter Vault]" says Yier Shi of the RNC, "It is the property of the RNC –- we’re the only ones with access to it and we intend to keep it." DNC spokesman Tony Welch echoed Shi, adding that, contrary to what critics believe, the intention of the DNC was not to cut people out of the political process, but to get them involved. "When you look at the Democratic party, one of the things you will here us say, repeatedly, is that its important for every American to get out and vote ... that’s the overall philosophy of the party, and this [Datamart] will allow us to touch far more people than we would otherwise."
At least, that’s what both parties are gambling on. "There’s a lot of implicit belief that there are connections between these types of things," says Weinstein, "People assume that the impacts are there, but if you look at databases the error rates tends to be really high." Even Oliver of Emily’s List had to agree, "We’re still seeing if this is working. I mean, we won’t know until this game is over in November whether all of this is truly effective."
It’s unlikely that campaigners from either party are going to spend much time mulling over the "long-term effects of data mining on a healthy democracy" –- after all, what are academics for? What matters is getting their man or woman into office. And if that means knowing every last thing about you, that’s what they’ll do. It may feel invasive, it may tempt some not to register, but on the other hand, it might just inspire one more person to vote. And, as Michael Cornfield points out, "That’s a good thing. At this point, voter turnout is so low that I would be satisfied with any effort that results in a bump up."
Kari Lundgren is an editorial fellow at Mother Jones.
