Last week, we posted a list of open questions about how voters’ personal information was used in the 2012 presidential election (and how it’ll be used now that the election’s over). We posted the questions on our blog and on the social news site Reddit, and lo and behold, someone from the Obama campaign’s data mining team answered!
Although he wasn’t in charge of the data mining operations, his answers still give some insight into how personal information drove the campaign. We found it especially unsettling to see how much detailed Facebook data is up for sale for marketing, regardless of whether you’ve set your Facebook account to be more private (see his answer to question 4).
We’re posting our questions and his answers below (note that we haven’t edited his answers at all, except to format their numbering):
1. Abine: How detailed did the profiles that the campaign built up about people get? Like, was it “category of single guys in OH who like microbrews,” or “John Smith in OH who hasn’t had a girlfriend in 3 years and always drinks Dogfishhead?” Bonus: can we see a (redacted) real profile?
Campaign Employee: “I wasn’t high enough to get end results, but I saw some of the print outs upstairs… They know everything Google and FB know about you, pretty much. They know what music you like, which Harry Potter book is your favorite, your voting habits, etc. It’s all in databases, you’re just a number in a DB with a name attached.”
2. Abine: It sounds like databases have become a valuable asset in a growing number of campaigns. Is there anything stopping, say, the losing party who’s in campaign debt and has to answer to creditors from selling the data to marketers and other third parties to help fulfill that debt?
Campaign Employee: “Again, I’m not the person to talk to about that, didn’t even think to ask that. I AM in advertising with a focus on DB marketing though and it depends on the conditions on which they got your data and what they agreed to in the process.”
3. Abine: Let’s talk about online tracking. Advertisers always argue that the data they collect doesn’t pose any privacy problems because it’s aggregated and “de-personalized,” while researchers say there’s no such thing as truly anonymous tracking. Given your experience, who’s closer to the truth?
Campaign Employee: “I’ve worked on advertising campaigns before. I can tell you that as long as the databases aren’t leaked, you’re fairly safe. They will have your name, CCN, SSN etc. but most of that is only used to match personally identifiable information to you (matching your Amazon spending habits to your FB “likes” for example). In other words, the latter, but it’s happened since the first targeted print ads. Magazines and newspapers have collected similar data without you knowing for decades.”
Campaign Employee: “List brokers are huge in politics AND regular marketing. Here’s an example from a very common list provider: http://lists.nextmark.com/market?page=order/online/datacard&id=351922 You’d use the universal rate/M(thousand) then add cost per thousand for all stratification and personalization options. This is for one time use data. List brokers do “seed” fake addresses that direct to the seller in order to prevent people from abusing and reusing that data.”
5. Abine: Be honest: do we have anything to worry about? Is this data collection truly for the limited purpose of the campaign, or is it going elsewhere? Has it been/could it be misused to harm people?
Campaign Employee: “Nah, the extent of the campaign within Hamilton County only used certain information to match voting habits, gender, address, phone number, name, registration and favored candidates for selection. They know more but they don’t use it. In the case that the New York Times endorsed Romney, though, they could, theoretically, target NYT subscribers in their DB (which they know) with Obama ads in order to counteract that. It’s a fine and strategic science. I don’t know what the other side does with their data, but neither campaign has money problems so there’s no incentive to sell the data, if that’s even allowed, depending on the user. Our only goal was to get people to the polls to vote for our candidates. If you want to know more, let me know! I’m not exactly a data miner (more of a PR/marketing guy, by training), but I know some and I know how data is used, so I hope this has been sufficient.”
We had a few follow-up questions about how the campaigns get your Social Security Number and whether your number in the database with a name attached is personally identifiable, and he had a joint response:
Campaign Employee: “It depends on what you give them, the venue and agreements you go through to give it to them. It gets around. If you’re scared of anyone, Facebook and Gmail go through your individual e-mails and chats.”
When we said that we were “shocked by not only the sheer amount and detail of information you can buy, but the fact that people who have set their Facebook pages to private still aren’t safe,” he concluded that “Yeah, it’s nutty, but that’s the nature of the beast!”
What do you think? Are you surprised by any of the ways in which your personal data fuels political campaigns?