data mining

by Chris Evans, UMN Law Student, MJLST Managing Editor

In “It’s the Autonomy, Stupid: Political Data-Mining and Voter Privacy in the Information Age,” I wrote about the compilation and aggregation of voter data by political campaigns and how data-mining can upset the balance of power between voters and politicians. The Democratic and Republican data operations have evolved rapidly and quietly since my Note went to press, so I’d like to point out a couple of recent articles on data-mining in the 2012 campaign.

In August, the AP ran this exclusive: “Romney uses secretive data-mining.” Romney has hired an analytics firm, Buxton Co., to help his fundraising by identifying untapped wealthy donors. The AP reports:

“The effort by Romney appears to be the first example of a political campaign using such extensive data analysis. President Barack Obama’s re-election campaign has long been known as data-savvy, but Romney’s project appears to take a page from the Fortune 500 business world and dig deeper into available consumer data.”

I’m not sure it’s true Buxton is digging any deeper than the Democrats’ Catalist or Obama’s fundraising operation. Campaigns from both parties have been scouring consumer data for years. As for labeling Romney’s operation “secretive,” the Obama campaign wouldn’t even comment on its fundraising practices for the article, which strikes me as equally if not more secretive. Political data-mining has always been nonpartisanly covert; that’s part of the problem. When voters don’t know they’re being monitored by campaigns, they are at a disadvantage to candidates. (And when they do know they’re being monitored, they may alter their behavior.) This is why I argued in my Note for greater transparency of data-mining practices by candidates.

A more positive spin on political data-mining appeared last week, also by way of the AP: “Voter registration drives using data mining to target their efforts, avoid restrictive laws.” Better, cheaper technology and Republican efforts to restrict voting around the country are inducing interest groups to change how they register voters, swapping their clipboards for motherboards. This is the bright side of political data-mining: being able to identify non-voters, speak to them on the issues they care about, and bring them into the political process.

The amount of personal voter data available to campaigns this fall is remarkable, and the ways data-miners aggregate and sort that data is fascinating. Individuals ought to be let in on the process, though, so they know what candidates and groups are collecting what type of personal information, and so they can opt out of the data-mining.

by Bryan Dooley, UMN Law Student, MJLST Staff

Most voters who use the internet frequently are probably aware of “tracking cookies,” used to monitor online activity and target ads and other materials specifically to individual users. Many may not be aware, however, of the increasing sophistication of such measures and the increasing extent of their use, in combination with other “data-mining” techniques, in the political arena. In “It’s the Autonomy, Stupid: Political Data-Mining and Voter Privacy in the Information Age,” published in the Spring 2012 volume of the Minnesota Journal of Law, Science, & Technology, Chris Evans discusses the practice and its implications for personal privacy and voter autonomy.

Both parties rely extensively on data-mining to identify potentially sympathetic voters and target them, often with messages tailored carefully to the political leanings suggested by detailed individual profiles. Technological developments and the widespread commercial collection of consumer data, of which politicians readily avail themselves, allow political operatives to develop (and retain for future campaigns, and share) personal voter profiles with a broad swath of information about online and market activity.

As Evans discusses, this allows campaigns to allocate their resources more efficiently, and likely increases voter turnout by actively engaging those receptive to a certain message. It also has the potential to chill online discourse and violate the anonymity of the voting booth, a central underpinning of modern American democracy. Evans ultimately argues that existing law fails to adequately address the privacy issues stemming from political data-mining. He suggests additional protections are necessary: First, campaigns should be required to disclose information contained in voter profiles upon request. Second, voters should be given an option to be excluded from such profiling altogether.