Data in politics is nothing new. Most politicians who have been active for longer than the past decade almost always have a book, or a notepad, or a stack of loose leaf paper with important information on lawn sign locations, volunteer contact information, and addresses of past supporters. That data would give them a real (though not insurmountable) edge against a new opponent who might not have been around long enough to gather it themselves.

But now, with unprecedented amounts of data on voters, the gap in electoral success between those who can access data and those who can’t is growing larger. What’s more, the way politicians are using that data is contributing to a more polarized democracy. The key to electoral victory is not a passionate argument, or a principled policy stance, but a race of data.

How Campaigns Use Data

University of London Professor James Curan argued that the key to democracy is access to information. One might think that with access to the world’s information in our pockets, our democracy would be flourishing. While we certainly have more ability to research policies and the positions of prospective politicians, politicians also have greater access to data on us, and are using it to target their campaigns.

Gathering data through a combination of traditional door knocking, surveys and population data, and big tech, politicians now craft their message to target key demographics. As Wired Shut: Copyright and the Shape of Digital Culture author Tarleton Gillespie wrote;

“We are seeing the deployment of data mining in the arenas of political organizing… where the secrets drawn from massive amounts of user data are taken as compelling guidelines for future content production, be it the next microtargeted campaign ad or the next pop phenomenon.”

Political machines take the data they’ve gathered and run it through algorithms, tiering prospective voters into groups based on the possibility of successful engagement. These tiers are based on past interactions and data on the individual, but they’re also based on predictions about a voter. For example, algorithms may predict your race based on your name and location, and assume voting characteristics based on that prediction. Then, political groups will start at the top tiered list and move down to gather volunteers, sign locations, and finally, get out the vote (GOTV) on election day.

The biggest (it could be argued only) role of local campaigns in party-affiliated elections is to gather data to confirm and expand databases. Volunteers go to likely supporters and confirm their support, and remind them to vote on election day. The election then becomes less of trying to convince new people to vote for you, and more about ensuring those who already do support you are galvanized enough to go to the polls. While this strategy was always a portion of the electoral process, it was Obama in his 2012 campaign that institutionalized it as a winning strategy:

“Obama’s campaign began the election year confident it knew the name of every one of the 69,456,897 Americans whose votes had put him in the White House. They may have cast those votes by secret ballot, but Obama’s analysts could look at the Democrats’ vote totals in each precinct and identify the people most likely to have backed him.”

Your Data Isn’t Available to Everyone

It should come as a surprise to no one, but politics is big business. While there are certainly examples of underdog candidates with limited financial resources who win the election, the big players in politics can bring in the billions of dollars it takes to overshadow their opponents.

If a candidate doesn’t have the financial resources to pay for the databases and algorithms, and especially the mountains of data to populate the databases, they are starting from a tremendous disadvantage. This financial barrier means that even if a candidate has the best policy ideas, or the best knowledge of the community, or any other important metric for politics, they don’t have the data to keep up with the GOTV of their wealthy opponents.

In party affiliated politics, this barrier means that it will be harder for anyone who’s not a part of a major political party to gain more than a handful of seats. At the municipal level in Canada for example, though not technically party affiliated, candidates without extensive financial backing or deep connections to the infrastructure of a major political party start at a disadvantage. Infrastructure includes access to the party’s database, which is a symbiotic relationship; the municipal candidate benefits from the data and algorithms, and the party benefits from having their data confirmed prior to their next election. Infrastructure also includes access to volunteer cores, sign locations and donor lists.

The Race for Data has Contributed to Increased Hyper-partisanship

Access to big data has made it more effective for politicians to target those likely to vote for them, and solidify that vote, than to make coalitions and work to convince new voters to support them. While persuasion is still necessary to an extent, it takes a backseat to focusing on solidifying the base. And with what can feel like a neverending election cycle, the election strategy has trickled into governing as well, where individual politicians are better served by digging in on partisan rhetoric versus compromising and building bridges.

Big data ensures that it’s inefficient to dedicate resources to bridge gaps, and so the gap has grown larger. What’s more, demonizing your opponent is a great way to convince your network to take action on election day, further ingraining the tribal mentality of voters, unlikely to even consider the possibility of connecting with opposing views.

Can We Take Back Elections from Big Data?

While big data has become the single biggest resource in politics, it isn’t an unbeatable force. If the populace distrusts a politician enough, they can cast their votes elsewhere. Or, if the political machine bets on their data incorrectly (as some argue happened to the United States Democrats in 2016), results can go either way.

But increased partisanship and reduced access for low-resource candidates are certainly issues that require solutions. The first requirement for any solution to be effective is a knowledgeable voter. The public needs to understand how their data is being used in order to generate the political will for any future changes.

Above that, a transparent process is key. That means transparent algorithms used by politicians, and full disclosure of types of data used by campaigns. Finally, political and social media platforms sharing their data sets with independent researchers and watchdog groups (who are currently priced out of accessing that data) could allow for future research. For any solution to be realistic, we need a voting populace who values the democratic process, and is willing to adopt new measures to protect it.

Justin Draper is a Canadian fiction and non-fiction writer who focuses on themes of politics and culture. He is currently completing his Masters degree in Communication and Technology at the University of Alberta.

Follow Justin on Twitter at @JustinDraperYEG