Can the Internet be a Force for Good in our Elections?

By Andrew Kaplan

Try convincing a skeptical friend about the value of voting, and you’ll hear a familiar refrain: “Why bother? My vote isn’t going to make a difference.”

The first time I had this debate, my brain ached with contradiction. Collectively, our votes matter a lot. Research shows that elections with high turnout tend to lead to more generous social policies that more accurately represent the views of the electorate. But on the other hand, statistically, it does seem that each of our individual votes matters little: as Jean Tirole points out in his book Economics for the Common Good, even in close races like the Florida 2000 presidential ballot, the weight of a single vote on the outcome is typically far less than 1%.

Considering these realities, it’s easy to see why turnout among younger voters is only about 50% during a presidential election year. The external benefits of casting my single vote are hard to see, while the costs (i.e.: turning up in person during the workweek, or applying for the correct form of ID if I live in a state with restrictive voting laws) can seem high by comparison. Our bias for immediate gratification can blind us to the less-obvious but important personal and social rewards that voting can eventually beget (for example, the chance to elect more accountable officials or legislators who will prioritize fairer tax laws).

What, then, can we do to motivate young people (or anyone, for that matter,) to show up to the polls? As Tirole might say, it comes down to incentives.

Economists believe that changing behavior is primarily a matter of finding the right incentives. While you may think of incentives as monetary rewards or other “carrots” (like a kid who gets to watch an extra hour of TV for cleaning the bedroom), there are actually many ways to provoke action.

To begin, there are “extrinsic” incentives like money, food, a certificate, etc. – things that are “tangible or physically given to you for accomplishing something”. By contrast, “intrinsic” incentives are “intangible award[s] of recognition, a sense of achievement, or a conscious satisfaction.” Tirole even describes a third category of incentive that centers around building a positive “image” of ourselves in our mind’s eye and in the eyes of our peers.

Combined, these three factors: extrinsic rewards, intrinsic rewards, and rewards to our image, drive a lot (if not all) of our behavior. Assuming an individual believes that the incentives for taking a certain action – say, voting – outweigh the perceived cost of that action, he or she is likelier to complete the action.

Any effort to encourage young people to vote should begin by examining the incentives that would convince them to head to the polls on election day, and whether those incentives are greater the costs of voting. These incentives can be difficult to tease out in focus groups and surveys because, as Nobel laureate economist Daniel Kahneman explains, we are often poor analysts of our own behavior. For instance, we may remember certain experiences fondly even if we were miserable for most of their duration.

The answer, then, is to verify incentives by running tests, an operation I think the internet (and online advertising, specifically,) is well suited for.

Say a random sample of voters 18-29 in a certain district tells you that they don’t vote because they don’t believe their ballot will make a difference in who wins or what policies our government enacts. In other words, their (extrinsic) incentive for voting might be to see evidence that their ballot contributes to a real-world outcome. In one test, we might target these voters with online ads emphasizing the very real impact of voting; for example, by promoting videos about reputable local representatives who won based on just a handful of votes from their fellow citizens, or interviews with real voters who feel confident that their vote played an important role in a referendum.

By monitoring turnout in our test district versus a control district (i.e.: one that’s demographically similar), it may be possible to correlate our ad campaign with higher electoral turnout and better understand the kinds of incentives that inspire some young people to head to the polls. Such understanding can help us shape more effective PSAs to drive greater turnout in future elections, or help our elected officials better understand voter attitudes and their impact on electoral outcomes.

Deploying the scientific method and targeted, online advertising to promote and measure voter turnout is not without precedent, and in fact it’s a growing area of innovation. As far back as 2010, Facebook found that showing people “social proof” that their friends had voted significantly increased turnout in certain races, an example of the self-image incentive at work. More recently, firms like the Analyst Institute and Acronym have begun leveraging data-driven, Web-based strategies to get out the vote. These organizations often adopt social missions to advance equitable, progressive political ideas or increase participation among younger voters.

I don’t want to downplay the risks that online content and social media carry. The example of Cambridge Analytica demonstrates that groups can wield online marketing and user data for ignoble purposes as easily as they can for virtuous ones. To put individuals’ privacy, data security, and vote integrity first, I believe that tests like the ones I’m proposing should only be conducted with clear, affirmative user consent, a verifiable guarantee that no personally identifiable data will be shared with third parties, and a system of independent monitoring to ensure the testing organization will comply with all election laws both in letter and in spirit. It’s vital that the organizations running the ads be transparent about who they are and who funds them, so citizens can feel secure knowing that malicious actors aren’t using propaganda and online experiments to undermine the democratic process.

As of late, discussions about the Internet’s role in our republic have concentrated on the spread of misinformation and the exacerbation of prejudice. But with well-designed, responsible testing, we can harness the power of the Web to make our government more representative, participatory, and progressive.

Andrew Kaplan works in online marketing in San Francisco.

Leave a Reply