Kelly Born is a Program Officer at the Hewlett Foundation, where she works on both Special Projects and our Madison Initiative. Tom Glaisyer is the program director of the Democracy Fund and Jonathan Kartt works in Programs & Evaluation for the Rita Allen Foundation.
How will voters find information in 2014? For those who care about US democracy, this question is front and center in a world where both the structure of the news media and the channels through which voters get information are in flux. In the not too distant past, voters received most of their information about candidates and ballot measures through mass market dailies and TV or radio—places where the message was mediated by gatekeepers. The only opportunity to directly communicate with voters was through paid advertising or in-person contact. Nowadays, candidates have limitless options to directly reach voters—even television, when delivered via satellite, permits hyper targeting of political advertising messages.
But it’s not just campaigns that are exploiting these new digital opportunities—a host of (mostly new) organizations, non-profit and for-profit, are seeking not to win a vote, but to inform voters about their options.
It’s an exciting time for the field. Abroad, websites that match voters to policy positions held by parties, so-called voting advice applications, have seen significant adoption. In Germany, for example, Wahl-o-Mat was queried 13.2M times in 2013—not bad when you consider there are only 80M people in the country. In the US, we have encountered dozens of similar sites such as Vote411, ethePeople, and Project VoteSmart.
The digitization of data permits an increasing amount of contextual information to be added to what was previously just a thumbnail sketch of a candidate or issue. For example, information on candidates or ballot initiatives can now be combined with “rules of the road” on where and when to vote, and what materials to bring. This digital “plumbing” is often under-appreciated—Google’s Civic Information API provide a way to lookup polling places in 2014 and listed the candidates on the ballot. It builds on data from the PEW Charitable Trust’s Voting Information Project and augments a recently developed iOS app.
Recognizing the possibilities in this emerging ecosystem of voter information, the Hewlett Foundation, the Rita Allen Foundation and the Democracy Fund partnered to explore the dozens of voter information websites that have developed in the last few years. We examined a number of dimensions:
Candidates vs Ballot Initiatives (or both): Many of the sites focus on candidates, while others like Healthy Democracy in Oregon and Washington State’s Living Voters Guide have (until recently) focused exclusively on ballot measures. Others like ethePeople, Project VoteSmart, and PollVault, cover both.
Geographic Scope: Many provide national coverage, whereas others, like ethePeople, partner with media and civics groups in specific states or localities. Maplight’s Voter’s Edge cover national races while also offering some down-ballot coverage in particular states (in this case, California).
Audience: Some, like Ballotpedia, provide detailed information that might appeal more to policy wonks like ourselves, whereas Voter’s Edge or Who’s On The Ballot seek to serve those who prefer a less detailed view.
Approach: Sites like Voter’s Edge provide “just the facts” (on a lot of dimensions, including candidate’s prior jobs, campaign funding sources, etc.). Others, like the newly launched Crowdpac,use campaign funding sources to predict candidates’ positions, in an attempt to address the challenge of comparing a 30-year incumbent’s record to that of a first-time challenger who has never held office. ISideWith uses matching algorithms – and has now paired more than 11 million users with their “ideal” candidates based on answers to basic philosophical and political questions (e.g., “what is your stance on taxation?”). Still others actually involve citizens in the deliberative process: Healthy Democracy in Oregon convenes a representative panel of dozens of citizens for a week to evaluate the pros and cons of a particular ballot initiative. The information is then shared with voters in the official voting guide. Research has shown how valued that information has been—a majority of Oregonians were aware of the tool, and roughly two thirds who read the CIR statements found them helpful when deciding how to vote. In Washington State the Living Voters Guide has utilized a deliberative platform to allow voters to share why they are in favor of or opposed to a particular initiative.
Business Models: Half of what we found are for-profit operations like Crowdpac and Poll Vault. The other half (most of what we’ve discussed herein) are nonprofit. So we spoke with venture capitalists who had invested in several of the for-profit competitors to understand their reasons for doing so, and to ensure that we felt there was a good rationale for philanthropic investment in this space.
Operating and Partnership Approaches: Some, like Project VoteSmart, rely on teams of dedicated interns, while others are striving towards more automated systems. We also looked at organizations’ partnerships—many like ethePeople are collaborating extensively as part of their model, others are closer to independent operators.
Use: Finally, we looked at use. Not much is known about the end-users of these types of voting information services beyond broad demographic statistics. In terms of numbers, some platforms have received a fair amount of uptake, whereas others are so new that no usage data is even available yet—however, no site appears to have come close to Wahl-o-Mat’s success in Germany.
This wide variety of activity left us with lots of questions: whether and how to support this field, who to partner with, and on what kinds of projects? We have begun to explore these questions, and will discuss our early work on this topic in a follow-up post next week.