The Astounding Reason Many People Don’t Know What They’re Voting For

For low-literate adults, navigating a polling place can be an especially intimidating process. Instructions to vote in a 2010 California election. Image via Flickr user Steve Rhodes.

In 1965, the passage of the Voting Rights Act banned the use of literacy tests to gain access to the polls. Disproportionately administered to African American voters, such tests were often written in a deliberately opaque vernacular that the most literate among us would struggle to comprehend. (Sample question from a 1963 Louisiana test: “Print the word vote upside down, but in the correct order.”)

Though it’s tempting to scoff at our misguided civic history, today’s election materials—from wayfinding signage in polling stations to overly complex voter instructions—don’t fare much better. Fifteen years after Florida’s infamous hanging chad, ballots all over the country remain plagued by confusing design choices and questions drafted in awkward, jargon-heavy language, making it unclear to even sophisticated readers whether a “yes” or a “no” vote will lead to the desired result.

CCD’s Field Guides offer helpful tips for usable ballot design. Image via Center for Civic Design.

These days, we can probably chalk up such difficulties to complex bureaucracies and budget constraints, rather than malice. Drew Davies, who with his Nebraska firm Oxide Design Co. leads the AIGA Design for Democracy project while serving as head designer for the Center for Civic Design (CCD), has been working to redesign America’s voting experience since 2007.

It’s important to note that when Davies talks about design in this context, he’s referring to anything that makes the election process smoother. The CCD team finds out what works by running extensive usability tests of various interfaces—and it turns out that “plain language” has proved as crucial to election design as white space or bullet points.

Dana Chisnell—who, as one of America’s foremost experts in ballot and voting systems design, serves as CCD co-director and UX researcher for the White House Digital Service—recalls a usability study she co-developed with plain-language specialist Ginny Redish. It compared two similarly designed electronic ballots containing similar content. The only difference was how the content was worded.

“We tested these two [ballots] against each other, and it turns out that the participants performed better on the plain language ballot,” says Chisnell.

The CCD has had a lot of success outlining its philosophy via the Anywhere Ballot, an idealized prototype intended to show the world what a truly accessible voting process might look like, as well as its Field Guides to Ensuring Voter Intent, which lay out guidelines for an easily comprehensible voting process.

In 2014, the Scottish Referendum ballot was revised to clarify its language after an election commission found the government’s proposed question misleading. Image via Twitter user @FBBDave.

Yet progress has been, as Davies puts it, a little slow. That’s partially because elections in America are run on a county-by-county basis. So that means the CCD needs to make its case to 3,143 counties or county equivalents before everyone in America will reap the benefits. And county officials are often hampered by tight budgets that don’t allow for shifts in the status quo until, says Davies, “a huge debacle shows up on the evening news.”

“We talk all the time in this country about how we cherish the right to vote,” he says.“Yet certainly at the federal and state levels, we aren’t willing to devote any resources to actually creating a process that’s inclusive enough that everyone who wants to vote can vote, or at least vote accurately and understand the system,” says Davies.

To be clear, the number of potential voters disenfranchised by what has turned out to be another literacy test is nothing to sneeze at. Angela Colter, who has collaborated with members from the CCD team in the past* and currently serves as a principal of design research at Philadelphia usability consultancy Electronic Ink, likes to put the problem in context by throwing out a rather startling statistic: About half of U.S. adults have low literacy skills, according to the National Adult Literacy Survey (NALS) and its counterpart, the National Assessment of Adult Literacy (NAAL).

Perhaps even more startling is that, between the two surveys, this half-and-half split between proficient and deficient has remained relatively consistent for well over a generation—which means we can’t blame evolving technologies, trendy educational theories, or an increase in the population of non-native speakers. (In fact, survey findings have demonstrated that, though minority groups are disproportionately affected by low literacy rates, the bulk of low-literate U.S. adults are white and native-born.)

But what is it really like on the other side of the literacy divide? Whitney Quesenbery, author of A Web for Everyone and co-director of the CCD with Chisnell, puts it this way:

“People who don’t read well, people who have low literacy, they’re not dumb. They’re not cognitively or developmentally disabled. They just don’t read well—they don’t go from printed word to idea easily. At the proficient end, there are about 13 percent of us who read complex material, quantitative material, logically complicated material. Around 40 percent of us are average readers, who read well enough, but maybe don’t love reading research papers. And below that, we get basic and below basic. These readers can sound out simple declarative sentences, labels, signs, simple forms, that sort of thing. But not much more. They’re not good at drawing inferences.”

So it’s when we start talking about job applications, medical forms, and yes, ballots, that things become more of a challenge for low-literate individuals. Though debate around this area gets heated—if someone has trouble making inferences, should they be allowed to participate in civic issues?—Colter believes that when we design and write for the functionally illiterate, we should consider it an accommodation, akin to shaping content for those with vision difficulties, unreliable motor skills, or trouble hearing. Says Colter: “You have to design for where people are now.”

Chisnell adds that people with low literacy skills tend to be lower on the socioeconomic scale:

“So this just means that they had less access in all kinds of ways to civic engagement, from getting the benefits that they’re entitled to, to taking part in voting in elections… If I’m not literate, I start thinking things like, ‘I have no way of knowing what the environment is going to be like at the polling place, because it’s probably described somewhere, as opposed to being posted in a video that I could watch that shows me what a polling place looks like and what it is actually like to vote on a ballot.’”

Though videos and imagery help, any way we can make the voting experience less intimidating is a huge step forward. And, thanks to a recent second-round grant from the Future of California Elections (FOCE) initiative, the CCD is doing just that—partnering with the League of Women Voters of California Education Fund on a set of best practices for the design of printed voter information guides.

Two weeks ago, Davies flew halfway across the country to join regional wonks and bureaucrats at the 2015 California Association of Clerks & Election Officials (CACEO) conference in idyllic Sonoma County. There, he worked to bring attention to CCD's recent successes, handed out copies of the field and voter guides, and began laying the groundwork for county-by-county training sessions and implementation. Quesenbery adds that:

“We will be working with three counties—Shasta, Santa Cruz, and Orange—to implement... the model voter guide for the June 2016 primary. The counties will do outreach to their peers, armed with the templates and samples that we will have created with them. We’re hoping that even if the counties don’t adopt the entire design, they will look at their current voter guides to see how they can be improved.”

On the left is Washington's statewide voter registration form from 2008. On the right is the modified form from 2009, which shows improvement in legibility, ease of use, and information flow. Image via AIGA Design for Democracy.

Though Davies is also looking forward to the 2016 election season, he’s a realist at heart, and suggests that a widespread redesigned election experience will be much more likely by 2020 or later. For now, he thinks the hottest topic buzzed about at CACEO was SB505, a California bill currently making its way through the California legislature, with just one more reading and then a vote in assembly. SB505 would allow the Secretary of State to “revise the wording [in the Voter Bill of Rights] as necessary to ensure the use of clear and concise language free from technical terms.”

In other words? To put election materials in plain language. Quesenbery says that if SB505 passes, the resulting Voter Bill of Rights would mostly be a test case. But if it works, there might be a broader bill down the line.

Which wouldn’t just be a celebratory new beginning for CCD, or even America’s low-literate population. It would be a step forward for everyone. After all, as Colter puts it, “We’re all low-literate sometimes. When we’re tired, or stressed, or searching for information on a hospital website in an emergency.”

There has already been some progress in a few states and counties. Quesenbery is also encouraged by election design she’s encountered abroad, though she says, “There are scattered examples of great practice, but none that are a complete package." She was impressed by the look and language of the recent Scottish referendum, which originally asked, “Do you agree that Scotland should be an independent country?” Before the vote, the question was officially revised to read, “Should Scotland be an independent country?”, per an election commission’s recommendations that it:

• be easy to understand
• be to the point
• be unambiguous
• avoid encouraging voters to consider one response more favorably than another
• and avoid misleading voters.

The most postive sign of progress might lie in CCD’s own research. Chisnell says her comparative study with Redish turned up some particularly intriguing results: “If participants used the plain language ballot first, they were more likely to do better on the conventional ballot than the people who used the conventional ballot first.”

More research needs to be done before we can be certain that plain language is an effective tool for honing literacy skills beyond the ballot. But, says Chisnell, “It shows a learning effect. And we’re okay with that.”

*This story has been updated to clarify that, though Angela Colter has worked with members of the Center for Civic Design, she has not worked on any of their projects.