Ballot design

While this example illustrates an extreme consequence of poor ballot design, there are many other design choices, made every election cycle, that attract fewer headlines but have significant impacts on voting behavior.

This design choice led to a 9% drop in the number of votes for that particular race compared to other counties in the same congressional district that used different ballots. This poor design choice likely impacted two distinct groups, the first being more experienced voters, referred to as “rushers.” These voters likely rushed through the instructions in the first column, skipping over the race in the bottom section.  The second group impacted was low-literacy voters, referred to as “skippers.” These voters likely just read the top of the instructions paragraph and decide to continue to the rest of the ballot. 

Poor ballot organization can also lead to overvoting, where a voter selects more candidates than is allowed for a single race. For example, California held a special election to replace retiring U.S. Senator Barbara Boxer in 2016. 34 candidates ran for the seat, presenting quite a challenge for county election officials who had to design ballots to accommodate a large number of candidates while also ensuring readable fonts and proper ballot order protocol. Bilingual ballots further constrained the number of candidates that would fit on a ballot page. 

As a result, the majority of counties opted to split the content into two columns, causing significant confusion for voters who were used to marking one name per column. After the election, it was determined that over 2.8% of ballots contained overvotes for the U.S. Senate seat. Researchers David Kimball and Martha Kropf found that the overvote rate was 3.4% in counties that split the contest into two columns and 0.8% in counties that opted for a single column, shown in more detail below.

Poor navigability can also arise with electronic ballots that list multiple contests on the same screen, leading voters to miss more questions than if each contest was displayed on its own. This is a basic principle of interface usability: people are far less likely to miss questions if they are asked one question at a time and proceed to a new screen only after they have answered the previous question. This principle has a wide range of digital use cases, ranging from ATM machines to online surveys. 

The 2018 race for lieutenant governor in Georgia is a good example of how interface usability impacts the navigability of electronic ballots. In this race, ballots in several jurisdictions displayed candidates for lieutenant governor on the same screen as other contests. As a result, 4% fewer votes were cast for lieutenant governor than for governor, and 2.7 percent fewer votes were cast for lieutenant governor than for secretary of state. That represents over 100,000 “lost” votes for lieutenant governor in a contest where the margin of victory was just over 123,000 votes.

Complex Races

Good ballot design - specifically intuitive ballot design - is especially important for elections that use instant runoff voting, given that voters are likely less familiar with the process. Instant runoff ballots have been designed in many different ways, but two of the most common formats use either an oval grid or columns with arrows (see below for examples of both formats). 

Recent research from the Ranked Choice Voting Resource Center studied the effects of ranked choice ballot design and found that using arrow-column ballots led to more errors than column-grid ballots, and that the rate of these errors increased as more candidates were added. More specifically, the Center found that arrow ballots lead to a 0.47-percentage-point increase in the error rate compared to oval ballots, meaning that in an election with roughly 200,000 voters, an additional 940 ballots could be  uncountable due to arrow ballot-related errors.

Many localities that previously used arrow ballots for RCV elections have since abandoned them in favor of more intuitive grid and column ballots. However, non-RCV races where voters mark preferences for different numbers of candidates tend to cause confusion as well. Examples include races where there is more than one seat to fill, such as district and city council races. In such cases, it is crucial for ballots to clearly indicate how many candidates a voter can choose for a single race. It may also be helpful for electronic ballots to limit the number preferences a voter can rank as they mark their ballot.

Confusing Details

Deceptively simple details like font type, font size, and paragraph alignment also present issues for ballot design. These elements make a huge difference in a voter’s ability to navigate a ballot. For instance, the Center for Civic Design recommends using lowercase letters, avoiding centered type, and using one cohesive font type and size. These changes make it less likely that a voter will skip information. Other recommendations include adhering to design best practices prevalent in other industries, such as: 

• Sentence-case (using both upper- and lower-case letters) makes shapes that are easier to recognize, and is more legible than uppercase letters. 
• Left-aligned type is easier to read than centered type, which forces the eye to hunt for the start of the next line. 
• Using only one font makes the ballot more unified, while different fonts force voters to stop reading and adjust.

Other, similarly small-sounding adjustments can be used to create a clear information hierarchy—a critical tool that helps voters navigate a ballot. This can include making the title of the ballot the most prominent thing on the page, followed by the contest headers, and bolding candidates’ names but not their party affiliation.

Likewise, visuals can be extremely helpful in maximizing ballot navigability, especially for low-literacy voters—but done poorly, they can create more confusion than they resolve. The placement, purpose, color, and shading of illustrations can all affect whether a ballot’s visuals are a help or hindrance; this holds true for paper and electronic ballots alike.

Ballot Design Implications for Absentee Ballots

These ballot design principles extend to absentee ballot envelopes as well. This has become more important than ever in recent years, as an increasing percentage of the electorate is voting by mail: 29% of voters cast a ballot by mail in 2024, according to the 2024 Survey of the Performance of American Elections. In many cases, ballot envelopes are covered with required information, which can lead to confusion and errors. It is even more critical for election officials to ensure absentee ballots have clear and obvious instructions for voters, because the voter may not have someone on hand to clarify or troubleshoot potential issues.

Signature verification is a common security measure states use to ensure mail ballots are cast only by eligible voters. Where signatures are required, ballots with incorrectly placed or wrong signatures make up a significant percentage of those that are sent back to the voter to be cured, or in some states, simply thrown away. This caused issues during the 2018 midterm elections in Gwinnett County, Georgia, where a confusing ballot envelope led the mail ballot rejection rate to skyrocket. 

The county’s absentee ballots that year did not have an “X” or box to draw attention to the signature line, and the instructions below the signature line read “Signature or mark of elector,” causing confusion with many voters who were not familiar with the term “electors” or did not think of themselves as an elector. The envelope’s language format was also inconsistent. While the English- and Spanish-language voter oaths were placed vertically, one on top of the other, the English and Spanish oaths for the “person assisting elector” were placed side by side. Ultimately, a report conducted by the Brennan Center found that the county rejected a total of 1,690 mail-in ballots ― 21 percent of those rejected in Georgia during that election ― even though Gwinnett County accounted for only 10 percent of the mail-in ballots submitted in the state.

Election officials continue to work on issues like these, often in collaboration with organizations and experts who study the intersections of behavior, user experience, and design. The Center for Civic Design, for example, has collaborated with the U.S. Postal Service to design ballot envelope templates that are more intuitive for voters to use. An example of one of the envelopes is shown below: in it, we can see that the signature line is clearly marked, and there’s a checklist to ensure voters have completed all of the necessary steps to get their vote counted. This kind of intentional design can also make election officials’ jobs easier: the suggested color palette in this template, for example, not only meets the official ADA requirements for accessibility, but also helps election workers sort and process votes accurately and efficiently.