The Effect of Usability Testing on Ballot Design
It is hard to overstate the importance of free and fair elections in democratic nations. To truly meet the criteria democracy demands, we must make sure that every component of the voting process is designed with the voter in mind, especially in order to minimize the room for human error on both the voting and processing ends. How can human factors help play a role in capturing the true will of voters?
There’s more than one way we can approach answering this question. We might apply human computer interaction principles to the ballot processing machinery and software. We might seek to explore the logistics of the postal service and determine how well the methods used to collect and deliver mail-in ballots account for human usability. One of the most direct ways we can observe the relationship of usability on a fair election is in the design of the ballot itself.
What makes a good ballot? The answer seems relatively straight-forward: a clear system for marking your candidate of choice. Historically, however, ballots have not always provided a successful system for voters to use.
To understand what makes a good ballot, let’s examine one of the most controversial examples of a poor ballot design in recent history: Palm Beach County’s “butterfly” ballot from the United States Presidential Election of 2000. In this extremely close election, it eventually became clear that the President-elect, George W. Bush or Al Gore, would finally be determined by the outcome of the votes cast in Florida. The spread of votes between the two presidential candidates came down to a difference of 537 among almost 6 million total votes (0.009%) in the state, leading to subsequent litigation regarding recounts over the following weeks and a controversial Supreme Court decision that resulted in George Bush being declared President.
So, why is this a bad ballot? For several reasons. To select their candidate, a voter was required to hole-punch a selection corresponding to their choice. Many selected the second hole-punch opting for Pat Buchanan with the intention of selecting Al Gore. A study published in the American Political Science Review determined that:
“The butterfly ballot used in Palm Beach County, Florida, in the 2000 presidential election caused more than 2,000 Democratic voters to vote by mistake for Reform candidate Pat Buchanan, a number larger than George W. Bush’s certified margin of victory in Florida.”
Further troubling was the issue of “hanging chads” (incomplete hole punches), “pregnant chads” (hole punches that were indented but not pierced), and both ballots with more than one vote or no votes at all. This ballot design may not be entirely to blame, but it certainly had a consequential impact on who was determined to be the President after this election.
Luckily, we do have design principles we can rely on to make effective ballots. The Center for Civic Design has compiled a ballot design checklist that give us eight guiding questions that can be used to aid ballot design. For example, “Is the design consistent, avoiding bias, throughout the ballot?” and “Do instructions have all the information needed to vote accurately?” The Brennan Center for Justice also offers some valuable resources for improving the quality of ballot design and correcting common ballot design flaws. The Center suggests that design expectations should not be rigid rules issued by the state, but rather that states should have flexible standards and provide resources to local officials.
However, there is not a uniform consensus on how states can help local officials create high quality ballots. According to Dr. Nichelle Williams, Director of Research at the U.S. Election Assistance Commission, the lack of emphasis on ballot design is a common problem and that states should play a more active role in ballot design. In this March 2019 government blog post, she writes:
“Most voting systems do not work to optimize ballot design, so experts recommend that states are proactive and assist localities by working with vendors to come up with uniform standards. States can provide vendors with legislative requirements and vendors can reply with how their voting systems can meet these state requirements.”
Whether or not states should maintain strict ballot guidelines or offer more freedom to local officials is a great question that there may be no clear answer to. But, we can recognize that guidelines only go so far. They can’t prevent every possible voter error, which is precisely why institutions invested in improving the efficacy of our voting system further recommend testing a proposed ballot’s usability — gather participants, observe their experience, document their problems, and implement the necessary changes.
If we want to practice elections as fairly as possible, we must minimize the risk of a poorly designed ballot influencing a vote’s outcome rather than capturing the true intent of the voters. Implementing design guidelines and usability tests will help us improve the quality of our elections, across the nation.
