So, you want to win an election, do you? You’re worried because the poll numbers suggest the vote will be close, and you’re a bit behind. You don’t want to engage in anything illegal like ballot-box stuffing or hacking voting machines. Essentially, you want to steal the election, but in a quiet, legal way, right?
Have we got a deal for you.
Given the current environment, with widespread apprehension about election security and media reports about voter suppression concerns, you might think this would be a challenge. Trust us when we tell you, it won’t be. Here is how you win the election: You control the design of the ballots.
Remember Florida in 2000? Recounts, hanging chads and court rulings. The “final” tally had George W. Bush win the state by less than 600 votes. An unusual punch-card ballot design in Palm Beach County, referred to as the “butterfly ballot,” led thousands of voters into making errors. Some lost their vote entirely and others accidentally voted for the wrong guy. The ballot design cost Al GoreAlbert (Al) Arnold GoreTrump's election fraud claims pose risks for GOP in midterms Don't 'misunderestimate' George W. Bush Why the pro-choice movement must go on the offensive MORE the state, and thus the White House.
What we’re selling is a way to capitalize on known psychological principles. We know, for instance, that the order that candidates are listed on the ballot influences how many votes the candidates get — the candidate listed first has a clear advantage. We’ll try to get your name first, but that’s hard to do. Instead, what we’re selling is deliberately flawed ballot design. Designs that are intended to lead voters — not all voters, mind you, but enough — into making errors that invalidate or alter their votes, so you can move the needle a couple points in your direction. It should win you that close election.
How do we do this? We make subtle changes to ballots in ways that make them just a little bit harder for voters to navigate correctly. If the voters are voting on an electronic system, we’ll change the layout for your race so it’s different. Maybe we’ll put two contests on the same screen where every other contest gets its own screen, as was done in Sarasota County, Fla., in 2006. Almost 15 percent of the voters there missed the smaller U.S. House race on the page.
You say your jurisdiction votes mostly on paper? That makes our job even easier. You’re in a long contest with lots of other candidates? Then we’ll make sure the list of candidates is split across two columns, like they did in Kewaunee County, Wis., in 2002 with the governor’s race. Some voters (about 11 percent in that election) won’t notice it’s not two separate races, and will vote twice, invalidating their vote. They’re even trying this in Georgia in 2020! Oh, you say your contest is small, just you and one other candidate? Great, we’ll hide it at the bottom of a column of instructions, like they did in Broward County, Fla., in 2018. Some 30,000 voters missed the race for U.S. Senate there, and the statewide margin of victory was about a third of that.
Voting by mail? Even easier. Not only can we use all the regular tricks for paper ballots, there’s no poll worker to ask for help. We’ll have complicated instructions, and all kinds of non-obvious requirements (like two envelopes, or a witness signature) that, if not followed, will lead to ballots being rejected. Easy-peasy.
Ahh, you noticed that these problems will apply to all voters, so how will it help you? Simple: we’ll only deploy these bad designs in jurisdictions where your opponent is popular. Getting a nice high error rate from a cluster of voters who predominantly vote against you, well, that’s as good as any computer hack.
Of course, the thing about these shenanigans is that we can’t make voters cast their vote incorrectly. Unfortunately, those pesky voters have the power to thwart us. How? If they make a list of who they want to vote for (perhaps a marked-up copy of a sample ballot) they can make sure they don’t make a mistake. They can ask a poll worker if they run into any problems. Worse, if they double-check their ballot carefully before casting it, our mischief won’t work.
But really, how many voters will go to the trouble of doing all that?
Mike Byrne and Phil Kortum are professors in the Department of Psychological Sciences at Rice University, where they specialize in research on ballot design, electronic voting and voter security. Together, the authors received the Human Factors and Ergonomics Society's Jack A. Kraft Innovator Award in 2020 for their significant efforts to extend the application of human-factors principles and methods into the realm of voting.