Back in 2000, I was teaching a “writing for the web” class at a local college. Since I was (and am) a Usability guy, I took a week or two to teach the basics of usability analysis.
Fortunately we had a prime case study in the news.
The “Butterfly Ballot”, viewed through the non-political, pretty-objective standards usability people use, was a complete disaster. My class and I came up with a solid page of feedback just by going over it; I would have added that “observing people using the ballot, with an aim toward improving the design, would have been interesting” – but of course, we got that all over the news for the next two months.
Before the election? That might have been helpful.
The Times, eight years later, gets the message:
The butterfly ballot in Palm Beach County, Fla., was one of the great debacles in election history. It was so confusing that it was hard to tell which hole to punch to cast a vote for a particular candidate. Many people intending to vote for Al Gore accidentally punched the hole for Patrick Buchanan or punched holes for both Mr. Gore and Mr. Buchanan, which disqualified their votes.
The design, by the way, was equally likely to draw votes away from Bush. But we’ll leave the Times alone for the moment.
The controversy should have led to sweeping reforms, but it didn’t. A study by the Brennan Center for Justice at New York University School of Law lists 13 ballot problems that show up around the country in election after election. One is creating a layout in which it is unclear what hole voters need to punch — or where they need to place a mark — to cast a vote for a particular candidate. Another is placing more than one contest on the same screen of a computer voting machine, which often leads voters not to vote in one of the races. Making matters worse, the instructions that accompany ballots are often confusing.
The response from some commentators – “anyone who’s too stupid to use a ballot is too stupid to vote” – is the kind of thing most interaction designers have heard from programmers, if we’ve been in the business long enough.
It’s wrong, of course; ballots should no more actively confuse, obfuscate or muddle people than should the software you use to do your job or balance your checkbook. In the private sector, the company with the more usable product usually wins. It’s a major reason “Quicken” has beaten “Microsoft Money” for the past decade and a half. It’s why Best Buy and Target and Amazon spend millions designing their real and online stores, and millions more observing how real people do interacting with them under real-world conditions.
Is the integrity of elections as important?
Congress should require that ballots used in federal elections meet minimum design standards. It should also mandate pre-election usability testing and make funds available for it. States and localities need to draw up better guidelines for how ballots are designed and clearer instructions to voters.
Every once in a while, the Times gets one right.
They should also publicly report after each election how many votes are lost because of miscast ballots.
There are, on the other hand, easier and cheaper ways to accomplish this.
Yeah, standards are nice, but we must be vigilent of the federal gov’t getting too close control of local elections. In a win-at-any-price atmosphere (see Washington State), the party that controls congress could try to work to make sure they win elections.
Which brings about some questions, the answers to which I think I know – but I’ve been wrong before.
Is the “usability” craft non-partisan? I know usability people on all sides of the political fence; some as conservative as me, others far to the left, many in the middle or apathetic.
And the actual craft and process of designing things to be usable should be politically agnostic; it answers some fairly empirical questions. Can people read what you wrote? Can people accomplish what the are supposed to do with the thing you designed? What percentage of attempts to do the thing they’re supposed to do are successful? What percentage is acceptable for your product? If not (for any of the above), how do we fix it? Do the fixes make things better? How much?
And so on.
From what you are saying, I think political advantage can be removed. You are talking something as simple as the type of ballot.
the type of ballot, how the information is laid out, the typface and size, the mechanism for voting, the ergnomics of the polling station and marking/punching/casting system…
Given the incredible number and diversity of voting machines in use usability experts would be gainfully employed for the next six decades just crafting standards for the machines now in use. Job security wasn’t one of your considerations was it?
The places with the worst ballot design I’ve ever seen (IL, Broward County FL, San Diego CA) were run by Democrats who seemed to ignore usability and emphasize style over clarity. At least the ballots in IL made it easy to do a straight party-line vote, though, rather than forcing you to figure out the ballot, although that might have been the reason the ballot was so badly designed.
They should also publicly report after each election how many votes are lost because of miscast ballots.
Is the Times insane? This would do nothing but convince the people who voted for the loser that the election was ‘stolen’ from them. This undermines the legitimacy of the vote & weakens the social contract. In the same op-ed piece the Times engages in this second guessing of results it doesn’t like:
The Palm Beach Post’s postelection analysis found that the butterfly ballot ended up costing Mr. Gore far more votes than the 537 by which he lost Florida — and the presidency.
Apparently the publisher & editors of the Palm Beach Post are now the authoritative source on the legitimacy of the FL ballot count. Because they say Gore really won! Note the hypocrisy of the Times‘ editorial writers — they don’t even mention analyzing ballot results in places other than Palm Beach that might have yielded other results.
Every medium & large election has ballots that are miscast or counted wrong. It doesn’t matter what the ballot looks like or how it is designed as long as misvotes aren’t systematic.
Having studied and used usability from a programmers perspective, I can’t think of anything that would be inherently partisan. Because when it comes down to it, it’s rather content independent.
It’s not WHAT is being displayed, it’s HOW, WHERE and WHEN. The same ballot listed above could be a ballot for favorite TV sitcom and the same lessons could be learned and same suggestions made.
Mitch,
I agree with you on good usability design being agnostic. With parties suspicious of each other and generally being design-ignorant, however, it’s unlikely they’ll welcome change easily.
Also, elections are local, and technologies — and therefore ballot designs — aren’t necessarily uniform across the state.
I was part of group from the state graphic design association that suggested Minnesota do a usability review of its ballots in the wake of the 2000 election. It went nowhere.
Politics aside, the public sector doesn’t place a premium on good graphic design, and too many consumers still associate design with cosmetics rather than function. That means there’d be little support for spending the money it would take to do the project right.