i-us Wire
DWH
Will-Harris Wire Home
Archives

Search the wire

Wire Logo
Wire Logo

<How bad graphic design
   helped decide the Presidency>

By Daniel Will-Harris 

After the election: In the end, it wasn’t just this one thing, but many discrepancies in Florida, and a Supreme Court that ruled the opposite of what they normally ruled about states rights that selected the current president. While I support the office of the president, I do believe that the president should be elected not selected and that election reform needs to take place to avoid a repeat of this unfortunate event. Clearly the leadership of the United States should not be decided by a few hundred votes from one state. Our modern technology certainly should allow for accurate vote counts, and the electoral college system needs to be reconsidered.

NOVEMBER 8, 2000--I write this the morning after the national election for President of the United States. Although Al Gore clearly won the popular vote, in this country the election is decided by the “electoral college,” a winner-takes-all system where if a candidate wins by even one vote in a state, he or she receives all the electoral college votes in that state.

So right now, the election has come down to only a few hundred votes, in one state, Florida. And while there are questions about voter irregularities and other issues, and all the absentee ballots won’t be counted for 10 days, one thing is clear—this election may be decided by graphic design.

I’ve been an expert witness in legal trials involving typography, layout and elections, and I know there are issues here that could lead to protracted legal debates on the validity of the Florida election results. Since layout is something that few people other than graphic designers understand, I want to take this time to explain where the problem is—and why I agree that there is a problem.

In a normal election, a loss of 3,000 votes to the wrong candidate might make little difference—even in a small, local election. And certainly in any other presidential race, 3,000 votes would be totally inconsequential.

But when you have two candidates who are only a few hundred votes apart, in a state with 6 million votes, and that state is the deciding one for the electoral collage, then suddenly the entire race can hinge upon the poor and misleading design of this one ballot.

>How bad design confused voters

 

Here you see the ballot. The first piece of bad design is the layout. In this type of “butterfly” ballot, the circles in the center are tiny holes you have to punch out with a small tool. This creates an old-style punch card for the computer to tabulate, and this system is used in many places—they used to use this where I live but changed last year to a much easier to use system.

The problem with the design of this system to begin with, is that the holes don’t appear to consistently line up with the person or proposition you’re voting on. This is such basic bad design that it causes problems from the start. When I used to use this system, I’d vote, take out the card, and compare the punched numbers with the numbers of the candidates on the ballet. Even I made mistakes and had to redo things—and I checked things carefully.

 

Here you can see the confusion caused by the fact that that line above “Democrats” points directly to the hole for Buchannan. Since this line is longer and more prominent than the small arrow next to the “5”—it’s understandable that people, especially more elderly people with poorer reading eyesight, could be easily confused.

They see this long line, pointing to a circle—and they think that’s the correct circle. Even younger voters were confused. The red and blue of the Republican and Democratic titles stands out from the rest of the black text, so the Democratic voter would be attracted to that title, and follow the line that’s just above it to the wrong circle.

It’s also confusing because we tend to relate things in order. The list on the left has Bush, then Gore, then Browne. So it’s human nature to assume that the circles are in this one-two-three order, too.

Because the name AL GORE is short, it has less of a visual connection with the vertically centered 5-> to the right—so the eye would once again go to the right from the horizontal line, pointing to the wrong circle.

Finally, because we do read from left to right, but are stopped by that thick black line, many people might not even see the names on the right. That’s of course not fair to those candidates from smaller parties, but it causes further confusion for people who want to vote in the left column, because they don’t even notice that the names on the right exist. And if you were a candidate in the right column, your voters would have a harder time figuring out to vote backwards, to the left.

Here’s how the South Florida Sentinel newspaper pictured it:

 

>Not just theory

Now, you might say, “Well, that’s a good theory, but can you prove it.” Yes, I can. In this county, Buchannan received over 3,000 votes. That was far out of range with what he received in other Florida counties. Furthermore, the demographic makeup of Palm Beach is more Democrat than Republican, and the predominantly elderly Jewish voters would not be inclined to vote for Buchannan, a candidate who holds views opposed to theirs.

This one county was responsible for 25% of all votes for Buchannan in Florida. That’s not realistic, or consistent with exit-polls that were accurate around the country. Furthermore, the socialist candidate received 50% of all votes in Florida from this one county, a result also inconsistent with the demographic--and clearly due to the ballot design.

So you have a badly designed ballot, voters who have complained they were confused, and poll results that show a disproportionate number of votes for a candidate—especially disproportionate given the demographics of the area.

>What can we do about it?

Because of this issue, and the closeness of the race, If Bush wins, Democrats could feel that the election results were unfair and inaccurate, given the misleading nature of the ballot.

The end result is that this could be sent to the courts to decide. Do I think it will get to that point? I don’t know—because graphic design is so esoteric to most people—including judges, that they may not understand what the problem really is.

The real problem is that almost 3,000 voters may not have had their true voice in this election, and in the end, that affects more than 250 million people in this country, and billions all over the planet, and the planet itself.

I suggest that these voters be given the chance to vote again--this time with a more well-designed ballot, so their voices, voices which can decide the election, are fairly heard.

So don’t let anyone tell you graphic design isn’t important.

And lest you think that I’m the only one who’s come to this logical conclusion, there are an increasing number of designers and UI experts who have reached the same conclusion, using different methodologies.

--DwH

11/9/00, 12:34pm PST: On ABC’s Nightline, it was reported that there were almost 19,000 “double-punched” ballots. These are ballots where someone punched two holes for the same office. These very well could have been because of this flawed ballot design.

If you agree there’s a problem here, send email to the following people with your comments:

webmaster@fec.gov (Federal Elections Commission)

letters@nytimes.com  (Letter to NY Times)

Also write to your elected officials, your congress person, senator, and the president.

To quickly and easily find out contact information (including email addresses) for your officials, visit:
ElectedNet
 

Read more Will-Harris Wire articles

 

 

Rectangle

[Previous Article]

[Next Article]

[Wire Home] [Archives] [ Will-Harris House]

Daniel Will-Harris is a designer and author whose work can be found at http://www.will-harris.com. His site features TypoFile Magazine and Esperfonto, the web’s only typeface selection system. He may be reached via e-mail.

Visit eFuse, the friendly place to learn how to build a better web site

Copyright Daniel Will-Harris, 2001, All Rights Reserved