Political Hay

Voting Error

It matters in close races, but only as an excuse.

By 9.18.02

Send to Kindle

"Here we go again!" media outlets gleefully chirped last week as they got to report another election breakdown in Florida. Involving a familiar cast of characters, the story line was almost too perfect. Numerous polling places failed to open on time, resulting in some voters -- including Democratic gubernatorial primary candidate Janet Reno -- being turned away. Gov. Jeb Bush had to sign an executive order to keep the polls open two hours later than originally scheduled. Voting machines at a number of sites, primarily in the southern counties, failed to start up or work correctly, confusing hundreds of poll workers who on primary day apparently were seeing the newfangled touchscreen voting terminals for the first time.

Such reports merely confirmed the nation's view of elections in general and Florida in particular: that both were a mess in 2000, remain a mess in 2002, and without millions of dollars spent on upgrades, will still be a mess when the presidential election rolls around in 2004. But that view is dead wrong. There was no widespread voting equipment failure in 2000, 2002 saw few problems that couldn't have been avoided with little expense, and 2004 will be fine as long as Congress and the states leave well enough alone.

Yes, that's right: There was no widespread voting equipment failure in 2000. But how can that be, given the close race in Florida, the hanging chads, the weeks of recounts? Combine the close division of the American electorate among Republicans and Democrats with an increase in first-time voters for the presidential election, and what you got in 2000 was an abundance of human error.

Technology (or a lack of it) was not to blame for Bush vs. Gore dragging on for weeks. In fact, voting machines have long been tested and certified at both the state and federal levels as meeting specific levels of accuracy. This testing results in extremely low machine failure rates, on the order of "no more than 1 in 250,000 ballots for federal certification and no more than 1 in 1,000,000 ballots in some states," according to a 2001 study by MIT and CalTech. Thanks to this certification process, only accurate voting machines are purchased by local officials for use in precincts around the country.

Although the machines themselves are tested to a high degree of accuracy, most precincts in the U.S. still see a voting error rate of about 2 percent. What drives this is, the MIT/CalTech study found, is "human factors."

The problem of human error was known to poll administrators and workers before 2000. After the 1998 election, the Dallas County elections administrator told the Dallas Morning News that voting problems are "never the same" and that "they come from out of nowhere." An election worker in North Carolina told the paper that "Murphy's Law does apply, especially for elections."

Social scientists, too, have been aware of the problem for decades. An article in the Political Science Quarterly stated rather bluntly: "A person of any intelligence, who knows how to read and has taken the trouble to find out the names of the candidates, ought seemingly to have no difficulty in casting his vote correctly even with the most confusing of these [voting] forms. But...figures show, if nothing else, the proclivity of the American voter to make mistakes where there is any possible excuse for his doing so." Those wise words were written in 1906.

The recurring nature of voter error shows the underlying folly of relying on technology to "fix" our election system. Fortunately, the most expansive and expensive national bills on voting reform died in Congress last year before a whole lot of money could be thrown at the problem. Also fortunate was that most states, expecting a wave of post-2000 "reforms," waited to find out what the Feds would do before budgeting their own money for equipment upgrades.

Which brings us back to the local level, where much of the blame is said to lie. Isn't our decentralized system of elections terribly unfair, with the poor and educationally disadvantaged getting stuck with machines from the 1920s -- machines that can't possibly be as accurate as the newer and sleeker computer terminals? Shouldn't there be a national office of elections to fix these problems, ensuring absolutely equal access to all?

In short: No. Lower-income districts often do have older voting machines, but that alone doesn't explain why they suffer from error rates that can be double or triple the national average. Technology can only do so much. In fact, first-timers and voters with lower levels of education often fail to follow simple directions correctly, make obvious mistakes like voting for two candidates for the same office, and fail to ask poll workers for help when they need it.

If reformers must spend money on something, it should be spent on education -- perhaps conducted in high school civics classes or immediately outside each and every polling place on election day -- to teach citizens how to correctly work the equipment and fill in the ballots. In the end, though, voters need to take responsibility for asking those gray-haired poll workers when they don't understand something. In this small way, we can all help to avoid another 2000 -- unless, of course, the election isn't close, in which case all such matters will be quickly forgotten.

Like this Article

Print this Article

Print Article
About the Author

Tracy Robinson is a 2004 graduate of the George Mason University School of Law.