Another election season draws to a close, and millions of U.S. voters cast ballots on electronic voting machines this time around, most for the first time.
It's often hard to see what's missing. But for those voters, what was missing was a paper trail.
Touchscreen voting, as is currently certified in California and the U.S., does not require there to be a paper trail. "Modern" means paperless -- no more ballots, no more chad, just data stored on a machine.
Paper-based voting systems, though less efficient, have many strengths that the touchscreen systems do not have. Paper-based systems allow voters to make their choices on a piece of paper that is inserted into a locked box and exists as physical proof of how they voted. This kind of security is transparent and understandable to even the most unsophisticated of voters.
With touchscreen voting, the technology that secures the election is not in plain sight, but instead is hidden from voters inside voting machines, the workings of which are beyond the comprehension of most people. The software that runs voting systems is proprietary and not open to public inspection.
While many election officials are hesitant to purchase touchscreens, California's Riverside, Alameda and Plumas counties already have. Others may have no alternative. Earlier this year a federal judge ruled that the nine California counties using the "Votomatic"-style punch card voting systems must replace them by March 2004.
There aren't many choices for the Votomatic counties. While paper-based optical scan and the "Datavote" punch card systems work well in many counties, they are not considered practical to administer in larger counties like Los Angeles. Several Votomatic counties have been trying out touchscreens in early voting experiments; only one, Sacramento, has used a touchscreen system that generates a voter-verified paper trail.
Proposition 41, passed by California voters last March, provides $200 million in state funds to help pay for new equipment. Though the measure requires counties to maintain a paper record of ballots cast on touchscreens, it does not require that record to be generated at the time the voter votes to insure that printed records reflect voter intent as opposed to, say, a computer malfunction.
Counties that purchase touchscreen machines, which typically cost $3,000 apiece, would have to spend millions to deploy several in each polling place. Just getting these machines to polling places is likely to be a huge logistical feat; many models are heavy and cannot be transported by pollworkers.
In Florida touchscreens were deployed for the first time in several counties for the state's September primary election. In dozens of Miami Dade and Broward county precincts, the number of votes recorded on computerized machines was far below the county and statewide averages. After the election, officials in those counties "found" hundreds of ballots stored in the machines that had not been collected on election night. Even if the election workers succeeded in "harvesting" lost ballots out of those machines, it's unlikely voters will feel confident that the final count was accurate.
In Dallas County, Texas, some voters who cast electronic ballots found that though they had selected one candidate's name, a different candidate's name appeared on the screen. Voters in some Florida and Georgia counties reported similar problems, which were attributed as "calibration errors" by election officials, possibly caused when machines were transported to polling places.
While some have characterized these problems as "minor glitches", I believe they are only the beginning of new voting problems that are in store for us if we convert to paperless, computerized voting.
First-time touchscreen voters often report that they like the machines and find them easy to use. But it's not enough to ask if people like the machines -- the bigger question is, do they trust the machines? Do they have confidence that their votes will be counted? Will they trust the results in a recount if conducting one is simply a matter of pressing a button?
I think not. In San Bernardino County's November 2001 local elections, the vote tabulation software used to count punch card ballots malfunctioned. The software error was not discovered until after the election results were released. Once discovered and corrected, the ballots were tabulated a second time, and the outcome of 13 contests changed; candidates who were told they won had in fact lost.
However, San Bernardino's voters were not content to accept the computer-counted results, and insisted the county conduct a hand recount of the punch card ballots, which it did. If there had been no paper ballots to count, I doubt the voters who supported the losing candidates in San Bernardino would have accepted the revised results. They might have instead filed a lawsuit, or insisted on a revote, or maybe they just wouldn't show up to vote the next time around.
It strikes me as reckless to experiment with paperless voting transactions in the one transaction that is most integral to living in a free society. Some suggest voting on touchscreens is like using an ATM, but few people would use ATMs if the content of their transaction was secret from their bank and there was no paper trail verifying the transaction. Yet this is exactly what we are asking voters to do with paperless touchscreen voting systems.
The vendors and election officials who deploy touchscreen machines seem to be asking the voters to "trust us", but many people will not trust something they cannot see or understand. Citizens in a democracy should be expected to exercise healthy skepticism about government, not blind faith, especially when it comes to elections, which we know from experience are subject to glitches and manipulation.
Computerized voting is a 21st century solution trying to fit into a 19th century system. If we are to truly modernize our voting process, it's going to take more than just switching out one machine for another.
Ours is a quaint, decentralized and highly inefficient voting process. Each election it's a struggle to locate and set up thousands of polling places across the state and recruit enough pollworkers to staff them. Meanwhile, a growing percentage of Californians opt to vote by mail and overall turnout declines.
It doesn't have to be this way. We could switch from precinct-level voting to county-wide voting and set up temporary, high-tech voting centers in numerous public places such as post offices, libraries, shopping malls and community centers. These voting centers can be highly-visible, staffed by professionally-trained public employees, with pollworkers serving as volunteers. Voters could cast ballots on computers that also print a paper version of the ballot that a voter can verify before it goes into a locked box to be counted alongside the digital ballots. The centers can be networked with the county election office to ensure the database of registered voters who have cast ballots is routinely updated to prevent fraud.
It is this kind of wholesale change that's needed before we usher computers into the voting process. Meanwhile, the clock is ticking and California's Votomatic counties will have to decide soon what to use for the March 2004 primary. Now is the time for California voters to speak up and ask questions, before counties spend millions of dollars on new equipment.
Kim Alexander is president and founder of the California Voter Foundation, a Davis-based nonprofit, nonpartisan organization advancing new technologies to improve democracy, online at www.calvoter.org.
This page was first published on December 21, 2002 | Last updated on December 21, 2002
Copyright 2002, California Voter Foundation. All rights reserved.