Return to Start
Computerized tabulation is here to stay. Lever machines and paper ballots are being replaced by increasingly complex computerized vote-counting systems. We are "progressing" towards automation, but first the implications of automation must be considered.
The past two decades have seen an explosion in the use of computerized voting systems. About two-thirds of votes cast in this years election will be count by computer. [Harrar] New technology has made elections faster, sleeker, more efficient: "No more pieces of paper. No more waiting until the early morning hours for weary election workers to finish counting paper ballots, only to find later that the count was off by 2% or 3% or more." When computerized vote-tallying equipment was first introduced general enthusiasm was high.
Today, some people are wondering if we haven't gone too far. The president of Election Data Services, Kimball W. Brace, says, "the election community is in a state of potential crisis. We're waiting for a volcano to erupt, in the form of a major election scandal. . . . We know it's going to happen, but we don't know when or how we're going to handle it." According to Brace, in the rush for state of the art technology not enough attention has been paid to the security risks that automation and centralization create. "Growth" of computerized elections cannot be considered "progress" until we are sure that the drawbacks do not outweigh the benefits. [Trombley, Part 1]
The biggest drawback is the possibility of a "major election scandal." When tasks were decentralized and counted by hand this possibility did not exist. According to a state official, major setbacks and embarrassment will only be avoided if all efforts are directed towards reforming the existing system. Like Brace, he is not optimistic that the necessary precautions will be taken: "unfortunately, it often takes a disaster before election offices are upgraded." [Trombley, Part 1]
Among computer security experts there seems to be some general consensus about minimal reforms that are needed. First, proper instruction for the people who operate the counting machines must be assured. "With some notable exceptions, they don't understand the technology. There is not enough money or training to enable our election officials to conduct elections correctly in the computer age. They are not up to speed." [Trombley, Part 2] Today's media tends to focus on scandals that involve intentional computer fraud. Much more common than fraud are failures that can be attributed to the large staff of overworked, poorly paid officials who try to run local elections without the necessary computer knowledge. Whether it is human error or computer error, the end result is the same--votes are invalidated or completely lost.
Until better training is established, election workers will be forced to rely on the providers of the vote-counting systems to assure the validity of the increasingly complicated equipment. The only way for them to obtain a recount will be to ask the computer for another copy of the same results. Education is a crucial step to avoiding dependence on equipment providers.
The next step is to assure that equipment providers adhere to some standard of performance and do not abuse their power. A voluntary standard written by the Federal Election Commission (FEC)'s National Clearinghouse on Election Administration has been approved for use by state and local election officials.
The FEC standard has three objectives: to provide certification of computer systems by independent testing authorities, to require audit trails, and to mandate the placement of vote-counting software in escrow so it can be checked in elections where results are contested. Unfortunately, many computer experts have argued that this standard has no practical future. A more technical, tougher standard is needed to assure the security and integrity of electronic elections.
Howard Jay Strauss, a Princeton University computer scientist and member of Election Watch says, "The result is a gossamer standard, so thin and flimsy that it will do little to improve the security and integrity of electronic elections." [Betts] Election Watch is a Los Angeles-based watchdog group of critics.
Election Watch arose from unease about elections that have gone awry, and a belief that better standards (both state and federal) are needed to assure fair elections. As long as citizens feel uncomfortable with the election process, groups like Election Watch will exist, and will pressure for the government for improvements.
The new FEC regulations have not eliminated the need for such groups. Many critics exist. They believe that the recommendations made by the FEC must be expanded to include rules for software testing, high-level programming languages and better computer security. "It's a good start, but I'm disappointed with the standard in several areas," [Betts] said Roy J. Saltman, a computer scientist at the National Institute of Standards and Technology and the federal government's top expert on voting systems.
As long as people do not feel they can trust their own government to count their votes, new legislation will be demanded and groups such as Election Watch and the Computer Professionals for Social Responsibility will continue to play a pivotal role. Perhaps we are advancing too fast. "An electronic voting system is to a mechanical one what a nuclear bomb is to a conventional one: if someone manages to sabotage it, the results can be catastrophic." [Dugger] Computer scientists today are the physicists of yesterday. We must stop to think about the consequences of our actions before we go ahead and create dangerous technology simply because it can be created.
"Since computers can sometimes be vulnerable to intrusion, there are safe-guards in place in state elections law to minimize the possibility of tampering with the vote tallying process." [Bush] Secretary of State Bill Jones is correct to advocate state prevention of election fraud. The problem is that with words like "minimize" it is no wonder that some citizens remain skeptical. When dealing with national elections, minimized risk is not enough. Until complete security against fraud is established there should be no rush for change.
It's true that one mouse click from a computer over the Internet is much easier than a trip to the voting booths in the high school gymnasium. It's also true that it is already possible to cast votes on the World Wide Web: users can vote for their favorite movie, favorite soft drink, or even their favorite college mascot.
In the future it might be possible for people to never leave their homes on a trek to the high school gymnasium. Unfortunately, this might damage society more than it helps. Cranor suggests that Intenet voting will help committees reach a consensus faster: "The larger your committee gets, the harder it is for all the committee members to negociate together, and so [on-line voting] would allow the negotiation process to sort of take place automatically." Committees are often blamed for the large amount of discussion that has to take place before any decision is reached. Although this can be inefficient, it also serves a purpose. According to the ACM committees could reach decisions by simply having every committee member rank their choices and preferences on a certain issue. If this happens, all of the arguement, negociation, and valuable feedback will be lost.
In conclusion, some people believe that the nation is moving from "representative government" to a system of "participatory democracy." Representation will vanish and citizens will vote directly on all issues. [Boyarsky]. Votelink asks, "Should NATO launch airstrikes on the Bosnian Serbs?" When the issue is the Bosnian Serbs and not the Stanford Tree sometimes I would rather rely on a slow committee of educated representatives. Society relies on representation not becuase voting would be too slow, but educating citizens on every issue to the degree needed for informed decision-making is too slow.
We should not rush to make changes just because they are possible. Slow, gradual progress is the best way to assure that our accomplishments do not backfire.