Skip to Content
Uncategorized

E-Voting Deadline Looms

Election officials must decide by January 1 which electronic voting system is least likely to cause problems next November.
December 13, 2005

The mid-term elections of 2006 are still 11 months away, but they’ve already generated a controversy – not over politics, but about what technology should be used to count all the votes.

After the voting fiascoes during the 2000 presidential election, Congress passed the Help America Vote Act, which offered a total of $3.2 billion to the states to improve their vote-counting processes, most notably to replace the punch card systems that were exposed as so flawed during the recounting in Florida. States that took the money must use replacement systems in all elections after January 1, 2006. Yet widespread rancor persists over which new system is best.

The trend away from punch cards and other mechanical voting systems and toward electronic voting was underway even before the passage of the Help America Vote Act. And since 2000, the number of machines that record votes directly into memory without printing a paper record, sometimes called “direct-recording” machines, has more than doubled, largely replacing punch-card systems, according to the U.S. Election Assistance Commission, an independent federal agency set up by the act to improve the election process.

It sounds like an obvious way to improve the democratic process. Yet the ascendance of direct-recording voting machines hasn’t brought accolades from everyone. “They’re not auditable,” says Rebecca Mercuri, a software consultant in Mercer County, NJ, whose PhD dissertation at the University of Pennsylvania was on electronic vote tabulation systems. She argues that there’s no way for voters to know whether the machines have faithfully recorded their votes. And if election monitors don’t trust the tally that the machine prints out at the end of the day, they can’t do a manual recount – just print it out again. “That’s a reprint, not a recount,” Mercuri points out.

This well-publicized problem with e-voting machines has led to more than just complaints from experts. “Without the capacity for meaningful audits, [direct recording machines] are unfit for use in actual elections,” says Matt Zimmerman, a staff attorney at the Electronic Frontier Foundation in San Francisco. On December 8, the EFF brought suit against both the Board of Elections and Office of Information Technology in North Carolina for certifying electronic voting machines from three vendors without examining the machines’ source code, as required by the state’s election law.

This fall, the U.S. Government Accountability Office summarized its own investigation of electronic voting, in a 107-page report with the rambling, but revealing title “Federal Efforts to Improve Security and Reliability of Electronic Voting Systems Are Under Way, But Key Activities Need to Be Completed.” Existing federal election standards are vague and not even mandatory, the GAO says. And, unfortunately, the agency added, any new standards issued by the Election Assistance Commission will not come in time to help with the 2006 election.

Without naming names, the GAO report cites instances in previous state elections in which the tallies from direct recording machines were open to tampering, the ballots themselves could be modified, locks were easy to pick, and power switches were exposed and unprotected. There were also breakdowns and cases of election officials who could not configure the machines. (Scores of election gaffs caused by direct recording machines are documented at www.votersunite.org, the website of a group of activists promoting fair elections.)

Mercuri and others also worry that direct-recording machines are vulnerable to human error. For example, the order of names on ballots is commonly rotated from ward to ward so that no candidate always has the coveted top slot. If a local programmer gets the names out of synch, though, suggests Mercuri, there might be no way to tell which votes were really cast for whom.

Mercuri would rather see jurisdictions use old-fashioned optical scanners instead of moving to direct-recording machines. In optical scanner systems, voters mark paper ballots with ink, then feed them into scanners. The paper ballots are collected in a bin inside the scanner, and can be hand-counted for an audit. In fact, optical scanner balloting is the most popular election technology in the United States, accounting for over 30 percent of the votes cast in 2004 (see table).

Rather than discarding the direct-recording machines they’ve bought, however, many jurisdictions have modified the tally printers inside the machines, which are normally used only after the polls have closed, to produce a “voter verifiable” paper audit trail. A person receives a ticket showing what votes he or she has just cast – at least what the machine has registered. The ticket, which contains no identifying information about the voter, remains in the polling place for auditing purposes.

Of the 50 U.S. states, 30 have laws requiring that voting machines produce a voter-verifiable audit trail, or they have already converted to a system with a voter-verifiable paper trail through their purchasing decisions, says Pam Smith, nationwide coordinator for the nonprofit Verified Voting Foundation.

Others claim that these concerns about e-voting machines are blown out of proportion. According to Michael Kerr, director of the Election Technology Council in Arlington, VA, every maker of direct-recording machines provides some way to audit the functioning of its machines, such as an audit log that can be checked against the contents of working memory. “The 2004 elections were for the most part problem-free, except for some isolated instances of technology process training glitches,” he says. The council represents makers of direct-recording machines such as Advanced Voting Solutions, Diebold Election Systems, and Election Systems & Software.

A main reason for switching to direct-recording machines, Kerr says, was to avoid the expense of paper and ink, especially where ballots had to be supplied in many different languages. Changing the machines to produce a voter-verifiable paper trail re-introduces that expense, as well as costing $1,500-3,000 per machine to do the upgrade, he says.

The Help America Vote Act required that direct-recording machines generate a paper record suitable for a manual recount – but not necessarily a voter-verifiable paper trail. Since the paper-trail debate arose after the law was passed, the federal government does not have a policy for or against them, according to Gracia Hillman, chair of the Election Assistance Commission.

Many jurisdictions may adopt direct-recording machines simply because it is the most expedient way to comply with the accessibility provision of the voting act, Hillman says. The act required that every voting place have a voting machine accessible by the disabled, and appropriated $100 million for accessible machines, but did not specify what kind of equipment should be purchased.

The election commission has already ruled that punch cards and lever machines do not meet the disabled-access requirement. But beyond that, no election technology is either required or outlawed. “In today’s market the [direct-recording devices] provide such access, although there are also a couple of electronic systems for marking optical scanner cards,” says Hillman. “The key is for blind voters to get audio instruction and confirmation through voice feedback.”

Hillman expects that most jurisdictions will hold their elections with mixed systems, each offering a technology-of-choice for most voters, as well as one accessible direct-recording machine for the disabled at each polling place.

As for the standards called for in the GAO report, the Election Assistance Commission may issue them as early as this week, says Hillman. “But the effective date will not be in 2006, so as not to disrupt the process for jurisdictions currently purchasing machines for the next election,” she says. “There is just not enough time for national certification between now and the 2006 election.”

Types of Voting Equipment in 2004 in U.S. General Election

Equipment Type

Percentage of Voters Using It

Optical Scan

39.8

Electronic

25.0

Lever

12.4

Punch Cards

9.0

Mixed Systems

7.7

Paper Ballots

1.8

Unknown

4.4

Source: Election Data Services

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.