Is Electronic Voting the Issue?
There's been quite a ruckus lately over flaws in e-voting machines. Much of this is the result of several embarrassing hacks into the public servers run by e-voting machine vendors. This has led many people, including many high profile researchers and professionals in the IT industry, to renounce electronic voting in its entirety. However, many others believe we may be throwing the baby out with the bath water.
Direct recording electronic (DRE) voting machines are generally free standing computers with touch screens, similar to ATMs. They record the vote directly to the computer's internal hard drive and do not create a paper trail, which prevents many types of scams (e.g. chain voting). However, it also means that the machine's tallies cannot be double-checked by a human or even another machine.
As such, we are reliant on the manufactures of DRE machines to produce reliable and accurate machines. Diebold Election Systems is one such vendor. When a Diebold server was hacked in March, the source code for their system was leaked out onto the Internet and fell under the eye of public scrutiny.
A group of researchers from Johns Hopkins University and Rice University published an analysis (Adobe PDF format) of the source code. In short, they found bugs and design flaws in the system which could make it susceptible to tampering. In at least one instance, it was discovered that a voter could cast unlimited votes without being detected by the system.
Amongst the other information retrieved from the Diebold site was an election data file. The time stamp on the file revealed that it was copied at 3:31 pm, during election hours. Since it is illegal to tally votes before the polls have been closed, this raises questions as to the file's origin and purpose.
Black Box Voting is a watch dog organization tracking problems with electronic voting machines and the vendors who supply them. They publicized information about the election data file, along with a long laundry list of other DRE inaccuracies and errors.
Despite all of this, there are many individuals in the industry who believe that these failures are the fault of the DRE vendors, not of DRE itself. In an interview with in the MIT Technology Review, Ted Selker, a MIT Media Lab professor and former IBM Researcher, highlights many of the benefits of DRE technology:
- Larger fonts and sharp contrasting colors will be welcome to those who have difficulty reading finely printed ballots
- Confirmation screens to ensure votes are casting for the correct candidate would have mitigated the "Buchanan effect" in the 2000 presidential election
- Systems can prevent users from accidentally voting for more than one candidate and other mistakes that can disqualify votes
- Machines can be configured with support for many languages without incurring the increased ballot printing expenses
- Voice enabled machines can provide truly private voting for many voters
- Photographs of candidates can appear on the screen to aid voters who may be illiterate
The problem, says Selker and organizations such as the Electronic Frontier Foundation (EFF), is not the idea: it's the implementation. Though many disagree over whether or not the machines should produce a paper trail, most proponents of electronic voting are opponents to a closed system, one made by private companies using proprietary technology and not open to public review.
With machines that can better communicate choices to voters and provide better privacy and instant feedback, it would seem that, individually, electronic voting machines have the potential to be less fail able than the ballot and counting systems currently in place. The real fear is that electronic voting systems connected to a network can be co-opted on an unprecedented scale.
This is why, according to the EFF, the only hope of reliable and trustworthy elections is through public review of voting systems. Selker argues that electronic voting machines should be designed in a public process, and the manufacture of the machines should be put out to bid. He notes that, overall, this process has worked well for countries such as Brazil.
To many, news coverage of the 2000 presidential election looked like something out of a banana republic. Perhaps in fear of a an event even worse, many would throw out electronic voting even though it could have prevented or mitigated many of the problems. However, as even many of the proponents of electronic voting note, before looking toward technology as a cure, we may need to study the lessons learned and examples set by other countries.