BlogTV: Electronic Voting Machines
BlogTV presents a primer on electronic voting machines, and the voting theory. Recently CNN presented a demonstration of several electronic voting machines that will be used in the 2002 elections. Unfortunately, the demonstration merely shows how the machines are used, and does not adequately explore the problems inherent in these voting systems.
Before we go further, please read this primer on paper ballot systems written by one of the foremost authorities on paper and electronic voting systems, Doug Jones, a computer science professor at the University of Iowa. Jones' series of papers on voting are a revelation. They treat voting like a problem in logic and mathematics, yet are written in plain language that anyone can understand. Jones does an excellent job of explaining an ideal voting system, so I will not belabor those points. Jones also does an excellent job of explaining how the voting system is designed to deter fraud. Let us examine how flaws in these electronic voting systems might be used to commit fraud and overthrow an election.
The most fundamental issue in American voting is the secret ballot. Secrecy must be absolute, nobody is permitted to know which ballot was cast by any individual voter. To insure secrecy, all ballots must be indistinguishable from any other, and must not contain any marks or numbers that might identify any specific ballot to any specific voter. Votes must be deposited in the ballot box in plain view of the voting public. Ballots should not be deposited in the ballot box in the sequential order, if they are deposited sequentially, the ballots should be jumbled together and randomized before counting. Even the voter must not be able to identify his own ballot, so he cannot prove he voted any specific way, so he cannot prove he cast his vote for any specific candidate, this prevents a voter from proving he voted a specific way, to someone seeking to buy votes. Votes must be randomized before counting, or else someone could match a list of incoming voters to the order ballots were deposited.
None of the new electronic voting systems meet these strict criterion, allowing numerous opportunities for fraudulent manipulation of the voting system. The systems were designed primarily to prevent overvoting and undervoting, and not with the goal of preserving security of the secret ballot, nor with the intent of preserving the overall integrity of the voting process.
Note that the voting systems all use an ID number for each voter, and each electronic ballot is tagged with that ID number. Normally, with paper ballots, the voter ID is written on the ballot stub which is detached before depositing the ballot, which removes any identifiable data that can track the ballot to the voter. With electronic voting, each registered voter is assigned an ID, either manually entered into the machine, or contained on a smart-card. There is no way to assure that ballot IDs are not stored with the ballot data. In practice, the ID must be stored with the ballot data. This completely strips away the privacy of the secret ballot. This opens up the system to vote buying. When secrecy is not maintained, it is easy for someone to prove they voted for a specific candidate on a specific ballot, and claim their reward for casting the vote. It also allows partisans inside the system to identify opposition voters and spoil their ballots.
Another violation of our ideal voting system involves the public deposit of votes in the ballot box, to prevent vote stuffing. The electronic ballot box is an incomprehensible device connected to the voting booth with wires. Nobody can see the votes being deposited. There is nothing to prevent an unscrupulous election official from secretly handing someone a dozen smart cards or ID numbers, in the privacy of the shielded voting booth they could vote many times without detection.
There are many other flaws in these electronic systems that are not immediately evident. Jones' paper also focuses on counting the votes, another opportunity for fraud. We cannot know how these electronic vote counting systems work, the internal mechanisms are top secret, and not open to public scrutiny. This is derisively referred to in the computer programming world as "security through obscurity." To prevent fraud, transparency of the system is essential, yet the voting mechanism is absolutely opaque. You have no idea if a rogue programmer has inserted code that discards every 10th vote for a certain candidate. You have no way to verify your vote was recorded, there is no physical record of your vote inside an electronic system. Votes can be manipulated instantly, secretly, and en masse.
This essay cannot possibly enumerate all the flaws in these systems, as the specific details of the electronic systems are not public information. This is the most astonishing fact of all. We have handed the voting system over to computer programmers working in secret, one programming error will affect all voting machines. In a paper ballot system, one unscrupulous official may spoil a precinct's votes, but not the whole system's votes. This is the biggest fraud of all.
My major intent in publishing this essay is to acquaint the public with the fine work of Professor Jones. His essay on paper ballot systems should be read by any voter who seeks assurance that his voting system is properly administered, whether it be on paper or electronic. Please examine his paper at length, and if anyone can suggest other potential flaws in the electronic systems in the video, please leave a comment.
|
|
Can't see BlogTV?
Click Here. |
Before we go further, please read this primer on paper ballot systems written by one of the foremost authorities on paper and electronic voting systems, Doug Jones, a computer science professor at the University of Iowa. Jones' series of papers on voting are a revelation. They treat voting like a problem in logic and mathematics, yet are written in plain language that anyone can understand. Jones does an excellent job of explaining an ideal voting system, so I will not belabor those points. Jones also does an excellent job of explaining how the voting system is designed to deter fraud. Let us examine how flaws in these electronic voting systems might be used to commit fraud and overthrow an election.
The most fundamental issue in American voting is the secret ballot. Secrecy must be absolute, nobody is permitted to know which ballot was cast by any individual voter. To insure secrecy, all ballots must be indistinguishable from any other, and must not contain any marks or numbers that might identify any specific ballot to any specific voter. Votes must be deposited in the ballot box in plain view of the voting public. Ballots should not be deposited in the ballot box in the sequential order, if they are deposited sequentially, the ballots should be jumbled together and randomized before counting. Even the voter must not be able to identify his own ballot, so he cannot prove he voted any specific way, so he cannot prove he cast his vote for any specific candidate, this prevents a voter from proving he voted a specific way, to someone seeking to buy votes. Votes must be randomized before counting, or else someone could match a list of incoming voters to the order ballots were deposited.
None of the new electronic voting systems meet these strict criterion, allowing numerous opportunities for fraudulent manipulation of the voting system. The systems were designed primarily to prevent overvoting and undervoting, and not with the goal of preserving security of the secret ballot, nor with the intent of preserving the overall integrity of the voting process.
Note that the voting systems all use an ID number for each voter, and each electronic ballot is tagged with that ID number. Normally, with paper ballots, the voter ID is written on the ballot stub which is detached before depositing the ballot, which removes any identifiable data that can track the ballot to the voter. With electronic voting, each registered voter is assigned an ID, either manually entered into the machine, or contained on a smart-card. There is no way to assure that ballot IDs are not stored with the ballot data. In practice, the ID must be stored with the ballot data. This completely strips away the privacy of the secret ballot. This opens up the system to vote buying. When secrecy is not maintained, it is easy for someone to prove they voted for a specific candidate on a specific ballot, and claim their reward for casting the vote. It also allows partisans inside the system to identify opposition voters and spoil their ballots.
Another violation of our ideal voting system involves the public deposit of votes in the ballot box, to prevent vote stuffing. The electronic ballot box is an incomprehensible device connected to the voting booth with wires. Nobody can see the votes being deposited. There is nothing to prevent an unscrupulous election official from secretly handing someone a dozen smart cards or ID numbers, in the privacy of the shielded voting booth they could vote many times without detection.
There are many other flaws in these electronic systems that are not immediately evident. Jones' paper also focuses on counting the votes, another opportunity for fraud. We cannot know how these electronic vote counting systems work, the internal mechanisms are top secret, and not open to public scrutiny. This is derisively referred to in the computer programming world as "security through obscurity." To prevent fraud, transparency of the system is essential, yet the voting mechanism is absolutely opaque. You have no idea if a rogue programmer has inserted code that discards every 10th vote for a certain candidate. You have no way to verify your vote was recorded, there is no physical record of your vote inside an electronic system. Votes can be manipulated instantly, secretly, and en masse.
This essay cannot possibly enumerate all the flaws in these systems, as the specific details of the electronic systems are not public information. This is the most astonishing fact of all. We have handed the voting system over to computer programmers working in secret, one programming error will affect all voting machines. In a paper ballot system, one unscrupulous official may spoil a precinct's votes, but not the whole system's votes. This is the biggest fraud of all.
My major intent in publishing this essay is to acquaint the public with the fine work of Professor Jones. His essay on paper ballot systems should be read by any voter who seeks assurance that his voting system is properly administered, whether it be on paper or electronic. Please examine his paper at length, and if anyone can suggest other potential flaws in the electronic systems in the video, please leave a comment.