From the editor: This article by Kurt Hyde was originally published in the October 13, 1986 issue of The New American. It is one example of how The New American and its parent organization, The John Birch Society, have been sounding the alarm on election integrity for decades, since long before concerns about the vulnerabilities of our voting system went mainstream.
Boston University was the site of an historic symposium on August 14th and 15th [1986]. Computer experts, attorneys, political campaigners, and election specialists from many parts of the United States gathered for two days to discuss the security and reliability of the computers used in American elections. While the emphasis was on computerized vote counting, the final afternoon featured a panel discussion aimed at keeping elections clean whether computers are involved or not.
One of the most startling revelations made was that computerized vote counting is not the fastest way to count votes. Mechanical lever machines are still the fastest! When the election is over, one has merely to open the machine and record the totals. The city of Manchester, New Hampshire implemented computerized voting booths on September 9th, less than a month after the symposium. After the polls closed, there were numerous complaints made by political veterans and moderators about the amount of time required to get the final totals. According to one moderator in Manchester’s Ward Twelve, the new equipment took over two hours to produce the final results, roughly three times as long as the old lever machines.
Mechanical lever machines are still the safest from tampering too. While it is possible to tamper with a mechanical lever machine, the tampering generally requires the use of a physical object, leaving a piece of physical evidence. Also, their public and protective counters can be used to detect attempts to rig an election. One brand of mechanical lever machine even uses a large carbon paper sheet to record initial and final vote totals on election day. However, mechanical lever machines do not have recount capability. Defects and corrosion sometimes cause the counters to stick. For this reason, mechanical lever machines should get regular preventive maintenance.
Computerized Voting
Most computerized vote counting equipment can be placed into one of two categories: Punch Card Systems or Direct Voting Systems. Punch Card Systems use computer punch cards as ballots. These ballots are counted after the polls close. Late in the day, during an election in a New Hampshire town, poll workers discovered that someone had taped over a candidate’s name in the election book and replaced it with someone else’s name. Punch Card Systems generally rely on a book to show the voter which hole to punch, and the punch card itself doesn’t have the candidate’s name. This means that the voters who had used that book had no way of knowing the name of the candidate they were voting for. Also, punch card ballots must sometimes be transported to counting facilities. This presents an opportunity for tampering and contributes to the centralization of power.
Direct Voting Systems
Direct Voting Systems are completely computerized voting booths. Because they generally do not create a ballot at all, there are no individual ballots to be recounted — and no physical trail that can be traced. The secrecy of internal procedures poses an even greater potential security risk in Direct Voting Systems. One manufacturer of computerized voting booths claims that only three people have complete knowledge of its internal operations. It is certainly possible that only three people have been authorized to know how it works, but there are many other employees who have been authorized to have partial knowledge. It is also possible that a curious employee has done some scavenging late at night while supposedly working overtime.
One speaker at the symposium noted that the Soviets are busy gathering American technology, and that organized crime is now using computers. The burning question is: Do subversive or criminal elements have the capability to tamper with a computerized voting booth? Secrecy-based security backfires in this case. How can anyone ensure that a computerized voting booth has not been tampered with if nobody is allowed to know how it is supposed to work?
The State of Illinois conducted controlled testing of its voting equipment and documented a number of errors. The most spectacular was in one county where errors were uncovered in 47 percent of the precincts. One New Hampshire town recounted the ballots after an election just to test their equipment. After three counts, they had three different totals. Fortunately, the differences were small in this case, about five votes. However, a similar experiment in Indiana had more dramatic results. A group of Computer Science students discovered that they could change the operator control cards and manage the election results. They successfully altered the results of a recount merely by changing the operator instructions.
One speaker at the symposium tried for 14 months to document an election in Dallas, Texas. The computer printouts had numerous discrepancies, and one precinct was never found. Votes from this precinct were included in the city-wide totals, but the precinct itself was never found. No moderator was identified, no voter list was found, nor was there any record of the address where that precinct held its election that day.
States generally establish very restrictive maximum time limits, some as low as 72 hours, for filing election complaints. Such time limits may make it impossible to file a well-documented complaint. The solution is clearly not in extending reasonable time limits, because that would allow a primary loser to harass the winner in the general election. The solution is in clearly defining what information must be reported and to whom, and in storing it in a way that makes it available to the public.
Paper Ballots
Paper ballots are still used in some parts of the country, especially for absentee balloting. One of the most common problems in paper ballots is the overvote, caused when a single voter casts a ballot for more than one candidate for the same office. This happens frequently when a voter votes a straight ticket for a party and then decides to vote for one candidate from another party. In some states, either the straight ticket or the individual vote may pre-empt the other, but in others it is considered an overvote and neither candidate is credited. The security problem occurs when vote counters with pencil lead deliberately cause overvotes. Fortunately, the abnormal rate of overvotes in a given contest provides some evidence of fraud.
A second problem with paper ballots is “chain voting”: A voter pretends to cast a ballot, but instead brings the ballot outside where it is marked indelibly. Another voter is then offered money for casting that ballot and returning his unused ballot, which is used to continue the chain.
Recommendations proposed during the symposium include the following:
- Punch card ballots should be redesigned to contain the names of the candidates by the appropriate punches. This would permit hand recounts and allow voters to read and understand their punch card ballots.
- There should be public domain computer programs for counting votes. This would do away with the current practice of using secrecy to protect manufacturers from copyright infringement.
- The vote counting process should be open to the public. To maintain public confidence in the electoral process, the public must be satisfied that votes are being counted correctly.
- Computerized voting booths should print a paper ballot for each ballot cast. This ballot can be passed under a transparent cover (to preclude stuffing the ballot box) for the voter to confirm that the computer has voted as instructed. In case of recount, the paper ballot would be the official ballot.
- All elections should be accountable at the precinct level. Ballots should not have to be transported for counting.
