Punch Card Voting
Computerized voting systems may do what the combined armies of all the earth could not do— that is, disenfranchise American voters and turn American elections over to machines and their operators.
The computer punch card systems are the most dangerous of all. There are four major punch card systems: Votomatic, Vote Recorder, Vote-A-Corder, and Vot-A-Maker. IBM holds basic patents for the equipment in all four systems. The Votomatic is distributed by IBM itself. The other three devices are distributed by other companies under IBM license. All four systems use patented IBM computer punch cards as ballots. Essentially, all four systems are alike.
The voter inserts the IBM punch card ballot into the machine beneath a booklet containing the list of candidates. The voter marks his ballot (which he cannot see at the time of marking) by punching a sharp stylus through fixed holes in the booklet, one hole being beside the name of each candidate. Each hole is supposed to fit perfectly above a small slit (chad, in computer language) on the punch card. The card is scored (partially cut through) so that each chad should punch out cleanly when the stylus is pressed against it.
When the voter has finished, his punch card ballot is perforated with slits—one slit for each candidate or party voted for. The location of the slits on the card tells the computer what the voter’s selections were.
If the voter can read a name and punch a stylus through a hole beside the name, he perceives no difficulty in using the punch card machine. If he can neither read nor punch, he can get a polling official to do both for him.
But speed and ease in using the punch card voting machine do not assure that ballot cards will be punched correctly, that votes will be counted accurately, or that they will be counted at all. In fact, the possibilities of error and fraud are enormous.
If candidates are not listed in the programmed order in the booklet—or if pages of the booklet are not assembled in the exact sequence programmed for the punch card—the voter’s selections will not be correctly recorded on the ballot card.
If the voting machines are not perfectly assembled, they can cause all ballots to be punched wrong, to be damaged so that the computer cannot count them, to be invalidated, or to register no vote at all. Even if voting machines are perfectly assembled, they are so fragile that they can be knocked out of alignment by rough handling or jolting while en route to the polling places.
There is no way for the punch card voting machine to keep a voter from accidentally voiding or distorting his own ballot—by “overvoting” (voting for more than one candidate in the same race), or by voting a straight political party ticket and then crossing over to vote for individual candidates of an opposing party.
The punch cards, mass-produced on cheap paper stock, have many imperfections that often invalidate votes, or foul up computer counting, even if the punch card machines and everything else in the system (including voters) are working perfectly.
Here are examples of punch-card imperfections which have marred elections in various parts of the country:
(1) Over-inked cards have gummed up computers.
(2) Cards have been scored (partially cut through) so deeply that chads were accidentally jolted out, thus voiding ballots or registering votes the voters did not intend.
(3) Cards have been scored so lightly that the chads did not punch out cleanly, but were left hanging to the card. Poll workers had to handle every ballot card and tear off all dangling shreds of paper before the cards could be fed into a computer. A worker handling a punch card to tear off paper shreds can easily (with little probability of being detected) punch out other chads on the ballot, thus invalidating it or recording votes the voter did not intend.
(4). Cards have been made of paper stock that absorbed too much moisture from the atmosphere, the results being that the ballots would not fit into the voting machine properly or go through the computer without jamming.
(5) In the mass-production process, punch cards are pressed tightly; and many of them stick together so closely that two or more cards look like one, and go into the voting machine like one. In such cases, the voter punch-votes two or more ballots, innocently thinking he punched only one.
Even if the punch card ballots are perfect, voters frequently mar them, without ever realizing it:
(1) Nervous or impatient voters, waiting in line to vote, with punch card ballots in hand, have bent the cards, perspired on them, stuck corners of them in their mouths—the results being that the cards would not fit into voting machines properly or go through computers.
(2) While waiting in line, shifting their punch card ballots from hand to hand, voters have accidentally popped out chads, thus unknowingly voiding their own ballots or casting votes they did not intend to cast.
(3) Voters have unknowingly voted wrong (or not at all), because they held the punching stylus at an incorrect angle, or used a ballpoint pen, pencil, or something else besides the stylus.
Voters, for the most part, are not aware of the many imperfections and resulting malfunctions that steal or distort their votes; but thousands have strongly complained about obvious dangers and inadequacies in punch card voting.
The punch card system often eliminates privacy of voting. In a large precinct with heavy voter turnout, punch card voting is about as private as dining at a crowded, stand-up lunch counter.
The secret ballot—absolutely essential to a free, self-governing people—is seriously endangered by punch card voting. In Dallas, for example (where the Vote-A-Corder was used in 112 precincts for the 1968 general elections), the poll card ballot was numbered; and a stub attached to it bore the same number. The voter had to sign the stub in the presence of an official. After voting, he detached the signed stub and put it in a locked box, placing his ballot in a ballot box.
The locked stub boxes can be opened only on court orders—which, generally, are issued in cases of lawsuits or demand for a recount. Such cases usually involve a local race. Feeling runs high. The community is often sharply divided. Partisan emotions are intense. The stub box is opened; and the political leaders of your community—of your neighborhood—can examine your ballot, knowing it is yours.
In punch card voting, there is no feasible way to guarantee the physical security of ballots.
In Dallas this year, ballots were taken out of ballot boxes at the polls and counted by hand. They were also handled and examined for shreds of paper improperly clinging to them, or for obvious mutilations. After they were carried downtown to the data processing center, they were stacked by hand for feeding into a mechanical counter. Then they were stacked by hand for feeding into the computer. After being computerized, they were again stacked by hand for another run through the mechanical counter.
It takes little imagination and not too much cynicism to visualize the possibilities of deliberate (or accidental) invalidating of ballots or adding of wrong votes, as the fragile punch cards pass (often under chaotic conditions) through so many hands in so many places.
In addition to being dangerous, the multiple handling and counting also looks a little ludicrous when one examines the end results. A spot check of 11 of the 112 Dallas County precincts using punch card machines in the November, 1968, general elections revealed that none produced a number of ballots equaling the number of voters.
Every election in which punch card voting has ever been used has been marred by the critical problem of “repunching” ballots that computers could not read.
For various reasons, many voters fail to punch out the chads of their candidates. Instead, they make pin-point holes, or indentations, somewhere between the chads. The computer rejects these cards. Election officials and party observers examine the cards and re-punch the votes on other cards. At best, they are guessing. It is manifestly impossible for them to know how the voters intended to vote.
In Dallas, use of the punch card voting system has caused an angry controversy.
The County Judge of the five-man Dallas County Commissioners’ Court—W. L. Sterrett—was opposed to the system from the beginning. Three Commissioners (M. G. Price, Jim Tyson, and Frank Crowley), acting as a quorum in Judge Sterrett’s absence, granted a $260,400.00 contract (July 2, 1968) to Datamedia Corporation for 1860 Vote-A-Corder machines (and 558 demonstrator units). The County was bound to keep and pay for the machines only if they worked satisfactorily in the November, 1968, general elections.
Long before the elections, there was much criticism of Commissioners Price, Tyson, and Crowley for getting the punch card system—because the system already had a record of poor performance throughout the United States (of disastrous failure, in some places). Hence, Price, Tyson, and Crowley were already defensive about the system before it was put to the test.
Since the election, Price and Tyson have become shrill and reckless in their efforts to prove the system worked well, despite compelling evidence to the contrary. While insisting that the punch card system handled the election well, Price and Tyson have accused County Clerk Tom Ellis and Judge Sterrett of sabotaging the election, and have even suggested that Ellis and Sterrett may be taking pay-offs from the manufacturers of other types of voting machines. Commissioners Price and Tyson seem determined to buy the Vote-A-Corders for Dallas County, regardless.
Testifying before a State legislative committee on November 15, 1968, Commissioner Price made an impassioned defense of the punch card voting system. He read telegrams (which he had solicited) from other communities reporting favorable experience with the system. He pointed out that, in those communities, an average of only one-fourth of one percent of all ballots cast had to be re-punched. In Los Angeles, one-half of one percent of all ballots cast had to be re-punched (13,423 cards re-punched out of 2,655,966 punch cards voted).
Commissioner Price cited these figures to show that malfunctions in the system are within the limits of tolerance. He ignored the fact that the presidential election this year was decided by a popular-vote margin of two-fifths of one percent.
Voting system malfunctions that can alter or invalidate one-half of one percent of all votes cast in one populous county could alter the outcome of all close elections, including presidential elections.
Is that tolerable?
Obviously, punch card voting lays elections wide open to undetectable and uncorrectable errors, to ballot-box stuffing, to ballot tampering, and to fraud.
But all of that is as nothing in comparison with what can occur when the punch card ballots finally reach the computer for counting.
Computers make errors. There have been many cases of computers writing checks for millions of dollars when the sums intended were in the hundreds. In a Dallas school this fall, one computer, making class assignments for students, gave one boy five lunch periods. Such computer errors are detectable, because they can be measured against known or reasonable quantities. But if a computer errs in tallying a million votes, the error is undetectable. No one knows what the correct sum should have been. There is no known quantity to check computer results against.
Computers can be rigged. They can be programmed to count votes any way the programmer wants them counted. A computer can even be programmed to count votes correctly in trial runs, and then to count them incorrectly in the actual run.
State legislatures should outlaw all voting systems that use computers to count votes.
Copyright by Dan Smoot, 1968.