Toggle Menu

In the Field of Chads

Why the hell isn't Al Gore--instead of me--doing this?

It's 11:30 in the morning, the third day of the new year, and that's what I am thinking as I sit in a bland conference room on the eighteenth floor of the Stephen P. Clark Government Center in downtown Miami. I am examining the infamous Miami-Dade ballots, the "undervote" punch cards that did not register a presidential preference when processed by tabulation machines on November 7. There were about 10,500 of these ballots (1.6 percent of the votes recorded in Miami-Dade) and up to 60,000 undervotes throughout the state. I've been at it--staring at one ballot at a time--for about two hours. The thrill is gone. The eyestrain has begun.

Only a few people are engaged in this history-making though drudgery-ridden exercise. Six organizations are paying the county $10 an hour for the privilege of reviewing the ballots. Those bothering are an accountant-reporter team from the Miami Herald (which is examining Florida undervotes in all but four small counties), a reporter from the Palm Beach Post, officials from the state Republican Party, a reporter from Inside Edition, several accountants retained by Larry Klayman's Judicial Watch (a conservative outfit that has filed dozens of lawsuits against the Clinton Administration) and yours truly. No Democratic Party officials are participating, and none of the media biggies have shown. A week later, the Wall Street Journal, the Associated Press, the New York Times, the Washington Post and others--rather than mount their own reviews--would form a cost-sharing consortium to examine the state's undervotes and overvotes. (The latter are ballots that were recorded bearing more than one vote in the presidential race; the state Supreme Court recount order did not extend to these ballots.)

There is a whimper-not-a-bang feel to the occasion. We sit at individual tables, and temps hold up the ballots. We are not allowed to touch the cards. We gaze at them, searching for dimples, bumps, bulges, punctures, jagged holes, pen marks, pinpricks, rips and hanging, swinging or dislodged chads. And we can judge the ballots entirely as we choose. As Miami-Dade officials repeatedly note, this is not a "recount"; it is an "inspection."

During three days of review, I will examine and consider the meanings of 3,409 ballots from precincts--including African-American neighborhoods that backed Gore and Cuban-American areas that went for Bush--that split the Gore/Bush vote 55 percent to 45 percent in Gore's favor. (The full countywide tally divided 53.2 to 46.8 percent Gore's way.) And the numbers? How many votes did Gore pick up? Would he have won Florida--and taken the nation--had the Miami-Dade recount not been thwarted by, first, the county elections canvassing board and, then, five Republican-appointed members of the US Supreme Court? Well, not so fast. I'll get to the totals. But here's a teaser: The results of this painstaking manual review contradicted the melodramatic spin of both the Bush camp and the Gore gang. The fundamental assertions pushed by each side--for the Bushies, it was that manual recounts are arbitrary acts of folly; for the Gore crowd, it was that if you count them, he will win--were undermined by these castaway ballots.

I and the other journalists arrived at the government center hoping to gather hard-and-fast answers to the murky questions floating in the wake of the messy presidential election. The Republican officials are present to keep an eye on the reporters. They are collecting ammunition, in case anyone in the media declares that Gore nets the 538 votes he needed to win the state. And Klayman is grabbing television face-time. As the review begins, he raises a fuss for the TV news cameras. He has asked that his accountants be allowed to sit at tables and review punch cards alongside the other participants. He claims he wants to speed up the process. It seems he is more interested in monitoring the inspections of others. David Leahy, the county elections supervisor, rejected the request, and Klayman huffs that the county "must have something to hide." His accusations are curious. Leahy, who holds a nonpartisan position, is a member of the three-person canvassing board that shut down the Miami-Dade recount the day before Thanksgiving, after khaki-clad Republican aides flown in from Capitol Hill mounted a thuggish protest on the nineteenth floor. The board's move enraged Democrats. And during the postelection period, Leahy consistently opposed conducting a manual review. (He says he did not believe a Miami-Dade recount would produce enough votes to alter the statewide results.) So why is Klayman giving him a difficult time? Moments later, Klayman informs me the county is worried because "there are irregularities here." He maintains that I will find such suspicious-looking ballots as punch cards with "chads scotch-taped back in."

As I proceed with the inspection, I do not encounter fraudulent-looking ballots, but I do see cards difficult to explain. This is what I am looking at: a seven-and-three-eighths by three-and-a-quarter inch card of heavy paper stock with twelve vertical rows of numbered boxes running from 1 to 312. Voters were instructed to insert this card into the plastic sleeve of the Votomatic machine. Then they turned through a ballot book that was attached to the device and that listed the various contests and candidates. Every time a page was flipped a different portion of the card was aligned beneath holes in the sleeve. To make their selections, voters stuck a sharp stylus into a specific hole--designated by arrows in the ballot book--to punch out the square-shaped chad of the appropriate box on the card. The first row on the card corresponded to the presidential race, and the candidates were assigned even-number slots. If a citizen voted for Bush, he or she broke the chad in box number 4. For Gore, it was box number 6. For libertarian Harry Browne, it was box number 8. And so on. Then the ballots were tallied by machines that counted the holes in the cards.

At my table Ruth Smith, a 76-year-old retired school aide from Queens, lifts each card. She tells me her son-in-law-the-attorney represents Mark Penn, Bill Clinton's pollster, and her grandson also works for Penn. (If Klayman or the Republicans find out that a woman this close to Clinton's most important adviser is handling the ballots--oh my!) The first eight ballots contain no marks on the presidential row. Then I spot several cards with the chad in the 7 box punched out. "What could this mean?" I ask Ivy Korman, the elections department official supervising the public inspection. "Don't ask us," she says. "We have no idea." Immediately I spot other clearly punched 7s and, soon after, a bunch of 5s. These boxes do not correspond to any selection in the ballot book. (More on this mystery later.) And there are many ballots that have no vote on the first row but are filled with well-defined holes elsewhere. Are these from voters who decided not to stab for a presidential candidate but who participated in down-ballot races? Many ballots contain not a scratch, hole, dent or bump. Did people take the trouble to go to the polling place and then not vote in any contest? The lines at voting sites were long, and some citizens left before reaching the Votomatics. Under the rules, their pristine cards were collected and placed with ballots that had been punched. Other ballots are more baffling: those with pinpricks across the portion of the card that does not match any contest in the ballot book. Some with a clear punch-out at 9, 11 or 13--or all three. Cards with punch-outs forming patterns--such as a straight line across the ballot--that are not in sync with actual races. Not everyone followed instructions. Are these willful political statements? Artistic expressions? Acts of ignorance? Or system-caused errors?

Within minutes I come across the ballots that drew me to this conference room. Here's a 6 that is plainly broken. The chad remains in place, but there is a hole along one side of it. How could this not have been a vote for Gore? Chads are sturdy beasts. They do not break on their own accord. Spend a moment with a ballot card and you will see that the Republicans prevaricated during the recount-a-rama when they claimed that ballots are fragile and handling corrupts them. As Leahy--no friend to the Democrats--says, "You can run a ballot through a reader 100 times and you'll never get any chads inadvertently punched out. The ballot won't disintegrate on the basis of normal handling." The hole on this ballot had to have been placed there. Intent is clear. And Florida law--and the statutes in many other states--says intent is what counts. I judge it an unrecorded Gore vote.

But a few ballots later, I am peering at a card with a slight indentation at box 6. No hole. No penetration. The perforation has held fast. What to do? It doesn't look like a manufacturing error. Did the voter--as some GOP spinners speculated--only consider voting for Gore and then, struck by remorse, withdraw the stylus before executing the final thrust? Unlikely, but possible. Then I spot a ballot with a sharp puncture mark in the chad for box 4, but the chad did not detach and no light shines through. Standards, I need standards.

It is not until I examine a couple of hundred ballots that I can construct guidelines. Regarding the 4s and 6s, I divide them into three categories. The first is for when the chad is absent. Why hadn't these cards been counted as votes? Perhaps the reading machines made a mistake or a hanging chad dropped after the card was tabulated. Also in this category, I place easy-to-recognize holes--puncture marks above the chad, openings that are partially blocked by swinging chads. The second category is reserved for marks that definitely seem a product of an effort to punch the card--deep indentations, punctures that allow a pinhole of light to pass, pushed-back chads that are perforated at spots. A fair-minded person looking at these cards would have to admit deliberate action was responsible for the disturbances. I also toss into this category my favorite anomaly: revolving-door chads. These are cards in which the chad completely turned around but remained tightly in place. The dot is now on the back side of the ballot, which likely means that a push of the stylus point spun the chad, as if it were on an axle. As for category three, it is for ballots with a small but discernible blunt or sharp bulge on the chad--a slightly pregnant chad. These marks are debatable. I record these votes, but I would not include them in a count.

As I continue, I find that my standards are not in accord with the rules adopted by the canvassing board during its aborted manual recount, which scrutinized the undervotes from 140 of the county's 614 precincts. (That review resulted in a net gain of 157 votes for Gore, but the precincts examined were heavily Democratic.) The evaluations of that recount were written on the back of the punch cards, and I see many ballots counted as votes for Bush or Gore that would not pass muster under my standards. On a few of these ballots, the barest bulge--do I see it or am I imagining it?--caused the board to award it to a candidate. (Of course, the lawyers of the other candidate challenged the determination.) These close calls are not irregularities; they are judgments. But the point is obvious: A hand recount should proceed under tight rather than loose standards. Especially in Miami--which has a recent history of vote fraud. During a break, a local reporter regales the out-of-towners with basic facts of Miami-Dade: "We account for 90 percent of the immigration fraud in this country. Twenty percent of our economy is underground. Twenty percent of our water is stolen, through meter bypasses. This is the way we do things here." Several feet from the entrance to the conference room is a sign, copies of which are posted by the Commission on Ethics and Public Trust throughout the government center, that reads, We Care About You! If You Have Information About Fraud, Waste, Corruption in Our Community, We Want to Know. Call Us.

By the end of Day Three, with my eyes screaming, I realize that clear answers will not be forthcoming. Republicans were correct to the extent that an attempt to evaluate certain punch cards does place a reviewer in the position of mind reader. But they were wrong in dismissing the value and legitimacy of hand recounts. It would not be difficult to create strict guidelines for a manual review. Slap each ballot on a light table, see if a beam passes through whatever mark is there. Count any ballot with a partially dislodged chad. Skip the subtle bumps and the maybe-it's-something impressions. And a manual recount of the undervotes need not have taken forever. The Klayman accountants, working at two tables, finished their review of the Miami-Dade ballots in less than three days (and Klayman did not immediately announce any findings). A hand review in Miami-Dade and other counties throughout the state--not only the four counties where the Gore team requested recounts--could have led to a more accurate tally without trampling on anyone's right to due process and equal protection.

Would such a recount have rewritten the outcome? Maybe not. After sifting through a third of the Miami-Dade undervote--a large-enough sample on which to reach conclusions while avoiding eye damage--I discover that 59 percent of the ballots contain no marks for President. Adding up the ballots in categories one and two, I unearth 119 votes for Gore and 114 for Bush. A measly gain for Gore. If category-three votes are included--and I wouldn't advise that--Gore's pickup increases by twelve. (After reviewing 4,000 of the Miami-Dade undervotes, the reporter from the Palm Beach Post discerned a modest boost for Bush.) Extrapolate these figures to the rest of the county, and Gore falls short of erasing Bush's statewide lead.

These numbers say nothing about other counties--where various news organizations have been and will be studying undervote and overvote ballots. And there's another nettlesome matter to consider: those 5s and 7s. In my sampling, 7s beat 5s 389 to 214. It seems reasonable to assume that most 5s were meant to be votes for Bush and most 7s for Gore, for there appear to be only two possible explanations for all these missing-but-unassigned chads. Either voters mistakenly placed the punch cards on top of the sleeve in the Votomatic (doing so lined up chad 5 with the Bush arrow in the ballot book and chad 7 with the Gore arrow) and then punched away, or there was a mechanical problem with the voting machines that caused hundreds of cards to misalign within the devices. Analyzing data from the county, Anthony Salvanto, a faculty fellow at the University of California, Irvine, found 1,012 7s among the Miami-Dade undervotes and 696 5s. Leahy denies that machine error--as opposed to voter error--could have produced these results, but Salvanto identified hundreds of undervote ballots where a citizen consistently punched unassigned holes one spot below those of Democratic candidates--as if the voter had attempted to vote a straight party line and had been undone by the machine. Add the 5s and 7s into the picture, and Gore bags enough votes to put the statewide numbers into question. But what judge would have ordered the inclusion of these votes?

My own review does not produce an unambiguous shift in the Bush/Gore count. Othere media recounts may well do so. But it indicates that accurate hand reviews could have been conducted--and that they had the potential to address, if not resolve, some of the doubt that shrouded the election. When the US Supreme Court halted the Florida recounts, a combative Justice Antonin Scalia wrote, "The counting of votes that are of questionable legality does in my view threaten irreparable harm to [Bush], and to the country, by casting a cloud upon what he claims to be the legitimacy of his election." That is, the American public had to be protected from information. The undervote ballots--though uncounted in the official tally--do speak, and they tell a story: of an election probably decided in part by voting-technology problems, and of election results that cannot be considered to represent definitively the will of the people who voted. These punch cards, which Scalia, Bush and even some time-to-move-on Democrats do not want to dwell upon, ought to cast a long and dark cloud.

David Corn

January 11, 2001

Why the hell isn’t Al Gore–instead of me–doing this?

It’s 11:30 in the morning, the third day of the new year, and that’s what I am thinking as I sit in a bland conference room on the eighteenth floor of the Stephen P. Clark Government Center in downtown Miami. I am examining the infamous Miami-Dade ballots, the “undervote” punch cards that did not register a presidential preference when processed by tabulation machines on November 7. There were about 10,500 of these ballots (1.6 percent of the votes recorded in Miami-Dade) and up to 60,000 undervotes throughout the state. I’ve been at it–staring at one ballot at a time–for about two hours. The thrill is gone. The eyestrain has begun.

Only a few people are engaged in this history-making though drudgery-ridden exercise. Six organizations are paying the county $10 an hour for the privilege of reviewing the ballots. Those bothering are an accountant-reporter team from the Miami Herald (which is examining Florida undervotes in all but four small counties), a reporter from the Palm Beach Post, officials from the state Republican Party, a reporter from Inside Edition, several accountants retained by Larry Klayman’s Judicial Watch (a conservative outfit that has filed dozens of lawsuits against the Clinton Administration) and yours truly. No Democratic Party officials are participating, and none of the media biggies have shown. A week later, the Wall Street Journal, the Associated Press, the New York Times, the Washington Post and others–rather than mount their own reviews–would form a cost-sharing consortium to examine the state’s undervotes and overvotes. (The latter are ballots that were recorded bearing more than one vote in the presidential race; the state Supreme Court recount order did not extend to these ballots.)

There is a whimper-not-a-bang feel to the occasion. We sit at individual tables, and temps hold up the ballots. We are not allowed to touch the cards. We gaze at them, searching for dimples, bumps, bulges, punctures, jagged holes, pen marks, pinpricks, rips and hanging, swinging or dislodged chads. And we can judge the ballots entirely as we choose. As Miami-Dade officials repeatedly note, this is not a “recount”; it is an “inspection.”

During three days of review, I will examine and consider the meanings of 3,409 ballots from precincts–including African-American neighborhoods that backed Gore and Cuban-American areas that went for Bush–that split the Gore/Bush vote 55 percent to 45 percent in Gore’s favor. (The full countywide tally divided 53.2 to 46.8 percent Gore’s way.) And the numbers? How many votes did Gore pick up? Would he have won Florida–and taken the nation–had the Miami-Dade recount not been thwarted by, first, the county elections canvassing board and, then, five Republican-appointed members of the US Supreme Court? Well, not so fast. I’ll get to the totals. But here’s a teaser: The results of this painstaking manual review contradicted the melodramatic spin of both the Bush camp and the Gore gang. The fundamental assertions pushed by each side–for the Bushies, it was that manual recounts are arbitrary acts of folly; for the Gore crowd, it was that if you count them, he will win–were undermined by these castaway ballots.

I and the other journalists arrived at the government center hoping to gather hard-and-fast answers to the murky questions floating in the wake of the messy presidential election. The Republican officials are present to keep an eye on the reporters. They are collecting ammunition, in case anyone in the media declares that Gore nets the 538 votes he needed to win the state. And Klayman is grabbing television face-time. As the review begins, he raises a fuss for the TV news cameras. He has asked that his accountants be allowed to sit at tables and review punch cards alongside the other participants. He claims he wants to speed up the process. It seems he is more interested in monitoring the inspections of others. David Leahy, the county elections supervisor, rejected the request, and Klayman huffs that the county “must have something to hide.” His accusations are curious. Leahy, who holds a nonpartisan position, is a member of the three-person canvassing board that shut down the Miami-Dade recount the day before Thanksgiving, after khaki-clad Republican aides flown in from Capitol Hill mounted a thuggish protest on the nineteenth floor. The board’s move enraged Democrats. And during the postelection period, Leahy consistently opposed conducting a manual review. (He says he did not believe a Miami-Dade recount would produce enough votes to alter the statewide results.) So why is Klayman giving him a difficult time? Moments later, Klayman informs me the county is worried because “there are irregularities here.” He maintains that I will find such suspicious-looking ballots as punch cards with “chads scotch-taped back in.”

As I proceed with the inspection, I do not encounter fraudulent-looking ballots, but I do see cards difficult to explain. This is what I am looking at: a seven-and-three-eighths by three-and-a-quarter inch card of heavy paper stock with twelve vertical rows of numbered boxes running from 1 to 312. Voters were instructed to insert this card into the plastic sleeve of the Votomatic machine. Then they turned through a ballot book that was attached to the device and that listed the various contests and candidates. Every time a page was flipped a different portion of the card was aligned beneath holes in the sleeve. To make their selections, voters stuck a sharp stylus into a specific hole–designated by arrows in the ballot book–to punch out the square-shaped chad of the appropriate box on the card. The first row on the card corresponded to the presidential race, and the candidates were assigned even-number slots. If a citizen voted for Bush, he or she broke the chad in box number 4. For Gore, it was box number 6. For libertarian Harry Browne, it was box number 8. And so on. Then the ballots were tallied by machines that counted the holes in the cards.

At my table Ruth Smith, a 76-year-old retired school aide from Queens, lifts each card. She tells me her son-in-law-the-attorney represents Mark Penn, Bill Clinton’s pollster, and her grandson also works for Penn. (If Klayman or the Republicans find out that a woman this close to Clinton’s most important adviser is handling the ballots–oh my!) The first eight ballots contain no marks on the presidential row. Then I spot several cards with the chad in the 7 box punched out. “What could this mean?” I ask Ivy Korman, the elections department official supervising the public inspection. “Don’t ask us,” she says. “We have no idea.” Immediately I spot other clearly punched 7s and, soon after, a bunch of 5s. These boxes do not correspond to any selection in the ballot book. (More on this mystery later.) And there are many ballots that have no vote on the first row but are filled with well-defined holes elsewhere. Are these from voters who decided not to stab for a presidential candidate but who participated in down-ballot races? Many ballots contain not a scratch, hole, dent or bump. Did people take the trouble to go to the polling place and then not vote in any contest? The lines at voting sites were long, and some citizens left before reaching the Votomatics. Under the rules, their pristine cards were collected and placed with ballots that had been punched. Other ballots are more baffling: those with pinpricks across the portion of the card that does not match any contest in the ballot book. Some with a clear punch-out at 9, 11 or 13–or all three. Cards with punch-outs forming patterns–such as a straight line across the ballot–that are not in sync with actual races. Not everyone followed instructions. Are these willful political statements? Artistic expressions? Acts of ignorance? Or system-caused errors?

Within minutes I come across the ballots that drew me to this conference room. Here’s a 6 that is plainly broken. The chad remains in place, but there is a hole along one side of it. How could this not have been a vote for Gore? Chads are sturdy beasts. They do not break on their own accord. Spend a moment with a ballot card and you will see that the Republicans prevaricated during the recount-a-rama when they claimed that ballots are fragile and handling corrupts them. As Leahy–no friend to the Democrats–says, “You can run a ballot through a reader 100 times and you’ll never get any chads inadvertently punched out. The ballot won’t disintegrate on the basis of normal handling.” The hole on this ballot had to have been placed there. Intent is clear. And Florida law–and the statutes in many other states–says intent is what counts. I judge it an unrecorded Gore vote.

But a few ballots later, I am peering at a card with a slight indentation at box 6. No hole. No penetration. The perforation has held fast. What to do? It doesn’t look like a manufacturing error. Did the voter–as some GOP spinners speculated–only consider voting for Gore and then, struck by remorse, withdraw the stylus before executing the final thrust? Unlikely, but possible. Then I spot a ballot with a sharp puncture mark in the chad for box 4, but the chad did not detach and no light shines through. Standards, I need standards.

It is not until I examine a couple of hundred ballots that I can construct guidelines. Regarding the 4s and 6s, I divide them into three categories. The first is for when the chad is absent. Why hadn’t these cards been counted as votes? Perhaps the reading machines made a mistake or a hanging chad dropped after the card was tabulated. Also in this category, I place easy-to-recognize holes–puncture marks above the chad, openings that are partially blocked by swinging chads. The second category is reserved for marks that definitely seem a product of an effort to punch the card–deep indentations, punctures that allow a pinhole of light to pass, pushed-back chads that are perforated at spots. A fair-minded person looking at these cards would have to admit deliberate action was responsible for the disturbances. I also toss into this category my favorite anomaly: revolving-door chads. These are cards in which the chad completely turned around but remained tightly in place. The dot is now on the back side of the ballot, which likely means that a push of the stylus point spun the chad, as if it were on an axle. As for category three, it is for ballots with a small but discernible blunt or sharp bulge on the chad–a slightly pregnant chad. These marks are debatable. I record these votes, but I would not include them in a count.

As I continue, I find that my standards are not in accord with the rules adopted by the canvassing board during its aborted manual recount, which scrutinized the undervotes from 140 of the county’s 614 precincts. (That review resulted in a net gain of 157 votes for Gore, but the precincts examined were heavily Democratic.) The evaluations of that recount were written on the back of the punch cards, and I see many ballots counted as votes for Bush or Gore that would not pass muster under my standards. On a few of these ballots, the barest bulge–do I see it or am I imagining it?–caused the board to award it to a candidate. (Of course, the lawyers of the other candidate challenged the determination.) These close calls are not irregularities; they are judgments. But the point is obvious: A hand recount should proceed under tight rather than loose standards. Especially in Miami–which has a recent history of vote fraud. During a break, a local reporter regales the out-of-towners with basic facts of Miami-Dade: “We account for 90 percent of the immigration fraud in this country. Twenty percent of our economy is underground. Twenty percent of our water is stolen, through meter bypasses. This is the way we do things here.” Several feet from the entrance to the conference room is a sign, copies of which are posted by the Commission on Ethics and Public Trust throughout the government center, that reads, We Care About You! If You Have Information About Fraud, Waste, Corruption in Our Community, We Want to Know. Call Us.

By the end of Day Three, with my eyes screaming, I realize that clear answers will not be forthcoming. Republicans were correct to the extent that an attempt to evaluate certain punch cards does place a reviewer in the position of mind reader. But they were wrong in dismissing the value and legitimacy of hand recounts. It would not be difficult to create strict guidelines for a manual review. Slap each ballot on a light table, see if a beam passes through whatever mark is there. Count any ballot with a partially dislodged chad. Skip the subtle bumps and the maybe-it’s-something impressions. And a manual recount of the undervotes need not have taken forever. The Klayman accountants, working at two tables, finished their review of the Miami-Dade ballots in less than three days (and Klayman did not immediately announce any findings). A hand review in Miami-Dade and other counties throughout the state–not only the four counties where the Gore team requested recounts–could have led to a more accurate tally without trampling on anyone’s right to due process and equal protection.

Would such a recount have rewritten the outcome? Maybe not. After sifting through a third of the Miami-Dade undervote–a large-enough sample on which to reach conclusions while avoiding eye damage–I discover that 59 percent of the ballots contain no marks for President. Adding up the ballots in categories one and two, I unearth 119 votes for Gore and 114 for Bush. A measly gain for Gore. If category-three votes are included–and I wouldn’t advise that–Gore’s pickup increases by twelve. (After reviewing 4,000 of the Miami-Dade undervotes, the reporter from the Palm Beach Post discerned a modest boost for Bush.) Extrapolate these figures to the rest of the county, and Gore falls short of erasing Bush’s statewide lead.

These numbers say nothing about other counties–where various news organizations have been and will be studying undervote and overvote ballots. And there’s another nettlesome matter to consider: those 5s and 7s. In my sampling, 7s beat 5s 389 to 214. It seems reasonable to assume that most 5s were meant to be votes for Bush and most 7s for Gore, for there appear to be only two possible explanations for all these missing-but-unassigned chads. Either voters mistakenly placed the punch cards on top of the sleeve in the Votomatic (doing so lined up chad 5 with the Bush arrow in the ballot book and chad 7 with the Gore arrow) and then punched away, or there was a mechanical problem with the voting machines that caused hundreds of cards to misalign within the devices. Analyzing data from the county, Anthony Salvanto, a faculty fellow at the University of California, Irvine, found 1,012 7s among the Miami-Dade undervotes and 696 5s. Leahy denies that machine error–as opposed to voter error–could have produced these results, but Salvanto identified hundreds of undervote ballots where a citizen consistently punched unassigned holes one spot below those of Democratic candidates–as if the voter had attempted to vote a straight party line and had been undone by the machine. Add the 5s and 7s into the picture, and Gore bags enough votes to put the statewide numbers into question. But what judge would have ordered the inclusion of these votes?

My own review does not produce an unambiguous shift in the Bush/Gore count. Othere media recounts may well do so. But it indicates that accurate hand reviews could have been conducted–and that they had the potential to address, if not resolve, some of the doubt that shrouded the election. When the US Supreme Court halted the Florida recounts, a combative Justice Antonin Scalia wrote, “The counting of votes that are of questionable legality does in my view threaten irreparable harm to [Bush], and to the country, by casting a cloud upon what he claims to be the legitimacy of his election.” That is, the American public had to be protected from information. The undervote ballots–though uncounted in the official tally–do speak, and they tell a story: of an election probably decided in part by voting-technology problems, and of election results that cannot be considered to represent definitively the will of the people who voted. These punch cards, which Scalia, Bush and even some time-to-move-on Democrats do not want to dwell upon, ought to cast a long and dark cloud.

David Cornis Mother Jones' Washington bureau chief. Until 2007, he was Washington editor of The Nation.


Latest from the nation