What the NORC Data Will Show

by Paul Lukasiak
1 November 2001

We don't need to wait for the media consortium to report the results of the NORC review of Florida ballots. It is already obvious from precinct-based results and previous media reviews that Al Gore got the most votes.

"There are conspiracy theories bouncing around the Internet that the news organizations studying last November's Florida vote are suppressing the results for fear of tarnishing President Bush in time of war. Not true. In fact, participants say, the results have not been tabulated or analyzed. No journalist has a clue whether they favor Bush or Al Gore."
(Howard Kurtz, Washington Post, October 15, 2001)

The controversy over the decision to withhold the data from the National Opinion Research Center's (NORC) Florida ballot survey has fueled a great deal of speculation regarding the conclusions that will be drawn from that data. While media drones like Howard Kurtz of the Washington Post continue to spin the fiction that nobody knows what the NORC data will show, the truth (and Kurtz must know it) is that the NORC data will merely confirm conclusions that can be drawn from the abundant information available from precinct-based returns, various other media recounts, and other sources.

The reporters and analysts from the media consortium have studied the other data sources in order to know what to look for when they get the NORC data. Some of these reporters and analysts have acknowledged privately that their newspapers have not done stories based on other ballot reviews because their editors and publishers want to wait until they have the NORC data.

In sum, the media consortium withheld election-based stories for the past nine months because they wanted to get the biggest bang from their NORC bucks. And as other reviews and investigative reports uncover the truth about what happened in the Florida election, they have realized another reason for withholding such stories. According to two independent journalists and other sources, the NORC data have been available since early July. Release of the data has been delayed by members of the media consortium that did not want to publish facts that would lead to the inevitable conclusion that the wrong person is in the White House — and that he got there through corrupt practices by Florida Republicans.

The consortium is now claiming that a "lack of resources" is what is holding up the release of the NORC data. This is demonstrably false. Their abandonment of wall-to-wall coverage of the war on Afghanistan for wall-to-wall coverage of the anthrax scare and reports on how Chandra Levy's parents feel about the FBI diverting investigators from their daughter's disappearance (1) show that the media's ostensible "lack of resources" is a fig leaf for something far more sinister. The media consortium is "managing" the news, deciding to withhold relevant facts from the American people in pursuit of its own agenda. But we do not need the NORC data to know what unavoidable conclusion will be based on it.


A "Perfect" Election

If no one had ever heard of hanging chads, if the butterfly ballot had never flown, if no voter had bungled in the booth, who would have won Florida and the presidency of the United States? In a race so tight, it may never be known for certain. But a Herald-commissioned analysis of voting patterns in each of the state's 5,885 precincts suggests that Florida likely would have gone to Al Gore — by a slim 23,000 votes — rather than George W. Bush, the officially certified victor by the wispy margin of 537.
("If the Vote Were Flawless," Miami Herald, December 3, 2000)

Under any consistent statewide standard for counting ballots, Al Gore would have won in Florida. Bush's victory was determined by the decisions made by some county election boards to selectively:

All of these decisions consistently benefited Bush, and were made almost exclusively in politically conservative counties.

Had the Florida election been free of faulty equipment, misleading ballot design, violations of election law, a partisan double standard for interpreting "voter intent," and partisan decisions by elections officials, Al Gore would have led George W. Bush by a sufficient margin to prevent any reasonable doubt that Gore won the state.

As many as 45,000 net Gore ballots were uncounted. Of these, 10,000 were spoiled because of poor ballot design, and more than 500 were unambiguously legal votes that were illegally not counted. Hundreds of others were ballots that contained errors, but from which the voter's intent could clearly enough be discerned that under Florida law they could and should have been counted. The number of ballots that legally should have been counted, but were not, can easily be determined from published figures and a statistical comparison of ballots that were similar in design.

We can also expect these conclusions to be supported by the NORC data:

Had it not been for the consistently pro-Bush bias of state elections officials and some county elections officials, the initial mandatory machine recount would have shown Al Gore in the lead by at least 200 and probably well over 2,000 votes. Had Florida's election laws been followed, and followed consistently, it is certain that Al Gore would be in the White House today.

Under the control of Republican operatives — Secretary of State Katherine Harris and Division of Elections Director Clay Roberts — Florida elections officials ignored their statutory duties to ensure that the law was followed consistently. The evidence indicates that they did so deliberately.

The significance of the NORC project is not in what the data will show, but in the fact that this nation's most influential news media organizations have chosen to ignore the data available from other sources in favor of its own "definitive" study of the Florida ballots. Stories of national importance have gone unwritten. Clear evidence of corruption on the part of election officials has been ignored.

Below are some of the stories that the media consortium has decided not to tell you because it is waiting for its own "evidence."


Gore's 45,000+ Vote Margin

"Neither the centre nor the consortium knows whether the data suggest that, had the uncounted votes been tallied, Mr. Gore or Mr. Bush would have won the state. [Steve Goldstein of Dow Jones Co., publisher of the Wall Street Journal] rejected the suggestion that the media might be avoiding the story for fear of embarrassing the President in a time of national crisis. 'It has absolutely nothing to do with that whatsoever," he said. "The priorities have changed. People are focused on the fact that we're at war.' "
(John Ibbitson, Toronto Globe and Mail, October 11, 2001)

Steve Goldstein should, and probably does, know better, because there is already a great deal of data available that reveals precisely what the NORC data "suggests."

A precinct-based extrapolation by the Miami Herald (2) of the more than 185,000 uncounted ballots found that Gore would have received an additional 23,468 net votes had there been no errors, and had every person wanted to vote for president. Of course, not everyone wants to vote for president; but in counties with voter protection, the percentage of those choosing to not vote for president was less than one half of one percent of the total votes cast. Gore's dominance of the overvoted ballots shows that, even allowing for intentionally unvoted ballots, his lead in a "perfect" election would have been close to 20,000 votes. (The Herald undervote analysis found approximately 28,000 ballots with no mark for president.)

The Herald's undervote review (3) found at least 1,300 more ballots with marks for Gore than ballots with marks for Bush. (That number, however, includes about 1,000 ballots with clear punches from Broward County that may have been part of the County's final total. Gore's net advantage in the undervote would be between 800 and 1000 votes if those Broward County ballots are not included.)

A review of the data tapes from Dade County also found evidence of more than 1,700 additional ballots that voters intended to contain votes. These ballots contained detached chads that were unassigned but were directly below the chads assigned to Bush and Gore, and which could only have been punched because the ballot had been misaligned in the voting device. An analysis of these ballots found that Gore would have received about 316 additional net votes had the ballots been correctly aligned.(4)

The Herald estimate (cited above) of the votes Gore would have gained from an "error-free" election may understate the case. Of the 111,251 overvoted ballots examined by the Miami Herald, 12,645 contained marks for both Gore and Bush. Of the remaining ballots, 25,079 had marks for Bush, and 71,548 had marks for Gore, a difference of more than 45,000 ballots with Gore marks than with Bush marks.


How Ballot Design Cost Gore More Than 10,000 Net Votes

Palm Beach County's "butterfly ballot" has received the most attention from the media, but Palm Beach County is not the only place where poorly designed ballots had played a significant role in determining the outcome of the election. Duval County, with a two-page presidential ballot and misleading instructions to the voter to "vote all pages," accounted for thousands of overvoted ballots, resulting in a net loss for Gore of thousands of votes.(5) Fourteen of the fifteen "optical scan" counties that tabulated ballots in the central office (not in the precincts) used a confusing "two-column" presidential ballot design that resulted in thousands more overvoted ballots, again causing a net loss for Gore of hundreds of votes.

Palm Beach County

Gore lost votes in two ways because of voter confusion in Palm Beach County. An estimated three thousand votes intended for Gore were cast for Pat Buchanan because, although Gore's name appeared to be second in the ballot booklet, Buchanan held the second position on the actual ballot card. But Gore lost thousands more net votes because of overvoting that can be attributed to the "butterfly" design of the ballot. The table below shows the chad assignments on the Palm Beach ballot.

Chad Number Candidate Assigned
2not counted

It was impossible to overvote the ballot by removing the chad for Bush (3) and the one immediately above it (2). That chad was not counted, and it was covered by a plastic "mask." But it was possible to overvote the ballot by punching the Gore chad (5) and the one above it (Buchanan); and 5,237 people did so.

It was also possible to overvote by punching out both the chad for the intended candidate and the chad directly below it: 1,600 people punched out both the Bush chad and the one below it (chads 3&4) while 2,856 people punched both the Gore chad and the one below it (chads 5&6). (These totals are for ballots that were overvoted for two candidates only, and do not include ballots on which three or more chads were detached.)

The butterfly design cost Gore about 6,493 votes on ballots that were overvoted. Adding these votes to the 3,000 votes lost because people punched the Buchanan chad instead of the Gore chad shows how the butterfly ballot cost Gore nearly 10,000 votes.

Of particular note here is that Gore voters were no more confused than Bush voters by the butterfly ballot. Gore received 63.8% of Palm Beach County's votes that were cast for either Bush or Gore. When you look at the people who voted overvoted by punching out the chad for either Bush or Gore and the chad directly below it, you find that a similar percentage (64.1% ) of those ballots were errors made by Gore voters (2856/(1600+2856)).

In other words, "Bush/Buchanan" votes and "Gore/McReynolds" votes occurred with the same frequency as Bush and Gore votes. But it was impossible for Bush voters to err by punching out the chad above Bush's chad, while 5,237 people punched out both the Gore and Buchanan chads. Thus it is likely that, had it been possible to punch out the chad above Bush's, and had that chad been counted as a "vote," Bush would have lost an approximate additional 2,900 votes because of overvoting.

To get some idea of how the butterfly ballot design influenced the election, compare the results from Palm Beach County to the results from Broward County, where the Gore chad was directly below the Bush chad. In Broward, 1,313 voters overvoted their ballots by punching out both chads (Bush chad and the Gore chad below it) while 2,015 punched out the Gore chad and the one below it.

The Herald data also demonstrate the difference between "ballot confusion" (created by poor design and instructions that lead voters to punch more than one hole), and confusion about the law concerning voting (you can only vote for one candidate). While 1,313 voters in Broward overvoted by punching out Bush and Gore chads, this error was made by only 364 Palm Beach County and 448 Dade voters. When adjusted for the total number of votes cast, voters were 250% to 300% more likely to overvote solely because chads for listed presidential tickets were contiguous, as opposed to intentionally casting votes for two listed candidates.

Duval County

Duval County's 8.29% overvote rate was the highest for a punch-card county in Florida, almost twice as high as the overvote rate in Palm Beach County. At least half of that overvote rate can be attributed to Duval County's decision to list the presidential candidates on two pages. Duval County exacerbated the problems that would be caused by a two-page presidential ballot by sending out a sample ballot that looked nothing like the ballot used at the polls, and which included instructions to the voter to "vote all pages."

Bush, Gore, and three other candidates appeared on the first page of the presidential ballot. Five other candidates (Hagelin, Buchanan, McReynolds, Phillips, and Moorehead, and a chad associated with the write-in instructions were on the second page.

The following table provides totals of overvoted ballots that contain both a vote for either Bush or Gore on the first page and a vote for a candidate on the second page. The table contrasts the Duval overvotes with the Broward County overvotes for the same combination. Despite the fact that more than twice as many voters went to the polls in Broward County, in Duval County there are 13 times more overvotes that included "second page" candidates than in Broward. When adjusted for the number of votes cast, the two-page ballot increased the likelihood of overvotes for the "second page" candidates by more than 2,600%.

Duval   Broward
Gore Bush   Gore Bush
1,888 1,987 Hagelin 161 45
3270 1905 Buchanan 122 45
526 147 McReynolds 68 13
655 278 Phillips 109 26
654 187 Moorehead 219 88
169 51 write-in* 0 0
7162 4555 (Total for Gore/Bush) 679 223
* These "write-in" overvotes were actually legal votes that Duval Election officials illegally decided not to count
(see Lost Votes — The Uncounted Legal Votes).

After adjustment for (a) the number of votes cast, (b) Broward's heavy democratic vote, and (c) Duval's heavy Republican vote, the data show that Duval's decision to use a two-page ballot and provide the wrong instructions cost Al Gore approximately a net 2,500 votes in Duval County alone.

Duval County's NAACP had made an extraordinary effort to register new Black voters in that county and to explain the voting procedure to them. The NAACP had no way of knowing that the ballot instructions they were giving these new voters was wrong, because the sample ballot included the "vote all pages" instruction.

The high "Gore-Buchanan" vote and "Bush-Hagelin" vote has been attributed to voters being told that if they wanted to vote straight-ticket Republican, to vote for the first candidate on each page, or if they wished to vote straight-ticket Democratic, to vote for the second candidate on each page. This explanation, however, assumes that voters were told that this was how to vote a straight ticket; and there is no evidence that such instructions were ever given to large numbers of voters. In the absence of evidence that such instructions were given, the data provide evidence of ballot tampering in which someone other than the voters punched out large numbers of "Buchanan" chads in Duval's four black majority voting districts.

Two-Column Optically Scanned Ballots

Under Florida law, the state Division of Elections sends out a sample ballot each year listing the candidates for statewide office, as well as the order in which different offices should appear on the ballot. This year, the sample ballot sent by the state had the presidential candidates listed in two separate columns, with the first eight candidates in the first column, and the last two candidates plus the write-in line in the second column.

Florida has 41 counties that use optical scanners. Of those counties, 26 have ballots that are tabulated at the precinct where they are cast; almost all of these 26 counties offer "voter protection" against overvotes, so that election officials notify voters that their ballots contained errors, and give the voter a new ballot. Every one of the 26 "precinct-scanned" counties designed their presidential ballot so that all candidates were listed in a single column.

The other 15 counties that use optical scanning equipment tabulate ballots at a central location. None of these counties can offer "voter protection." Only one of the centrally-scanned counties (Lake County) put all of the presidential candidates in a single column.

In other words, with one exception, the counties in which voters did not have the opportunity to correct voting errors used a ballot design that could cause confusion and thus cause voting errors.

The table below lists overvote totals from the 14 "two-column" counties, and compares them to the combined totals from two "single-column" counties: Lake County, which was centrally scanned; and Escambia County, which chose to disable the "voter protection" feature on its precinct scanning equipment. Slightly fewer than 189,000 votes were cast in the "two-column" counties, and slightly more than 212,000 votes were cast in the "single column" counties used in the table below.

Subtracting the one-column overvotes from the two-column overvotes, makes it obvious how the decision to use a two-column ballot in fourteen counties was responsible for more than 3,700 overvotes. These overvotes cost Gore more than 400 net votes. A decision (contrary to state law) not to count ballots that contained a vote in the printed list plus a write-in vote for the same candidate cost Gore another 100 net votes.

Hagelin, Buchanan, and McReynolds all appeared in the first column on the two-column ballots; Moorehead and the "write-in" line appeared in the second column on all two-column ballots; Phillips was in the second column in almost all of the two-column ballots.

Two-Column   One-Column
Gore Bush   Gore Bush
61 45 Hagelin 63 34
194 175 Buchanan 114 94
82 28 McReynolds 21 8
974 1138 Phillips 25 58
1098 512 Moorehead 102 28
2409 1898 Subtotal for Bush/Gore 325 222
254 163 Write-in* 547 328
2663 2061 Totals for Bush/Gore 872 550
* These "write-in" overvotes were actually legal votes that were not counted because of decisions made by county officials. Six of the "two-column" counties (Gadsden, Hamilton, Jackson, Lafayette, Levy, and Liberty) counted these votes properly. Two other "two-column" counties (Suwanee and Taylor) and one "one-column" county (Escambia) counted only their absentee ballots properly, while failing to count the poll-cast votes.

The precinct-scanned counties had a special incentive to ignore the state's two-column design — voter protection meant additional expenses to the county for printing enough ballots (at approximately 23 cents each) to allow for giving voters replacement ballots when errors are detected.

Of the 14 centrally scanned counties that used two-column ballots, all but one voted for Bush. All of the 19 precinct-scanned counties that voted for Bush used a single-column ballot.

It is easy to see how the two-column design could lead to voter confusion and ballot errors; the question is why all but one centrally scanned county chose to use a design that would increase voter confusion. Republicans dominated the overwhelming majority of optically scanned counties, and there is an apparent pattern of election officials in these counties using ballot designs that would increase voter error when "voter protection" was not available.


"Lost Votes" — The Uncounted Legal Votes

"So they have the data, but they have no way of counting them. We have the counters written as software programs, but we do not have the data. So — [LAUGHS] because the two have never been joined, nobody — nobody on earth — knows what that data shows."
(Washington Post Database Editor Dan Keating on NPR, October 6, 2001)

More than 2,000 ballots containing votes that were specifically defined as legal votes under Florida law went uncounted in the Florida presidential election. Had these votes been counted, Al Gore would have received over 500 additional net votes.

Thousands of Florida ballots contained both a mark in the list of Presidential candidates and a mark in the space adjacent to the write-in instructions. Under unambiguous Florida law, a mark adjacent to the write-in instructions is not a vote (the name written in the write-in space constitutes the vote, not the mark next to the space), and cannot be counted as a "vote" when deciding whether a ballot has been overvoted.

Florida statutes provide mandatory provisions to ensure that ballots with marks next to the write-in instructions are not counted as overvotes. These provisions are to be followed every time ballots are counted, and not just during a "manual recount" of the ballot. Under the law, every ballot must be examined for the presence of a possible write-in vote during every canvass of the returns. If a possible write-in vote is found, then the ballot must be counted based on specific definitions in Florida of what constitutes a legal vote when there is a possible write-in vote on the ballot.

Florida's write-in laws say that a write-in can only be counted as a vote if the name of a "qualified" write in candidate" has been written in, and says that candidates whose names appear on the ballot cannot also be "qualified write-in candidates." Unless the name of a "qualified write-in candidate" is written in, the write-in vote is considered not to exist. The law states:

An overvote shall occur when an elector casts a vote on the ballot card and also casts a write-in vote for a qualified write-in candidate for that same office. Upon such an overvote, the entire vote for that office shall be void and shall not be counted. However, an overvote shall not occur when the elector casts a vote on the ballot card but then enters a sham or unqualified name in the write-in space for that same office. In such case, only the write-in vote is void. [FAC 2.0031 (7)] (italics added)

The examples in the following table illustrate how ballots should be counted.. ("Chote" was the name of a qualified write-in candidate in the Presidential election. "Mickey Mouse" was not.)

Hole or mark for GORE Hole or mark for BUSH Hole or mark next to write-in instructions Name written in Should be counted as...
yes yes not relevant none overvote
no yes yes Chote overvote
no yes no Cote overvote
no yes yes or no Gore Bush vote
no yes yes or no Bush Bush vote
no yes yes or no Mickey Mouse Bush vote
no yes yes none Bush vote
no no no Bush discretionary
no no yes Bush discretionary

As unambiguous as this definition is, many Florida counties simply ignored it. As a result, more than 2,000 legal votes were never counted. Hundreds of people who marked their ballots for Gore and wrote in "Gore" or marked their ballots for "Bush" and wrote in "Bush" had their ballots completely ignored.(6)

The counties where this egregious violation of election laws had the greatest impact on the election totals were all counties where Bush won by significant margins. Five heavily Republican counties accounted for nearly 80% of these uncounted legal votes, yet accounted for only slightly more than 10% of the votes cast statewide.

County Bush "Lost" Votes Gore "Lost" Votes Net Gore Loss Total Bush Votes Total Gore Votes Total Bush % Total Gore %
Charlotte 86 139 53 35,426 29,645 54.4% 45.6%
Duval 51 169 118 152,098 107,864 58.5% 41.5%
Escambia 64 148 84 73,017 40,943 64.1% 35.9%
Lake 264 399 135 50,010 36,571 57.8% 42.2%
Manatee 120 163 43 57,952 49,177 54.1% 45.9%
Totals 585 1,018 433 368,503 264,200 58.2% 41.8%

Charlotte and Lake Counties were optically scanned, centrally tabulated counties.

Escambia and Manatee Counties were optically scanned, precinct-tabulated counties that chose to turn off the voter protection feature at the polls. Escambia County chose to count its absentee ballots (where Bush received an extraordinary 72% of the votes) using a "voter intent" standard that recovered the "lost votes," while ignoring the similar fully legal votes that were cast at the polls. (See Counting Ballots Differently in the Same County — And Statewide.)


Duval County

Duval County deserves special attention. As noted above, Duval used a two-page presidential ballot, with the "write-in instructions" on the second page. Duval County used punch cards, but unlike virtually every other large counties (250,000+ voters), only in Duval County was it possible to punch a hole next to the write-in instructions. Other counties used the plastic mask to prevent such holes being punched, because such holes serve no useful purpose with punch-card ballots. Punch-card tabulators cannot flag ballots marked for "write-ins" the way optical scanning equipment can.

Duval County also violated the law concerning the wording of the write-in instructions that appear in the ballot booklet.

In Duval County, unlike most punch-card counties where write-in candidates are written in at the bottom of the ballot card, voters who wanted to cast a vote for a write-in candidate wrote the name on the "secrecy envelope" in which the ballot was placed before it was dropped into the ballot box. Duval County election workers were told to examine only the secrecy envelope, and not the actual ballot card, for evidence of write-in votes — despite the statutory requirement that both the ballot card and secrecy envelope be examined when the names of write-in candidates are to be written on the secrecy envelope.

This decision to violate Florida's election laws and ignore "best practices" followed in other punch-card counties cost Al Gore more than 118 net votes.(7)


Counting Ballots Differently in the Same County And Statewide

According to the Orlando Sentinel, the totals from the mandatory machine recount included more than 10,000 Florida ballots were evaluated based on a "voter intent" standard, then "duplicated" and tabulated by machine. Virtually all of these "hand-counted" ballots were from 26 optically scanned counties.(8)

Six of the 15 centrally scanned counties checked all of their ballots for "voter intent."

Of the optically scanned counties that provided "voter protection" at the polls, 11 checked only their absentee ballots. According to election officials from these counties, the "hand counting" of the absentee ballots was done to provide at least something akin to the "voter protection" that voters at the polls received. From these 11 counties, Al Gore would have gained an additional 9 net votes from among the overvoted ballots had the poll cast ballots also been "hand counted."

More significantly, four counties which did not provide voter protection at the polls chose to "hand count" only their absentee ballots using a "voter intent" standard. In other words, in four Florida counties, absentee ballots were given "error protection" that was not available for poll-cast ballots.

Two of these counties (Taylor and Suwanee) were centrally tabulated, and could not offer voter protection at the polls. Two other counties (Escambia and Columbia) were precinct-tabulated counties which chose not to offer "overvote" protection at the polls. As a result of the differing treatment of these ballots, Al Gore was denied an additional net gain of 186 votes.

In all four Bush won with margins of at least 21%.

The table below shows the number of "recoverable" poll-cast overvoted ballots (those showing voter intent, based on the criteria used by the Miami Herald), as well as the total Bush and Gore percentages for each county.

  Bush Recoverable Gore Recoverable Bush Total % Gore Total %
Columbia 8 14 60.9% 39.1%
Escambia 132 289 64.1% 35.9%
Suwanee 2 16 66.3% 33.7%
Taylor 10 19 60.5% 39.5%
Totals 152 338    

The Herald data also indicate that Gore would have picked up an additional 26 net votes from the undervoted ballots cast at the polls in these four counties. The total net gain for Gore would have been 212 votes if these four counties had treated all their ballots equally, instead of giving special preference to absentee ballots. If all of the counties that counted absentee ballots by hand had also counted their poll-cast ballots by hand, Gore's net gain would have been 221 votes.

Gore's net gain would have been even greater if all of the optically scanned counties had "hand-counted" all of their ballots.

The decision of the optically scanned counties that counted all of their votes using a "voter intent" standard is supported by precedents established by both the state Division of Elections and the Florida Supreme Court. In 1990, the Division of Elections issued an opinion that Pinellas County election officials should hand-count overvoted and undervoted ballots when elections are so close that a machine recount is required. However, under Florida law, this opinion was binding only on Pinellas County, and only for that election, although counties are urged to act in a manner consistent with these Division of Elections opinion.

In 1998, the Florida Supreme Court ruled in Beckstrom v. Volusia County that overvoted and undervoted ballots met the definition of "damaged or defective" ballots which, under Florida law, must be counted by hand using the "voter intent" standard.

It is possible to excuse the decision by centrally scanned counties that did not use the "voter intent" standard. It is also possible to understand the decision by the counties that provided voter protection at the polls to hand-count only absentee ballots using the voter intent standard. But there is no reasonable excuse for the four counties that did not offer "voter protection" to count absentee ballots using an entirely different standard than the standard for ballots cast at the polls.

The Herald analyses of the undervotes and overvotes found more than 5,000 optically scanned ballots that contained votes according to their criteria for "voter intent." The Herald totals indicate that if all of the optically scanned counties had counted their votes according to legal precedent, Al Gore would have picked up an additional 1,001 net votes.

Another way of looking at the issue of disparate vote counting is to consider what would have happened if the selective hand-counting had not taken place. If all 10,000 ballots that were hand-counted were excluded, Bush's net vote total would probably have been at least 3,000 votes lower, based on the 2-1 advantage in absentee ballots that Bush held in the counties that counted absentees using the "voter intent" standard. In other words, if only machine countable votes were included in the totals, Bush would have received 3,000 fewer votes, and thus would have lost the election decisively.


Illegally Counted Late Absentee Ballots

According to an investigative report in the New York Times (9), at least 680 of the 2,490 "late" absentee ballots were counted illegally, with Bush receiving an estimated net gain of 292 vote from these illegally counted ballots. The Times detailed how Bush's lawyers selectively raised issues of patriotism in those counties where Bush was strongest, urging that ballots be counted despite unequivocal evidence that these ballots violated the anti-fraud provisions of Florida election law. By contrast, in those counties where Gore gained a majority of absentee ballots, Bush's lawyers insisted upon strict enforcement of the anti-fraud provisions of the law. The intention of the Bush team was obvious from this disparity: not to make sure all legal votes were counted, but to make sure Bush received the greatest possible number of votes, legal or illegal.


What If...

There are only two ways to count the Florida ballots without blatant "spin." The first is to consider all the legal ballots on which there was a "clear indication of voter intent." The second is to consider what would have happened if "voter intent" had not been a factor — if only machine countable and legally prescribed votes had been included in the totals. All other vote-counting criteria rely on inconsistent application of Florida law.

Under both the "voter intent" scenario and the "no interpretation" scenario, Gore won in Florida.

What If...

...All ballots showing voter intent were counted, and all illegal votes were excluded?

Gore wins by 1,054 votes.

  Gain for Bush Gain for Gore
Machine Recount 300  
(New York Times) Legal late absentees 447  
(Herald) Marked Punch Card ballots*   800
(Herald) Marked Op-Scan Ballots   1001
Total Gains 747 1801
* There is evidence that some Broward County ballots that were part of the Herald undervote count were already included in the machine-counted totals. The Herald estimates that Gore's lead in punch-card ballots should be reduced by about 500 votes. Further analysis of these numbers shows that the proper estimated reduction in Gore votes would be approximately 275 votes. In either case, Gore still wins.

There will no doubt be attempts to raise questions regarding whether a "dimpled" ballot actually represents a vote. However, the standard in Florida for manually counting ballots is a "clear indication of voter intent"; and any mark for a single candidate — be it a dimple on a punch-card ballot or a circled name on an optically scanned ballot — is an indication of intent.

Dimples are in fact votes, and are the result of either the same kind of voter error that results in "circled names" on optically scanned ballots, or of problems with the voting device that somehow made it impossible to dislodge the chad. Numerous studies show that it takes far more force to "dimple" a ballot than it does to dislodge a chad; and the number of "non-voters" is consistent with the number of "blank ballots" and completely inconsistent with the number of dimpled ballots. Additionally, Gore's ballot position in many of the punch-card counties was directly above a "support bar" that significantly exacerbated problems associated with "chad build-up" and thus made it far more difficult to completely dislodge the chad.

All of this information is readily available to the various members of the media consortium. Any attempt to present numbers that cast doubt on whether a dimpled ballot represents a vote must be considered pure spin. One of the "acid tests" to determine whether the consortium members are spinning the NORC numbers will be how they handle the issue of dimpled ballots. Those members that explain why such ballots must be counted as votes should be regarded as far more credible than any media outlet that gives credence to the notion that tens of thousand of people "changed their minds and decided not to vote." The facts simply do not support such a claim.

What If...

...Only ballots that met the strictest standards (machine countable, or prescribed by law) were counted?

Gore wins by 2,755 votes.

  Gain for Bush Gain for Gore
Machine Recount 300  
(New York Times) Legal late absentees 447  
Bush loss from "voter intent" absentee ballots (est.) -3000  
(Herald) Gore gain in prescribed legal votes   502
Net Total -2253 502

This analysis is based on the assumption that counties would follow the law with regard to the counting of votes, but would not hand-count any ballots solely because voters did not follow instructions. The "candidate-plus-write-in" votes are included because Florida law includes both specific provisions detailing the procedures to use when counting those ballots and specific criteria defining what constitutes a vote on such ballots.

This analysis is key because it demonstrates how the entire process was perverted by Bush and his cronies. It was only because of the selective inclusion of hand-counted "voter intent" ballots in the "machine recount" numbers that the issue of "manual recounts" in Palm Beach, Broward, and Dade Counties became important. If only machine-countable and legally defined votes had been included in the "machine recount" totals, Gore would have had a comfortable lead of more than 3,200 votes prior to the inclusion of late absentee votes, and would have been presumed the winner.

The NORC data will demonstrate the impact of the decision by Bush, who received the benefit of selectively hand-counted ballots, to do everything in his power to prevent the rest of Florida's ballots from being examined. Ultimately, this is the most important story that should be reported by the media consortium. Whether this story is given the prominence it deserves will be another test of which media outlets are "spinning" the results of NORC to Bush's advantage.


The Stench of Corruption

It is impossible to estimate what Gore's margin of victory would have been without the corruption of the electoral process by state and local election officials. There is no question that corruption and fraud played a large role in the theft of the election for Bush.

Bush's percentage of the absentee vote was completely anomalous when compared with previous elections. Much of this gain in absentee ballots can be attributed to the Florida Republican Party's encouraging its members to commit absentee vote fraud. In Florida, you cannot vote absentee as a matter of convenience. The state GOP, however, sent out a solicitation telling party members to do just that. The solicitation included a letter from Governor Jeb Bush that was printed over the Governor's seal, making it appear that fraudulently cast absentee ballots were in fact "officially" not fraudulent. As a result of this letter, the number of Republican absentee voters went up by tens of thousands of votes. Even assuming that most of these people would have gone to the polls and voted for Bush, the increase in absentee voters was so high that even if a small percentage of them had not shown up at the polls, Gore would have skated to victory.

In at least two counties (Seminole and Martin), elections officials permitted improperly filled out applications for absentee ballots to be altered by Republican Party operatives. This opportunity to alter signed federal election documents was made available only to the Republican Party, in counties controlled by Republicans. Democratic voters who submitted faulty absentee requests were not sent ballots, and in most cases were never even notified that their request for an absentee ballot would not be honored. Thanks to this corruption of Republican election officials, Bush gained thousands of extra votes.

The NORC data will also expose other likely instances of corruption. As noted above, many counties simply refused to count legal votes — and the worst offenders were counties that were dominated by Bush sympathizers. Duval County is an especially egregious example of Republican Party operatives who chose to ignore virtually every law concerning "write-in" voting in order to avoid counting Gore votes. These violations of Florida law are already obvious to those who know about the data that is already available — including the people who will be relying upon the NORC data itself. The media consortium has refused to fully report on this scandal "while waiting for the NORC data."

The Herald data also reveal startling evidence of fraud in the way that absentee ballots were counted in at least three heavily Republican Florida Panhandle counties. These three counties had impossibly low rates of overvoted ballots when compared to counties with similar voting systems and voting population. Even though reporters from the media consortium acknowledge the importance of these findings, these stories remain unreported because the consortium members "are waiting for the NORC data."

In Escambia County, one out of every nine absentee ballots was "duplicated" so the vote could be machine-counted. (Florida law permits duplication of a ballot that is damaged or otherwise unreadable by the tabulating machine, if the voter's intent can be determined.) After these "duplicated" ballots were counted, not one absentee voter (out of more than 21,500 absentee votes cast) overvoted by marking a ballot for exactly two candidates — but 296 absentee voters marked their ballots for three or more candidates. Only 11 absentee voters in Escambia County did not indicate a choice for President — an astonishingly low number.(10)

Bay and Santa Rosa Counties had impossibly low overvote totals. In Bay County (which also "duplicated" one out of every nine absentee ballots), out of more than 14,000 absentee ballots cast, only six were presented to the Herald as "overvoted" (and three of those ballots showed "voter intent" based on the Herald criteria). None of the six ballots had marks for more than two candidates. And in Santa Rosa County, only one of more than 7,000 absentee ballots was overvoted. Both counties duplicated their absentee ballots after "hand-counting" them using the "voter intent" criteria.

In contrast, heavily Republican Okaloosa County did not use a "voter intent" method to count their absentee ballots. Out of more than 14,000 absentee ballots cast in Okaloosa, the Herald examined 65 overvoted ballots. On 48 of these, voter intent could not be determined, and at least 17 of those overvoted ballots were marked for two candidates only. (In addition to being heavily Republican, Okaloosa is also Florida's best educated county, and has the highest percentage of Ph.D.s of any county in the United States.)

There is also evidence of possible widespread ballot tampering with poll-cast votes. A statewide analysis of election returns that compared uncounted ballots with results in both the Presidential and Senatorial races reveals a clear pattern of "errors" that can best be accounted for by fraud. But because confirmation of fraud requires an examination of the ballots themselves, the consortium members will not follow up on this evidence until the NORC data are released — if they follow up at all.

There is no doubt about corruption in the state's role in the election. Florida Secretary of State, Katherine Harris has the statutory obligation to see to it that all counties interpret and enforce election laws the same way. Nevertheless, despite considerable publicity that was given to the selective hand-counting of ballots in certain counties, Harris supported the efforts of the Bush lawyers to prevent the hand-counting of all the state's presidential ballots.

If there is any question regarding Harris' corruption, it can be answered by examining her actions in regard to late absentee ballots. Harris was asked to provide guidance, based on Florida law, regarding what constituted a legal late absentee ballot. This was a perfect opportunity for Harris to set specific guidelines regarding the absence of postmarks, ballots postmarked after election day, ballots without valid signatures, ballots without witnesses, and ballots that were received although no absentee ballot was requested. Such specific guidelines would have allowed Harris to fulfill her obligation to see to it that the law followed consistently throughout the state. The examination of Harris' computer hard drives showed that such specific guidelines were contemplated, and in the works. Yet Harris' response to the request for guidance on late absentee ballots provided no specific guidance. Harris told the counties that in order for these ballots to be opened, they must be "properly executed" without any specific criteria detailing what "properly executed" meant. Harris left open the door for Bush's strategy of selective counting of late absentee ballots, and the Bush camp's abhorrent strategy of accusing Gore of attempting to suppress the vote of military personnel.


What The Reporting on NORC Will Show

Ultimately, the real information that will be derived from the NORC data will be the nature and extent of media bias. We already have sufficient evidence to know what the NORC data itself will show. The real question at this point is how that evidence will be reported.

To date, the consortium's reporting on election issues shows a clear tendency to do whatever it can to avoid questioning the legitimacy of Bush. The members of the consortium reported the results of the Herald examinations of the undervote and overvote verbatim, without any attempt to confirm or challenge the conclusions by analyzing the Herald data or the biased assumptions that the Herald used in reaching its conclusions. And the members of the consortium have consistently isolated each and every instance of corruption of the electoral process, never making the effort to consider the cumulative impact of the corruption of the election as a whole.

Perhaps the most notable example of this was the New York Times report on the late absentee ballots. In its story, the Times noted that the estimated 292-vote advantage that the illegally counted ballots gave to Bush would be insufficient for Gore to overcome Bush's 537 certified vote victory. No mention was ever made of the fact that the Florida Supreme Court had ordered that votes found in the Dade and Palm Beach County recounts should be included in the final totals, and that Gore's net gain of more than 300 votes from those counties would have changed the results of the election and given Gore a razor thin victory had the Supreme Court majority not stopped the counting of Florida ballots. Nor was any mention made of various other ballot studies that, when combined with the Times estimates, would have revealed a Gore victory.

It is this tendency to view each election controversy in isolation that demonstrates the bias of the media, and of the media consortium in general. Fortunately, regardless of the spin placed on the NORC data by the media consortium, the NORC data itself will be made available to the public. History will eventually record that this election was stolen. If the news media spin the NORC data in ways that do not confirm the illegitimacy of Bush, history will also confirm that the media played a key role in the destruction of the bedrock principle of "consent of the governed" on which this nation was founded.


  1. Petula Dvorak and Allan Lengel, "Levy Family Reflects on How 'World Changed'," Washington Post, October 29, 2001.

  2. Anabelle de Gale, Lila Arzua, and Curtis Morgan, "If the Vote Were Flawless..." Miami Herald, December 3, 2000.

  3. "Who Won? A Herald Special Report." The Miami Herald index of articles related to their review of the undervotes.

  4. Geoff Dougherty, "Spoiled ballots favored Gore." Miami Herald, January 28, 2001.

  5. See "Lost" Votes and a Stolen Election.

  6. For a complete explanation of the relevant laws, see "Lost" Votes and a Stolen Election.

  7. See "Lost" Votes and a Stolen Election.

  8. David Damron and Roger Roy, "Mangled absentee ballots resurrected," Orlando Sentinel, May 7, 2001 (not available online)

  9. David Barstow and Don van Natta Jr., "How Bush Took Florida: Mining the Overseas Absentee Vote." New York Times, 15 July 2001

  10. For a statistical analysis and discussion of the implications, see "The Evidence for Ballot Tampering in Escambia County, Florida."

Copyright © 2003 by Paul Lukasiak
Permission is granted to link to this page or to reproduce the contents, as long as the author is credited.