(Please see update below for important corrections.) Lance deHaven-Smith is a professor in the Askew School of Public Administration and Policy at Florida State University. He has long been deeply immersed in the politics of Florida, and has just published a book through University Press of Florida entitled
"The Battle for Florida:An Annotated Compendium Of Materials From The 2000 Presidential Election." (
Amazon link)
He has an interview on the book in Florida State's Research in Review Magazine that you really have got to read. It's all about the disaster of 2000, and the man is full of insights and wise words (and by the way, terrified that we are in the late stages of losing our democracy).
But the real zinger is his comment on overvotes.
There were 175,000 votes overall that were so-called "spoiled ballots." About two-thirds of the spoiled ballots were over-votes; many or most of them would have been write-in over-votes, where people had punched and written in a candidate's name. And nobody looked at this, not even the Florida Supreme Court in the last decision it made requiring a statewide recount...The write-in over-votes have really not gotten much attention.
Yes, we were all focussed on the "undervotes," caused by hanging chads, and the like. I've never heard anyone analyzing the "overvotes," ballots with more than one indication of a presidential choice. I had always assumed that there were useless, because they contained two different votes for two different candidates. I was wrong:
Those votes are not ambiguous. When you see Gore picked and then Gore written in, there's not a question in your mind who this person was voting for. When you go through those, they're unambiguous: Bush got some of those votes, but they were overwhelmingly for Gore. For example, in an analysis of the 2.7 million votes that had been cast in Florida's eight largest counties, The Washington Post found that Gore's name was punched on 46,000 of the over-vote ballots it, while Bush's name was marked on only 17,000.
So the 'overvotes' were actually more clear than a normal vote -- they were cases in which a voter punched the card and then wrote the selected candidate's name in as well, just to be sure. And there were twice as many of them as of the undervotes. And they went overwhelmingly for Gore.
A total of 46,000 uncounted overvotes unambiguously for Gore and 17,000 of them unambiguously for Bush. If I understand this correctly, this means that Gore won the state not by a few hundred or a thousand votes, but by approximately 30,000 votes. It wasn't even close. (Of course, that is without accounting for the butterfly ballot or the African-American non-felons illegally purged from the voter list.)
Why did the overvotes overwhelmingly favor Gore? deHaven-Smith points out that they were overwhelmingly African Americans, and offers the following interpretation:
One of the things I found that hadn't been reported anywhere is, if you look at where those votes occurred, they were in predominantly black precincts. And (when you look at) the history of black voting in Florida, these are people that have been disenfranchised, intimidated. In the history of the early 20th century, black votes would be thrown out on technicalities, like they would use an X instead of a check mark.
So you can understand why African Americans would be so careful, checking off Gore's name on the list of candidates and also writing Gore's name in the space for write-in votes. But because of the way the vote-counting machines work, this had the opposite effect: the machines threw out their ballots.
This could be the most heartbreaking thing about Florida 2000: African-American voters, who were used to getting screwed out of their votes, took measures to be absolutely sure that their votes counted -- and as a result got screwed out of their votes.
But at least we can say to everyone: Not only did Gore win Florida and thus the election, but it wasn't even close.
UPDATE: A number of commenters have questioned whether the 46,000 Gore and 17,000 Bush overvotes in question are valid overvotes in which a voter made a consistent and redundant choice (call this option I), or invalid overvotes, in which a voter punched choices for two different presidential candidates (call this option II). I was pretty sure that it was the former, given that in the interview the 46,000/17,000 figure comes in the same breath as the discussion of the valid redundant overvotes. But on rereading the interview (yet again) I did see that there was some ambiguity about this in Dehaven-Smith's wording. So I went to Lexis-Nexis to find the original Washington Post article he's referring to (it's from Saturday, January 27, 2001, on page A1. I couldn't find an online copy.). As you can see from the date, it is from _before_ the media consortium hand-recount analysis. In fact, the article draws entirely on computer records from punch-hole readings:
A consortium of media organizations, including The Washington Post, will next month begin an examination of all votes statewide that did not register when passed through an automatic counting machine. A similar, separate count by the Miami Herald is underway.
Unlike those efforts, which will be based on a hand examination of the physical ballots, The Post analysis of the eight counties was based on an examination of the computer record made for each ballot when it passed through the automatic counting machines. These mechanical readers shine light through each card to detect which holes have been punched out.
Although the files -- which have no voter identification -- include the details of every mark detected by the equipment, revealing the exact pattern on each ballot, they do not contain information about marks that the machines could not read. This includes such such things as dimples or partially detached chads that were the subject of so much attention.
Thus, this article, from which the 46,000/17,000 figures come, is _only_ about multiple-punch overvotes, and hence it is all about invalid, multiple presidential candidate overvotes. I was wrong: Option II is correct, not Option I. I'm sorry to have misinterpreted the figures that way, and I'm very grateful to the alert Kossacks who prodded me to dig into it.
So, the correct interpretation is as follows: Counting all of the valid, redundant overvotes that indicated an unambiguous choice for president would give a narrow victory to Gore. Analyzing the invalid, multiple candidate overvotes shows a pattern that suggests probably several thousand additional people were trying to vote for Gore and made an error. Many thousands of voters punched for Gore and one other presidential candidate, vastly more than punched for Bush and one other candidate. In addition:
In the eight counties examined, people who overvoted for president but cast a valid vote in the U.S. Senate race favored the Democrat 70 to 24 percent.
So this finding, which is still new to me and not well known, is of the same type as the butterfly ballot: Strong evidence that the number of voters attempting to vote for Gore was considerably larger than the numbers attempting to vote for Bush. In fact, this overvote analysis shows that the butterfly ballot itself is relatively small potatoes, compared to the many thousands of net Gore votes apparently lost due to overvotes.
UPDATE2. First, a special vote of thanks to shumard for an avalanche of facts and clarifications. Second, I send Prof. deHaven-Smith email, asking if he could clarify the above questions, and he very kindly reponded with a great, punchy note, which confirms the point of the update above and adds some additional insight. I don't think he would mind if I reproduce it here:
Thank you for your interest in my research on the 2000 election and the interview in the FSU research magazine. Before going into some of the technical details, let me answer your question directly. Commenters are correct that a good portion of the “write-in overvotes” were truly spoiled ballots, so Gore’s margin of victory would not be as large as you have estimated. If all of the votes had been counted in Florida, Gore would definitely have won, but his margin of victory would probably have been less than 10,000 votes and could have been as low as 100-200, depending on the criteria used in deciding how to count ballots with various types of ambiguities, e.g., those with partially punched chads, write-ins, etc.
A number of studies have been conducted of the “spoiled ballots” in Florida. The most thorough was by the National Opinion Research Center (NORC) at the University of Chicago. It examined all of the 175,010 spoiled ballots that were excluded from the vote tabulations in Florida in 2000. Ballots were examined by multiple coders to assure reliability, and a dataset (available on NORC’s website) was constructed that included each coder’s judgment on each ballot. NORC also provided a program for applying different criteria and calculating the results. If the most restrictive criteria are used, Gore’s margin of victory is 158 votes. By “most restrictive,” I mean ballots are counted only if the voter’s intention is totally unambiguous (no partially punched chads, no unclear write-ins, etc.).
Please note that this estimate of Gore’s margin is very conservative. It only counts ballots that are totally unambiguous. Also, it deals only with votes as they were actually cast – it does not address the votes lost by Gore because of the butterfly ballot in Palm Beach County or the faulty ballot instructions in Duval and Gadsden Counties. Nor does it take into account the effects of Florida’s flawed program of felon disenfranchisement, which excluded many eligible voters from casting ballots and also created delays and confusion at polling places. During the election controversy, most observers acknowledged that more voters had gone to the polls to vote for Gore than for Bush, but it was unclear whether Gore had actually received more legally valid votes. We now know that he did.
Today, this would be widely recognized if NORC had published a report on its findings, but instead it chose simply to release the data and let everyone make sense of it on their own. My assumption is that NORC and the research sponsors did not issue a report because they did not want to embarrass President Bush. The study was concluded just after 9-11-01.
Emphasis mine. Note that the public mostly still do not know any of this, all due to the deference the press showed Bush post 9/11. That's important to remember.
The initial misunderstanding in this post was mine, not the professor's. He is careful and precise, eloquent and full of insight. I suspect that the book will be the bible on the disaster of 2000.