The first thing one would do to investigate a suspicious overvote is talk to the voters, and of course the Democratic Party did so. ABC's "Nightline" broadcast a special on Duval's overvotes March 28, in which I saw again the exculpatory evidence first reported in an 11/16 New York Times article on Duval's overvotes: Democratic Party officials putting the blame on voter error. On the Nightline broadcast, the manager of Gore's campaign confirmed the evidence of voter confusion in Duval. In her words, "I talked with the voters." But actual voter error does not rule out tampering. The two are not mutually exclusive.

Voter error most certainly happens. There were 3,000 overvotes in Palm Beach in 1996, Duval Elections Supervisor John Stafford also said that 3-4,000 is normal. Add to that Duval's and Palm Beach's confusing ballots, and the large turnout of first-time voters, and quite of few of the overvotes probably were innocent. Innocent overvote is especially likely in Duval, which used a 2-page, so-called "broken ballot" for the presidential race. There were 4,000 (maybe more) broken-ballot overvotes in optiscan counties, which I believe to be innocent overvotes. (That figure came from the Orlando-Sentinel study of 15 "most error prone" counties. That study covered about 250,000 votes, not all of which were from broken ballot counties. There may be additional broken-ballot overvotes in counties not included in this study.) So broken ballot does cause a fair number of overvotes.

In fact, the occurrence of real voter confusion may answer of the most perplexing questions I've had: Why isn't the Democratic Party investigating this? It's not because fraud never dawned on them. In the first days following the election, Rep. Robert Wexler in Palm Beach, Rep. Corrine Brown in Duval and Democratic Party Vice-President Gus Garcia in Miami-Dade expressed suspicions about ballot tampering. This may be the answer: They were thrown off the track by the coincidence of actual voter confusion.

(Lucky Republicans. Or maybe creating voter error through confusing ballots was part of the plot to begin with, to create cover. The perpetrators knew that broken ballots cause overvotes, and that many voters think you have to vote for vice-president, and might be tripped up by a butterfly ballot. There's something fishy about how the ballot designs were picked in both Duval and Palm Beach. The Secretary of State (Kathy) officially recommended a two-page ballot design, and Duval, a Republican-run county, followed directions. Also, Duval had the infamous sample ballot printed in the Jacksonville newspaper, which inexplicably instructed voters to "vote every page." Even odder, the sample printed in the paper was a single-page ballot, looking nothing like the 2-page ballot actually used in Duval. Where did the County even get the copy to give to the paper, for a sample ballot that looked nothing like the real thing? The inaccurate sample ballot prevented Democratic Party workers from alerting voters to the broken ballot trap. As for Palm Beach's truly unfair butterfly ballot (I was furious that the Florida courts gave such short-shrift to the seniors who voted for Buchanan by mistake), I have heard that Palm Beach's Democratic Party officials swear they were never shown the ballot design in advance.)

But evidence of real voter confusion doesn't mean the case is closed. The Buchanan confusion and/or vice-president error (if they did happen) account for only half of Palm Beach's overvotes. And while we do not yet have the detailed overvote data for Duval, Elections Supervisor Stafford reported that, in one representative precinct, only about half of the overvotes had punches on both pages (fit the broken-ballot error). Most perplexing of all, there is no voter confusion story to explain the 18,000 overvotes in Miami-Dade, nor the 1,000 in the rest of the suspicious counties (Broward, 8,000; Hillsborough, 5,000; Pinellas, 4,000 and Pasco, 2,000).