Forget Florida. What about California? With all attention focused on the Florida recount, what's being overlooked is that California has at least 1 million and possibly as many as 1.4 million absentee and "unprocessed" ballots yet to be counted. This huge block far exceeds all other absentee and similar uncounted ballots in the rest of the nation combined. The uncounted California votes won't change the allocation of the state's 54 electoral college seats to Al Gore because he won California by 1.2 million of the votes counted so far. But they could swing the national popular vote count to Bush, making him the winner as the people's choice overall.
Right now, in the national popular vote, Gore leads by slightly more than 200,000. (In the closest modern election in popular-vote terms, John Kennedy defeated Richard Nixon by 118,574 votes.) But the popular vote may be turned on its head when the uncounted California ballots are added into the equation. In recent presidential elections, about 60% of California absentee votes have been Republican. Should George W. Bush receive about that same percentage of uncounted California ballots, he will surpass Al Gore in the popular vote. Not by much--perhaps by even less than Kennedy's 1960 margin over Nixon--but by enough to be the winner.
Should George Bush turn out to be the winner in the popular vote, disputes about Florida will dramatically change in tenor. Right now, Al Gore's strongest argument is that questions about Florida vote accuracy are causing the people's will to be thwarted, because he won the popular vote and fundamentally deserves to become president. But there is a strong chance Gore has not won the popular vote--at the least, right now we have no idea who the popular-vote winner is. This should be listed everywhere as "too close to call."
So what is California doing about its million-plus uncounted ballots? Dragging its feet. In an announcement totally missed by the national media (only the Los Angeles Times reported it), California Secretary of State Bill Jones said yesterday that because state law specifies the final count does not have to be certified until December 5, it may be almost a month before California reveals its absentee ballot totals. For all the outrage (real and pretend) now being directed toward Florida, it's astonishing that California has said it may take up to a month to disclose the totals that would let us know who won the popular vote--yet this has attracted no attention at all.
Add to this that Florida, a state with a Republican governor, has instantly focused its full resources on recounts and is concentrating the hand-recounts in those Democratic counties likely to help Gore. (No hand-recounts in Republican counties, where Bush might pick up extra votes, are currently planned.) In contrast California, a state with a Democratic governor, is taking its sweet time about finishing its initial count. And California state officials know full well that the longer they hold back information about what the real popular-vote outcome is, the longer Al Gore can press his claim for special treatment with the seeming backing of the popular majority in his favor.
Now for the unnoticed point on Gore's side: Why did almost every county in the Florida recount find more votes for Gore? If you watched www.cnn.com or any of the websites that were posting Florida recount totals throughout the day yesterday, you were struck by the fact that the proportion of new Gore votes was usually parallel to the proportion of counties reporting. That is, 10% of the counties would report, and the Bush-Gore gap would shrink by 10 percent. Such proportions held throughout the day. Yet if what the recount was finding were simply human errors, such errors should have been randomly distributed--there's no underlying statistical reason each new reporting county would bring about the same fraction of gain to Gore. It turns out (assuming the Associated Press tally is correct) that 44 of the 67 Florida counties gave net added votes to Gore on the recount, while 12 had no change and 11 gave net additions to Bush. If what happened were standard human errors, in this split-down-the-middle race there should have been roughly as many counties adding to Bush as to Gore. There weren't.
Also, several counties gave (relatively, considering the overall closeness) large additions to Gore, while only one gave a large addition to Bush. Palm Beach County on its first recount (another will happen tomorrow) netted Gore 642 more votes. Pinellas County netted Gore 478. Duval County netted Gore 166. Bush's only three-digit gain came in Martin County, which netted him 105 votes; all other Bush gains were slight.
Maybe the fact that most of the recount favored Gore shows Florida officials went out of their way to give Gore ballots the benefit of the doubt on second viewing. But it surely suggests, also, that Florida officials as a group disfavored Gore ballots the first time around. The swing is small enough (a 1,457-vote gain for Gore, out of almost six million cast) that no one could claim conspiracy. But the fact that the recount consistently turned up more for Gore than for Bush seems to show that on election night, there was at least some subtle favoritism toward Bush when Florida officials made the first counts.
This seems a far more substantive objection to the Florida vote than a flawed Palm Beach ballot design (confusing perhaps, but a design specifically ordered by a Democratic official!) or the problem of people who punched their ballots twice (too bad, but there were discarded double-punched ballots all over the nation for both candidates, and no one's proposing re-votes for the entire nation) or people who might have meant to vote for Gore but didn't. (The Constitution does not allow mind-reading of voters; we might as well complain that people who forgot to vote should now get a second chance if they "meant" to vote.)
The steady drip-drip-drip of adding votes for Gore shows there may have been small but real systematic Bush favoritism in Florida. This, not Palm Beach, should be what Florida complaints focus on, though the logic leads to this: Hand-recount every county and then accept the result.