[This was originally a posting to a mailing list, part of a discussion we were having at the time, so you will see some references here to other posts or people on the list. I think this essay can still stand on its own.] Date: Fri, 29 Dec 2000 19:14:55 -0500 From: Ofer Inbar To: [mailing list] Subject: who won (was Re: Politica) On Thu, Nov 23, 2000 at 05:09:25PM -0500, [another list subscriber] wrote: > But really... I personally don't find it "clear" at all that the > election was won by Al Gore. Last I heard Bush was still 930 votes > ahead. I think it's clear that the country was by and large divided > on who is the better candidate. The election was won by Al Gore. True, it was a very close election, perhaps the closest we've ever had. Almost a statistical dead heat. But not quite. I. The popular vote Yes, Al Gore won the popular vote by a narrow margin. That margin was not quite narrow enough to be within what I believe would be the margin of error for a nationwide election, and it is also likely that most of the "error" would skew the results towards Bush, so it is pretty clear that Gore beat Bush in the popular vote. However, personally, I think this is meaningless. The election was carried out under the electoral college rules, which split the nation up into "districts" known as states, and the campaigns dealt with each such district as a separate entity. It is unclear what would have been different if the rules had been different. It is certain that "turnout competition" between states, partly driven by these district election rules, can twist the "popular vote" results in strange ways. For example, before the election, many people were predicting that we might have the opposite of what we got: Gore wins the electoral college, but Bush wins the popular vote. Why? Because polls showed Bush slightly ahead (although well within the margin of error for the small samples those polls used). Electoral College calculations showed that most of Bush's EC votes would come from states where he would win very wide margins, while most of Gore's would come from states where the result would be much closer. And this is, indeed, the classic case of how a districted result might differ from a popular result. Say you have only two states, one with 1,000 people and one with 2,000 people. Say candidate X wins state 1 on a vote of 800/200, but candidate Y wins state 2 on a vote of 1050/950. Y wins the election, because state 2 is twice as large. But X was only 100 votes behind in state 2, and 600 votes ahead in state 1, for a significant popular vote win. What actually happened, however, was that precisely *because* Gore's victory depended on several very close states whose populations happened to be very high, the voter turnout in those states was particularly high. A Bush voter in Idaho might easily stay home and not care, and Ralph Nader got one of his best results in the Bush stronghold of Montana. But a Gore voter in California went out and voted, with the feeling that the result of the election rested on them. It was the high turnout in large, democratic-leaning states like California that gave Gore his popular vote win. Overall voter turnout in this election was about 50%-51%, but turnout in California was at a record high, above 70%. Florida, Minnesota, and other "battleground states" either set new turnout records, or came close. California's above-average turnout alone probably netted Gore a gain of a few hundred thousand, and without that the popular vote would be within the margin of error for a nationawide vote count. Bush himself may have caused his own deficit here, by visiting California several times shortly before the election, giving the false impression that the state was up for grabs, and stimulating higher Democratic turnout. [ http://www.calvoter.org/cvfnews/cvfnews110800.html http://csf.colorado.edu/mail/pfvs/2000/msg05713.html ] In fact, one of the reasons we have district elections is so that turnout won't distort district representation. If, on election day, there's a snowstorm over Minnesota and bright skies in Florida, that doesn't mean Florida deserves more than its share of representation in the election results. A district system in a way makes it so that people who don't vote are also represented, by letting their fellow citizens from the same district speak for them. But this same system that makes sure states aren't helped or harmed by varying voter turnout, also itself encourages skewed voter turnout. This is OK, as long as we ignore the popular vote results. Of course, neither candidate even bothered to campaign in states such as Idaho and Montana. If the election were running under different rules, they probably would have. There was a much more striking national divide than the one between the populated and unpopulated states, though: the divide between city and country. In cities with populations 1 million & above, Gore won 2/3 of the vote! In small towns and villages, Bush got 60%! These are not close results. If they were both simply tageting the popular vote, Gore would have campaigned in the cities and Bush in the country. The electoral college system forced them to campaign for districts, rather than demographics. Gore couldn't ignore small towns in Florida, and Bush couldn't ignore big cities, because in a close state, a few more votes might make the difference district-wide, even if those votes come from a group that gave you little support. Bush had no chance with blacks, but that 10% of the black vote he did get may have swung a couple of states his way, so he could not ignore them. If the election were targeting the popular vote, Bush could safely ignore blacks. So, OK, the popular vote is meaningless. The fact that Gore won it doesn't tell us anything useful. But, Bush certainly did not win the popular vote, either. To the extent that the popular vote means anything at all, it doesn't point to Bush being the winner, that's for sure. II. The district vote A much more meaningful result, however, is that Gore also won the *districted* vote. In a district system, each district is supposed to be represented by a number of votes proportional to its population. This is consistent with all of the benefits of a district system, and all of the strategic considerations in campaigning for districts still apply. But our electoral college system has a built in distortion, originally meant to protect the slave states (who artificially lowered their population by insisting that the constitution undercount slaves and Indians). The distortion is to add senators to a state's vote count. Now, even if we just counted Representatives, the very smallest states would still be overrepresented somewhat, because a state whose population is enough for 1/2 a Representative still gets a whole one. But adding Senators to the count greatly increases the level of distortion to extreme levels. For example, it takes more than three Massachusetts voters to equal the representation of one Wyoming voter. [ MA population is about 6,175,000, with 12 electoral votes that's about 514,500 people per EV. WY population is about 479,600, with 3 electoral votes that's about 159,900 people per EV. See http://www.census.gov/population/estimates/state/ ] Bush won almost all of the extremely overrepresented states, and his victory *depends* on this overrepresentation to reach his narrow victory margin of 271 to 267. If you were to decrease the overrepresentation factor just a little bit, from two senators to one and a half senators, we'd have a tie of 256 to 256. If you were to decrease it to only 1 senator, Gore gets a victory. And without the senators at all, Gore wins 225 to 211, a decent and comfortable victory. Some people claim, and rightly so, that the states don't make for very good districts. Some, like Florida and California and Texas, are just way too big. New England is represented in a very different manner than New York, due to its being 6 districts instead of one. It's not clear how exactly this uneven division of the country affects things (if we don't consider Senatorial electoral votes), but it's probably not the best way to district the county. A much better way would be Congressional districts, probably, since each one theoretically covers about the same number of people. Two states, Maine and Nebraska, actually do this: the two "Senatorial" electoral votes go to the candidate who won the state, and each congressional district picks an elector based on who won that district. I heard on NPR that they went through all of the precint results in the country and figured out what candidate each congressional district voted for. According to them, if all states used the Maine/Nebraska method, the electoral college vote would have been... 271 to 267. Exactly the same. Of course, none of the statewide results would be different under this system, so this boils down once again to Bush depending on the "Senatorial" electoral college votes for his victory. If only congressional districts voted, it would be Gore by 225 to 211. [Although if we really were running the electoral college like that, the Greens might have decided to campaign strategically, and could have picked up a few of Gore's districts - and could then instruct their electors to vote for Gore in exchange for concessions, if they had enough of a swing vote to determine the victor.] The bottom line: Even if you include Florida in Bush's column, the total population of the states that went for Gore exceeds the total populations of the states that went for Bush. In a districted election, this is the equivalent of the morally powerful "popular vote". It is on *this* basis that I can say that Gore did, indeed, "win the popular vote" in a meaningful way. [And there's also the fact that NH was almost certainly swung to Bush by Nader/Gore voters choosing to vote Nader over Gore, and those voters almost certainly would prefer Gore to be president rather than Bush - I'm not talking about all of NH's Nader voters, only the 1/3 of them that would have been necessary to give Gore the state.] III. The Florida vote But lets forget all that for a moment, and play by the rules. As we all know by now, Florida voted for Bush in the electoral college, so it might seem odd for me to say things like "even if you include Florida in Bush's column". But keep in mind that at the time Erin posted this question, Florida's results were still very much in question. The reason they were in question is that, in fact, the official result in Florida was a tie. A 930 vote margin is about 0.015% of votes cast. A 537 vote margin is about 0.008% of votes cast. Even if you postulate the smallest of all plausible margins of error, Florida's count was within that margin. The only other state that ended up a statistical tie was New Mexico. So really, those 25 votes Bush got from Florida, and the 5 votes Gore got from New Mexico, were not based on valid wins, they were based on technicalities and sheer chance. What if these ties were counted as the ties that they actually were? Well, take 25+5 EC votes, and split them up evenly, awarding 15 to Bush and 15 to Gore, and what do you get? Bush 261, Gore 277. Oh, right, but we were gonna play by the rules. You can't split up a state's vote if its law didn't say you could. Even if it's a tie, you have to figure out a winner and a loser, and vote accordingly. So who, if anyone, won Florida? III.1. The Butterfly Ballot Well, for one thing, it's hard to dispute the fact that more Floridians went to the polls to vote for Gore than to vote for Bush, and the butterfly ballot severely skewed the results. Buchanan got 3K votes in Palm Beach where he should've gotten a few hundred at most, and there were 19,000 ballots rejected for double-voting, several times as many as had been rejected for double-voting in the previous election. The Republicans tried a few truly stupid ways of twisting the statistics to dispute this. They cited increased rates of third party registration, but obviously most of those new third party registered voters did so well before anyone knew Pat Buchanan was going to be the Reform Party candidate, and most of them weren't even Reform Party anyway. Those were not Buchanan voters. The Republicans also tried to pretend that the rejected ballot rate wasn't all that much higher than it was in the previous election. They did this by comparing the number of presidential double-votes in 2000, with the *total* number of ballots rejected for all reasons in the previous election. If you look at real, valid numbers, it's obvious that Buchanan got an order of magnitude more votes in Palm Beach than he should have. It's also obvious that the rate of double-votes multiplied. And, guess what, we have a very likely, certainly plausible, explanation for these two otherwise very strange facts: Gore voters confused by the butterfly ballot. No other plausible explanation has been proposed by anyone. This one is almost certainly it. This is further borne out by the exit polls, which are more in line with the hypothetical statistically corrected Palm Beach results than they are with the "real" results. [Note: At the time I wrote this essay, I was not aware of all of the facts surrounding the butterfly ballot issue. It is unfair to pin all of it on "confused" voters, although that was certainly a factor. Having seen some actual pictures of how these ballots looked in the voting machines, I now know that in many cases the arrows on the card did not line up very well with the actual punch holes on the machine, and because of the butterfly design, the markings on the card were close enough such that the Gore arrow in many cases appeared equidistantly between the Gore & Buchanan marks punch holes. Voters could make this mistake easily, even if they were smart and well educated. Since the Bush marking was first at the top, and his punch also first at the top, it did not lead to many errors. Also, as it turns out, there were thousands of complaints from people who *did* realize their mistake and asked for another ballot, or asked for help, and were denied. This was not legal - a voter has a right to assitance with the voting machinery, and the right to invalidate their ballot and cast another one if they've made a mistake. The polls were poorly staffed and many poll workers were not properly trained and lacked resources. Oh, and by the way, it seems that Florida law already had a provision that said punch holes must be to the right of candidate names, listed in a single column. The butterfly ballot was illegal. It is unfair to blame the Gore/Buchanan confusion on the voters. The system failed them. ] Without the butterfly ballot, Gore would probably have had over 20,000 more votes than he did. Even if we're extremely cautions about this prediction and cut it by 1/10, 2,000 extra votes for Gore would have been more than enough to put the margin in his favor. III.2. Absentee Ballots There was no counter-example anywhere in the state of Florida that would have skewed any significant number of votes towards Bush to overcome that butterfly ballot factor. If there were, I'm sure the Republicans would have tried to make a big deal out of it. They did try to make a big deal out of overseas absentee votes not counted due to lack of postmarks, but there, again, they were playing with statistics. They kept citing the total number of absentee ballots received but not counted, implying that all of these were lost Bush votes. First of all, while Bush did win the overseas absentee vote by a good margin, it wasn't 100%. I think it was something like 60/40. So if 10,000 absentee ballots were rejected unfairly, that wasn't a loss of 10,000 for Bush, it was a loss of 2,000 for Bush. But that's not all: The implication that all, or even most, of these votes were rejected for missing postmarks, is extremely misleading. I heard an interview on NPR with a county canvassing board member about the rejected overseas absentee ballots in his county. By far the most common reasons for rejections were: - Voter isn't registered - Voter didn't request an absentee ballot (i.e. someone got a hold of someone else's ballot and sent it in) - Duplicate votes from the same person These are all perfectly legitimate reasons to reject a vote. Postmark problems accounted for something like 10% of overseas absentee ballot rejections in the county in question, and even among those, the majority were rejected because they were postmarked *too late*, not because they lacked a postmark. In other words, they probably came from people who voted after the deadline. But let's be very generous to Bush. Let's say 10% of overseas absentee ballots statewide were rejected due to missing postmarks, and let's say Bush won the overseas absentee vote in Florida by a margin of 75%/25%. Even so, the total Bush loss from missing postmarks is a mere 50 votes per 1,000 ballots rejected. This rate is clearly not enough to match up to the butterfly ballot factor. And if you accept the argument that voters who had their vote invalidated by a trick of the system, should be counted, then you really have to count both sets. You can't have it one way on the postmark problems, and the other way on the butterfly ballot. That's clearly not fair. [Note that I'm talking specifically about *overseas* absentee ballots. The Republicans liked to say that there were 23,000 absentee ballots in Florida this year, but most of those were already received and counted on election day, and included in the first count. Only about 3,000 overseas absentee ballots arrived after election day, of which less than 1,000 came from military voters.] There were, however, other problems with absentee ballots that more than made up for Bush's loss on the postmark issue. Absentee ballot fraud is in fact a time-honored tradition all over Florida. Several years ago, an egregious case of absentee ballot fraud, which resulted in an overturned election in Miami (Xavier Suarez for mayor), prompted Florida to pass a new law: It is expressly illegal for anyone other than the voter or their legal guardian to in any way alter an absentee ballot application after the voter has filled it out. But all over Florida, Republicans, and for the most part *only* Republicans, did just that. The most egregious case was in Seminole County, where the Republican party realized that they had goofed by failing to include voter ID numbers in the absentee ballot applications they had mailed out to their registered voter lists. Normally, that would mean they'd have to send out new applications to their whole list. But this was not a normal situation, because the county officials were Republican party members. What they did instead was to let a team of Republican party campaign operatives sit in the county office for ten days, and correct already filed absentee ballot applications, *unsupervised*. In the meantime, faulty applications sent in by non-Republicans just sat neglected. The Republicans actually ended up benefiting from their own mistake! Rather than losing some voters due to incomplete applications (remember, those voters *could* have read the rules and filled in their own voter ID numbers!), they actually gained votes. We'll never really know how many, because we don't know how many of the 4,700-odd ballots they corrected were ones that would have been fine if it weren't for the party's earlier goof. But it's very likely that this number is larger than the number of overseas absentee votes the Republicans lost to missing postmarks. And what's more, what the Republicans did was very clearly illegal. In any other election, the courts would have been much more likely to throw those ballots out. All across the state, Republican county officials gave Republican party members undue access to absentee ballot applications, while Democrats did not get such access. In some cases, they even allowed party members to take the applications home with them, and return them later! The courts ruled that while these actions were in fact illegal, there was no evidence that there was fraudulent intent, so they would not throw out the votes. In other words, well-intentioned corruption is fine. And this from the same Florida Supreme Court that everyone was claiming was partisan and biased towards the Democrats! It wasn't biased towards the Democrats, it was a fair court with a strong bias towards counting as many votes as could be counted, which is certainly a defensible position. In this particular case, that bias worked in favor of the Republicans. Without it, Bush could not have won - if the courts had thrown out all absentee ballots in Seminole County, certainly also a defensible position, then Bush would have lost about 10,000 votes and Gore would only have lost about 5,000, in this heavily Republican county. III.3. Corrupt Certification Oh, but I almost forget, we were going to play by the rules. Sure, those confused voters in Palm Beach intended to vote for Gore, but they didn't, and we can't count them. The absentee ballot irregularities are hard to pin down, but they do also look to favor Bush overall, and in any case don't come close to cancelling out the butterfly ballot. So, we're back to a tie in Florida. Well, it doesn't matter if there's no winner, as long as you can get the right official to certify that you were the winner. Imagine, for a moment, if the secretary of state of Florida weren't Katharine Harris. Let's not put a hypothetical Democrat in her place. Instead, let's put an honest, impartial person with no preference for Bush or Gore. Or even, a Republican, one who supports Bush, but one who does not display quite the level of brash corruption that Harris did. What would have happened? If it weren't for Harris, the recounts the Democrats requested would all have happened. Harris singlehandedly supplied the pretext on which the Republicans based their court challenges and were able to run out the clock and prevent the recounts. There were plenty of arguments about whether Harris was acting properly by refusing to accept any recount results after November 14th, and whether the courts could force her to do so, but nobody disputes the fact that had she chosen to, she was certainly *allowed* to let the recounts go on, and include them all in the certified total. So, it was her choice not to do so, that led to the court battles. In fact, a plainly obvious reading of Florida law makes it clear that the November 14th deadline was never meant to apply to manual recount results anyway. The law in question refers both to "certification" and "the official certification". It says that counties should turn their numbers in by 7 days after the election, and the state certify the total. But then it also says that the "official certification" should include both final absentee ballot counts and the results of manul recounts. Obviously the "official certification" can't be the one that happens 7 days after election, because not all of the absentee ballots are in by then. The law is consistent about its use of "certification" vs. "the official certification", and no explicit deadline is set for the official certification. Florida law also allows manual recounts to be requested up to 6 days after election day, which would be clearly ridiculous if all recount results needed to be in 7 days after election day. All of these arguments were made by David Boies during the first Florida Supreme Court recount case, and e court accepted them all, but the Republicans did not, and this later became fodder for Rehnquist. Basically, this was Harris making a choice that may or may not have been a valid one for her to make, but that sure as hell was not the choice she *had* to make. And, of course, the manual recounts from Seminole and Voluisa counties, both of which were in before the November 14th deadline, were never disputed by anyone after they were certified. So clearly there was no legal case here against manual recounts in general, the issue was only about Harris' discretion to reject recounts that came in after the supposed deadline. One thing we didn't hear much of, is any truly convincing argument that claims Harris did the right thing. Although the Republicans tried to imply that in their media blitz, their arguments were never very compelling. As one person on NPR put it, "a secretary of state refusing to accept vote counts is like a secretary of the treasury refusing to accept revenue." Imagine if, after tax collection, a corporation discovered that they'd made a mistake, and actually owed the state a bit more money... and then the secretary of the treasury told them, "too bad, you missed the deadline, we're not gonna take your money." That's the equivalent of what Katherine Harris chose to do, in blatant violation of the moral duties of her office. Why? Harris said she'd accept late results in case of natural disaster, or machine malfunctions, but not accept manual recounts conducted simply because a county thought their voting machinery hadn't accurately counted all the votes and this may have changed the result of the election. That doesn't take us very far, since it's merely a claim that the reasons for recounts weren't quite good enough to compel her to accept them. That presumes that the reasonable default action would be to reject recount results, which is preposterous. What we need is some good reason for her to reject the recounts, and this certainly isn't. Going by this argument alone, no harm would have been done by accepting them. After the fact, two other arguments were brought up. First of all, that manual recounts were unfair and innaccurate. I'll deal with that one in the next section, but basically, total bullshit [see below]. The other argument was that we needed "finality and closure". Why we suddenly needed closure on November 14th and couldn't afford to wait another week (which is all it would have taken to complete all of the recounts if they hadn't had to keep stopping and waiting for court decisions), I never understood. But more importantly, how could anyone have even imagined that prematurely stopping the recounts would bring immediate finality? That's preposterous. It must have been obvious, even to Harris, that her action would result in court challenges which would likely take longer than just letting the recounts happen, and would in any case make the whole situation murkier. If finality, clarity, and closure, were what she really desired, the logical choice would have been to let all the recounts happen without interference. As it is, we'll never have true "closure" from this election. So what reason does that leave for Harris' decision? The only one left, as far as I've heard, is the very simple one: she wanted Bush to be President. She'd been chair of Bush's Florida campaign since before the primaries, and it's very likely she was expecting a posting of some sort in the Bush administration (which she's fortunately not gonna get). For this, she was happy to abuse her office and betray her electorate. So, what if we didn't have Katharine Harris, and we did have recounts? III.4. Punch Cards and Recounts As I said in a previous email, one of the things that skews vote counts all across the country, is the voting machinery used. The poorer a county is, or the more populated, the less it can afford to redirect its resources (money) towards upgrading old voting machines. And the older voting machines are much more likely to fail to count a vote than the newer ones. Here in the state of MA, punch card voting machines are banned. But in Florida, in general, the poorer and more urban areas used punch cards, while the more affluent areas used optical scan systems. There is a remedy for this problem, built into the law: manual recounts. If an election is close enough that it's possible the margin of error of the voting machines changed the result, candidates usually request a recount. This is not a new or unusual event - it happens almost every time an election is very close. And this election was far closer than most. In an echo of his father's 1988 campaign, the younger Bush succeeded in consistently attacking a part of our democratic structure so much that he was able to cast significant doubt among people who didn't know better. But while daddy Bush targeted the ACLU, his dimwit son attacked... manual recounts of ballots! Unfortunately, he succeeded, with the collusion of a media so wedded to "objectivity" they would rather give credence to the ridiculous, than appear partisan. And so now we have the mind-boggling situation where half the country don't trust manual recounts. As any election expert anywhere will tell you, counting ballots by hand is by far the most accurate and correct way of counting votes. Until we replace all our voting machinery with large-print menu driven touch screen systems running open-source software with extensively user-tested interfaces, manual counting will remain the *only* way to accurately count votes. There's no good substitute for a small group of human beings actually looking at a ballot, to see what it says, and determine who the vote is for - short of interviewing every voter. When automated voting machinery was introduced, it was not done for accuracy. It was done because it allowed a much larger number of ballots to be counted, in less time, by fewer people: in other words, it's *cheaper*. It was always understood that using these machines would introduce a much larger error margin than manual counting, and that if an election result was very close, then you'd have to count by hand to determine the winner. It was never the intention to force people to follow technical rules about how to mark a ballot, and not count their vote if they fail, or if the machines fail them. Almost every state in the country, FLORIDA INCLUDED, has a law on the books saying that a legal ballot is a ballot which you can look at and tell the intent of the voter. If the voter simply scrawls "Al Gore and all them demicrats", as one Florida voter apparently did, that's still a legal vote and ought to be counted. When one of the Florida counties set a strict standard of only counting ballots where the chad had at least two corners detached, a court told them that standard would not stand up in court. The reason was simple: If you have a ballot with a poke that looks like it came from the voting machine, in one hole and not in the others, then you can very clearly tell that ballot was cast by a voter who pulled that particular lever. Whether the chad detached or not, the intent of the voter is clear and unmistakeable. Almost every state in the country, FLORIDA INCLUDED, has provisions in its law for doing manual recounts, usually at the request of a candidate. In many states, the candidate who requests the recount has to pay for it, again reinforcing the fact that the main point of counting by machine is to save the state money. This is the way it has always been done, in Florida and everywhere else. When a machine count comes in very close, and someone thinks the margin of error of the machines may have swung the result, they ask for a recount. Until last month, I had heard of numerous cases of recounts, and many occasions where a recount overturned an election result, but I had never heard of the validity of manual recounts being challenged in court. It's utterly ridiculous. If you were following the news at the time, you probably already heard that Texas not only has a law providing for manual recounts in close elections, but that it has a recent clarifying law that says specifically that manual recounts are the preferred way to decide a close election, and that dimpled chads ought to be counted. George W Bush was Governor when this law passed, and he supported it. You may have also heard that Jimmy Carter, in an interview, said that manual recounts were the right thing to do. If you can't trust Jimmy Carter about how to run a fair election, then you can't trust anybody about anything at all. The Democrats had some election experts testify in court that manual recounts are the only way to accurately count these votes, and I don't recall the Republicans being able to find anyone with election logistics experience who would counter this. The Republicans made a big deal out of the "selective" nature of the recounts: only 3 counties out of 67 were at issue, and these were the Democratic strongholds. Recounting those counties would give Gore an advantage, that might be somewhat offset by uncounted votes in other counties, they said. Once again, they were playing games. Actually, the Democrats requested re-counts in 4, not three counties, but that's a small matter. The three counties in question, however, in addition to leaning Democratic, are also the three most heavily populated in Florida. Dade, Broward, and Palm Beach, put together, account for more than 25% of Florida's voters. So saying "3 out of 67" is very misleading to begin with. Even if it was selective, it wasn't as selective as they made it sound. Secondly, the Republicans had the opportunity to request recounts in 4 counties of their choice, and never did so. When Al Gore offered, *twice*, to re-count the entire state, drop all lawsuits, and abide by the result, they brushed it aside. Jimmy Carter also suggested recounting the whole state by hand, but they ignored that. Why? Because they realized that their argument against selective recounts didn't stand up to the facts. Like I said before, the more Democratic-leaning a county is, the more likely it is to be poor and large, and thus use older punch card voting machines that have a much higher undercount rate than the optical scan systems used in the counties that had the money to upgrade to them. It's not entirely coincidence that these three counties, the large Democratic strongholds, used punch cards, while 2/3 of the other counties used optical scan. The Republicans knew full well that the bulk of uncounted legal votes in the state were in the three counties they were contesting, and that failing to count those hurt the Democrats badly. They didn't want to recount the whole state because they realized that the results wouldn't be much different from the "selective" recount, and they wanted to hold on to the cry of "selective!" for its propoganda value. [ Although Florida law didn't specifically include a provision for any candidate to request a manual recount of the entire state, county canvassing boards could do this, or a court could have ordered it. If Bush had accepted Gore's proposal, I have no doubt the courts would have happily seized on the opportunity of the two sides actually agreeing to something, *anything*, and ordered it be done. So claims that a statewide recount was not possible are silly. ] There were a few other rather interesting cases of selective recounting. On election night, the entire state did a legally mandated recount because the margin of victory was less than 0.5%. In this legally mandated recount, each county had full discretion about how to do it. Most of them just ran their machine count again, and reported the new result. A few may have just resubmitted their original result without doing any recount, if they felt that it was accurate. Nassau County, run by Republicans, did a machine recount, and discovered that Gore had gained about 50 votes... so they decided to discard the new result, and resubmitted their original count. And one county did a full hand recount. If you were watching the recount results being updated county by county on the days immediately after the election, you might remember that very soon, we had results from 66 of 67 counties, and Bush was up by a little more than 200 votes... and then a couple of days later, the 67th county reported, and now Bush was up by just over 300. That was Seminole County, heavily Republican, doing a full hand recount. Although they used optical scan, with a very low error rate, their recount did turn up a few hundred more votes, with a net gain of more than 80 for Bush. Did we hear the Bush campaign yelling about how this was selective and unfair, and how manual recounts are biased and innaccurate? Of course not. But more significantly, did we hear the Gore campaign complain about this? Hardly. The Miami Herald published an analysis of the undercounts in Florida. Precint by precint (a precint is a much smaller unit than a county!) the figured out what the count would have been if all of the uncounted ballots had been votes in the same percentage distribution as the votes that were counted at that precint. The analysis is at: http://www.miamiherald.com/content/archive/news/elect2000/decision/104268.htm The headline of that article said that Gore would have won by about 23,000 votes (note that this does *not* take into consideration any other factors, such as the butterfly ballot, the absentee ballot corruption, allegations of poor service to black voters, etc). The Bush people dismissed this by saying that number is based on the assumption that every single uncounted ballot was actually a legal vote for President, even though there are in fact some people who do not vote for President, just for other offices, or who cast blank ballots. Some of those undercounts really were legitimate non-votes. OK, true enough, but anyone who accepts that hasn't actually read the article or looked at the numbers intelligently. While a 23,000 win for Gore is based on the extreme assumption that every single undercounted ballot was an uncounted vote, the certified result giving Florida to Bush by a mere few hundred votes is based on the much more extreme assumption that almost every undercounted vote was a non-vote. The Herald article includes a sliding scale showing what Gore's win would probably have been depending on what percentage of undercounted ballots were legitimate non-votes, and it shows that even if 9 out of 10 undercounted ballots were non-votes, Gore would still have won by about 1400 votes. Yes, that's extremely narrow, but it's not as narrow as Bush's supposed win. The average undercount rate of all the counties using punch cards was 3.9%. The average undercount rate of all the counties using optical scan was 1.4%. Does it make any sense to assume that a voter using punch cards is almost three times as likely to decide not to vote for President as a voter using optical scan? I haven't seen any plausible argument yet why this should be so. The much more likely explanation is simply that punch card machines are much more likely to fail to read a legal ballot, a fact we already know to be true. So let's for a moment assume that the optical systems were nearly perfect (they weren't quite), and the undercount rate for the whole state was about 1.4%. This would mean that on average, in Gore-leaning punch card counties, about 64% of the undercounted ballots were legal but uncounted votes, while in more Bush-leaning optical scan counties, most of the undercounted ballots were non-votes. If that were the case, Gore would probably have won with a margin on the order of 15,000 votes. [See also this: http://www.miamiherald.com/content/archive/news/elect2000/decision/072960.htm ] Remember those exit polls? The ones that caused VNS to call Florida for Gore right after the polls closed? Exit polls have a substantial margin of error, but that margin isn't particularly skewed in one direction or another as far as I know. On the other hand, exit polls are not tainted by the punch cards, or the butterfly ballot, two things which did skew the actual vote count decidedly towards Bush. If you take those two factors into account, Gore was ahead in the Florida vote by about 30,000 - 45,000 votes, and that's what the exit polls would have predicted. The original call of Florida for Gore was not a bad call. It was probably correct, and the actual vote count was really the "bad call". [ The later mistaken call of Florida for Bush was not based on any facts, it originated from a reporter at Fox who was connected to the Bush campaign, and either got too excited or deliberately wanted to give the Bush campaign a propoganda victory. In reality, at the time of that call, all the facts pointed towards "too close to call". But as usual, once one news source makes a call, the rest follow blindly so as not to be left behind. ] IV. Summary More people voted for Gore than for Bush, and this is not disputed, but doesn't matter. We use a district voting system for president, which has certain advantages, but which skews the vote. However, the population represented by the set of districts that chose Gore also exceeds that of the set that chose Bush. This is true whether the districts being considered are states, or congressional districts. Bush's electoral college victory was the narrowest in history, and depended on almost every last inch of the overrepresentation of conservative leaning states in the electoral college. In one state, New Hampshire, it seems likely that Bush's victory depended on third party voters who would likely have preferred Gore. In two states, Florida and New Mexico, the result was so close as to be called a tie. In New Mexico, it probably was more or less legitimately tied, as far as we can tell, although it's possible that Gore really won, and some of the usual factors that work against poorer voters may have made the margin closer than it ought to be. In Florida, documented irregularities demonstrate that Gore almost certainly won. If you either don't count these two tied states, *or* if you split up their electoral votes, *or* if you give NM to Gore and FL to Bush but flip NH back to Gore... in any of those three cases, Gore would also win the electoral vote despite the overrepresentation of conservative states. Throughout the country, poorer voters and urban areas suffer from a higher likelihood of having their votes rejected due to older voting machinery. We know that almost certainly turned the result in Florida. It doesn't seem to have turned the result in any other state, this year, but in general, Gore's narrow victories in several states were probably somewhat less narrow than they appear. We do not know of any cases of large-scale vote skew towards Gore, and we can be pretty sure that if any such cases had turned up, the Republicans would have made quite a big deal out of it in the news and we'd have heard about it. Look at the mountain they made out of the postmark molehill, after all. But that was apparently the best they could find. Bush's certified victory depended on the following: - The butterfly ballot in Palm Beach - Punch card ballots being used more in poorer precints - Republicans in charge of Florida, especially Katherine Harris - A media blitz successfully using Big Lie techniques, supported by the media's bias towards "objectivity", in conjunction with endless court challenges, running out the clock on recounts - Chief InJustice Rehnquist of the Supreme Court If any one of these factors had been absent, Gore would be President elect. I think I can say this with a fair amount of certainty. Note that these are mostly things which biased the results towards Bush. Gore wouldn't have needed them replaced with something biasing the results towards him, to win. As I said before, Gore would have won if Katharine Harris were replaced with an impartial secretary of state. Similarly, if Rehnquist were replaced by a fair and sensible Chief Justice, even if he were conservative, Gore would have won. If the voting machines in Florida were fair, Gore would have won. And so on. It was only the combination of a series of factors, each of which gave Bush a further advantage, that, all put together, delivered the narrowest of victories to Bush in an election which he lost by almost every standard applicable. Except the corrupt certification standard, where whoever controls the relevant organs of government, and is willing to abuse that power, gets to steal close elections. -- Cos (Ofer Inbar) -- cos@polyamory.org http://cos.polyamory.org/ -- WBRS (100.1 FM) -- info@wbrs.org http://www.wbrs.org/ Don't throw your vote away! Let Katharine Harris do it for you. ---------------------------------------------------------------------- Date: Fri, 29 Dec 2000 22:58:24 -0500 From: Ofer Inbar To: [mailing list] Subject: Re: who won (was Re: Politica) [another list subscribed] wrote: > Gore won Pennsylvania. Pennsylvania loses many representatives next time > redistribution comes around. Maybe life just isn't fair. > > If only congressional districts voted, it would be Gore by > > 225 to 211. > > unless they did redistribution of representatives more often than they do > now. in which case, gore would probably lose some votes. Good point. New York is also set to lose some representatives, and Texas is going to gain. Then again, I think all three west coast states (CA, OR, WA) are also set to gain representation. I'm not sure exactly what the overall effect of this would be on the Senator-free electoral vote count, but I do know this: The total current population of the set of states Gore won, exceeds the total current population of the set of states Bush won. So, whatever the numbers turn out, it would still lead to a Gore victory. > > Note that I'm talking specifically about *overseas* absentee > > ballots. The Republicans liked to say that there were 23,000 absentee > > ballots in Florida this year, but most of those were already received > > and counted on election day, and included in the first count. Only > > about 3,000 overseas absentee ballots arrived after election day, of > > which less than 1,000 came from military voters.] > perhaps i'm confused, but the last time i checked with retired military > (otherwise known as papa), all overseas mail (shipside or landside)needs a > postmark. how does it get sent otherwise? Theoretically, yes, all mail, including military mail, needs a postmark. However, military mail goes through its own postal system, and they apparently don't always follow the rules. There were absentee ballots received from overseas military voters with no postmark at all. Some others had postmarks from the postal center in the US that received them from the military (so theoretically, someone could have voted the day before election, sent in the ballot, and only had it postmarked several days later, when it arrived in the US). There weren't a very large number of these, but given how close the count was, they might have made a difference, and the Republicans did make a very big deal out of it. > I'd like to say something... They've got two years, down in Florida. > Will the voters remember this fiasco two years from now? I'd wager > to say they will... The Republicans are probably going down in > florida. they'll take a hard hit. And I'll lay a wager that > governor of Florida stays Democrat for fifteen years. let a > generation go by, and they might be willing to change. but I think > now they want the republicans out. Oh, it's not just Florida. As of now, I think the Republicans are headed for losses all over the place in 2002, including both houses of Congress. Things might happen in between that would change that, but they're gonna have to be rather big things. And Florida has long tended a little to the Democratic side. The Republican dominance is fairly recent, and they've abused it to such an extent that they're likely to lose it soon. Jeb Bush might actually pull through, though. Unlike his big brother, he's got a brain and some political skill, and doesn't need to depend on his dad's people to win an election. He's also got some very relevant experience. The first time he ran for governor of Florida, he lost narrowly to Lawton Chiles, an extremely charismatic folksy well-loved man who campaigned door to door (he got the nickname "the walking governor" because he literally walked through most of the state during his campaign). Jeb studied hard, made changes in what he said and who he said it do, and the next time around, he narrowly beat Chiles. Chiles is now dead, so he's not around for a rematch. Jeb's got a big disadvantage to overcome because of election 2000, but he also did a decent job of maintaining plausible deniability (if you want to, it's not that hard to believe that he had little to do with the fiascos in this year's election), and he's got the brains and the skills to overcome it. Maybe. On the other hand... Jeb Bush got 14% of the black vote in Florida the last time he ran, up from 7% the time he lost to Lawton Chiles. Shortly after being elected, he tried to reform affirmative action. His plan actually wasn't bad, but politically, that's not going to matter, because he didn't have the rapport and trust established with the black community to be able to propose something like that and get away with it. They feel betrayed, and he'll be lucky to even match that 7%, next time. Also, black voter turnout is on a steady rise, election 2000 will encourage even more, and the likely upgrading of voting machinery throughout the state (and the country!) will edge their vote up even more. If Chiles were still around, I think he'd have a really good shot at taking the governorship back, but he's not, and I'm not sure who the Democrats have to throw at Jeb. -- Cos (Ofer Inbar) -- cos@aaaaa.org cos@wbrs.org cos@polyamory.org The law doth punish man or woman, That steals the goose from off the common, But lets the greater felon loose, That steals the common from the goose. Anon, 18th cent., on the enclosures. (courtesy of John Whiting)