Neo, a cyber-friend Jason Cawley has done a terrific job of putting the statistical evidence of voter fraud together. It's a bit long, but well worth it. A MUST read...
freerepublic.com
The following is a letter I recently sent to Scott Rasmussen about some of the possible fraud issues in this election.
The original posts were on another thread. Someone there suggested that they be reposted as an article of their own. I have chosen to post the letter sent, in the "vanity/opinion" section, so that others may see it and comment on it if they wish.
Dear sir,
Since the election, a number of us at Free Republic have been discussing a theory of events that suggest, among other things, that your pre-election polling results might have been more accurate than you may think. The theory involves large scale vote fraud by the Democrats. The evidence for this is statistical, supported by anecdotal evidence, spin, etc. I thought you might be interested in this possibility, or in testing it.
First, though, a word about why most republicans aren't shouting about this. The following are pieces written for Free Republic, by me except where otherwise indicated. In earlier stages of working out the idea, the input of other regulars was vital. I will take responsibility for this way of stating things, however.
Proven fraud is about the only thing a judge will allow to invalidate an election and call for a new one. As long as Bush is going to win anyway, it is more dangerous to establish the fraud occurred than to leave it in doubt. While you and I can easily see through the "heads I win, tails we flip again" aspect of this, that is not necessarily how it would play on CNN with the liberals all spinning for a re-vote.
In addition, the evidence for more widespread and massive fraud is either anecdotal or statistical. The ancedotes prove local things but nothing more. The statistical proofs don't play well on CNN - what is a cross correlation, a T-square, a 4-sigma event? The Dems have prepared rival spin explanations for every aspect of their frauds. Those rival explanations do not withstand statistical scrutiny. But as mere spin, they are perfectly adequate.
The Story Itself
I can give examples of some of this, and some of the additional "coincidences" that all make sense with the theory, but are a pile of unrelated coincidences without it. Start with overall turnout. If you graph election turnout against time, you see a smooth, slowly declining curve over the last 30 years. Broad demographic trends account for this. People in some demographic categories, more common today, just vote less and always have. OK, project the curve ahead with the same slope to this year, and determine the predicted number of votes, total. The answer comes out "100 million". The actual votes officially cast are above that figure, by on the order of 5 million votes, perhaps more now that all the absentees have been counted.
The scale of the difference is on the same order of magnitude with the difference between the election day vote total percentages, and the pre-election polls of most pollsters. (50-47 to 50-45, => 3-5% difference between polls and official votes, is about that many million votes). Coincidence number one.
The rival explanations for this "uptick" in "turnout" are - null hypothesis - it is just a close election with people interested, so turnout was up. Fraud hypothesis - people voted multiple times, or voted absentee as well as in person, or voted for others who did not turn out, effectively stuffing the ballot boxes. Can we test which theory is more likely? Only crudely, but yes.
The null hypothesis predicts a broad increase in turnout in every district of the country, with turnout rates varying from place to place, but randomly, and thus in a normal distribution. The fraud hypothesis predicts the turnout increase will be concentrated in those places where the frauds are easier - liberal absentee ballot rules, liberal proof of ID practices at polls, particularly one-sided partisan districts. I have not seen exhaustive testing of the rival theories, but the press reports I've seen favor the latter. The higher than expected turnout is not evenly spread, but concentrated, particularly for the Dems and in core urban areas.
Now, what about the double-punching component of the fraud? Well, the idea is that half of it is ballots fraudulently double-punched to leave intact votes for Gore, give him unpunched ballots, and disqualify votes for others [see below for more details on this]. The null hypothesis is that double-punches are caused merely by voter confusion.
First, look at the scale, how many are there? Press reports have said that disqualified ballots usually run around 0.5% of all votes cast. That comes to 500k and change. Total disqualified ballots nationwide were actually 2.1-2.8 million, according to a professor's study reported in the press. The excess is 1.6-2.3 million disqualified ballots, or half of the difference from the pre-elect polls. The half pulling the Bush vote down, on the fraud hypothesis.
But can the fraud hypothesis be distinguished from the null hypothesis, "there were just more errors this year"? Let's look at it. Where were they? In Florida, there were 170k total DQed ballots. That is 170k/~2.45m nationwide, = 6.94% of the nationwide DQs. You can check portion of population, roughly, by number of representatives, which is electoral votes minus 2 (the sentators). So for FL, that is 23 out of 435 = 5.3%. There are 1/3rd more DQs in FL than the average by population. Similarly, there were 120k DQs in Chicago alone, about 5% of the nationwide DQed ballots. Chicago is nothing like 5% of the population.
DQs are highest in Dem strongholds. In FL, there are 5/3rds as many DQed ballots in the less than 10% of the counties with total DQ rates over 10%, as in the statewide average. These counties are heavily pro-Dem in the official vote. In addition, the rate of double-punches disqualifying votes is 5 times as high for the presidential race as for other state-wide races in Florida. This is getting to be quite a pile of coincidences.
The Dem spin on the concentration of DQed ballots in the counties that went their way, is that their voters are the ones that are dumb and thus vote funny, and fail to cast legal ballots. We've heard that spin incessantly, since election day, and in Palm Beach (probably location of the late-night, Gore-in-limo, operation, perhaps along with Broward) they hired a telemarketing firm to find honest double-punches or people willing to allege the same to match their story. More on these spin issues later.
Can we test their confusion hypothesis against the fraud hypothesis? You bet. This is their weakest point, the place where their spin cannot account for the facts and fails most decisively. Unfortunately, that failure is too complicated to play well on CNN, but I can explain it easily enough. The way to distinguish the two theories is by their implications for true turnout.
On the Dem "dumb" hypothesis, every double-punched ballot (or the majority of them, minus a small-scale honest-error background for all parties) represents a Dem who turned out but was not counted in the official figures because his or her vote was invalid. On the Dem "crooks" hypothesis, every double-punched ballot represents a republican or independent who turned out, voted for Bush, a minor candidate, or didn't vote on the Prez line at all, whose vote was thrown aside, or in the case of the last, turned into a vote for Gore. OK?
So, what are the measured turnouts in these places with many double-punches, to start with? And we see the official numbers show like 45-50% of the registered republicans voting for Bush, and ~100% of the registered Dems voting for Gore - sometimes more. Add the "dumb" hypothesis, and you have to assume the independents broke overwhelmingly for Gore, turned out strongly, *and* all the Dems turned out and voted for Gore, *while* less than half the registered republicans turned out. That yields turnout rates for the Dems that are *boosted*, compared to their turnout rate in other counties. And republican turnout rates that are depressed, compared to other counties - even with the "errors" supposedly "pulling" the other way.
But try the other, fraud hypothesis, and guess what? Then the implied true turnout from both registered Dems and registered Republicans are both in the 2/3rd to 3/4ths range in these areas, *just like they are everywhere else*, with the independents breaking fairly evenly or a little pro-Dem, with relatively low turnout.
It is reasonable to expect turnout rates to vary from county to county. But randomly, and thus in a normally distributed fashion. But if you make a histogram of turnout percentages, you will find the high double-punch areas out at the ends of the distribution, low for Republicans and high for Dems. The fraud hypothesis pulls them back into the rest of the distribution, lowering the implied variance in true turnout. The error hypothesis pushes the same areas even farther out, away from the rest of the pack.
In addition to this internal, statistical trail, we have anecdotal evidence of an entirely different nature to back up the theory. A Dem is found with a voting machine in his car. Another is put up for sale on E-bay. Voters in Wisconsin are seen voting several times, some of them brag about it. IDs are not checked at polling places, in Cook County, Chicago, in Wisconsin, etc. Mass mailings in Spanish in CA encourage non-citizens to vote. Votes are found cast by felons. The Dems loudly declare they will do anything to win. After the election, they prove it with multiple recounts in selected areas, eaten chads, taped ballots, rejected military votes, etc.
And what makes "dimpled" chads? You can't make them with one-at-a-time voting with the poll-booth stylus, even if you try. But they occur naturally a the bottom of a stack of partially-punched ballots, all rammed at once. [see below]
The cumulative evidence takes on far more than the character of mere circumstances or mere anecdotes. There are dozens of anomolies that require many independent things to all have turned out differently than expected, some of them extremely unlikely from in a statistical sense, to cover even some of the facts. But all of them, including some not explained otherwise, are easily handled by a single theory.
That being, the Dems attempted to steal this election with massive vote fraud, on the order of 2 million Bush votes cancelled and about as many, perhaps somewhat more, manufactured for Gore, overall 4 million votes plus or minus 1 million. That when even this did not succeed, they tried to steal it again in a late-night operation in Palm Beach and/or Broward, literally as we sat waiting for Gore to make his concession speech. And when that too did not get them enough, they challenged the entire outcome with the circus we have seen, a centerpiece of which is their own fraud being used as evidence that the process was "unfair", so it should be done over, or outright victory awarded to them.
Gore is a fraud. As the dictionary defines the term, "one who lies intentionally to induce another to part with a legal right or something else of value".
I hope this is interesting...
Question - was the late-night 50,000 vote surge cause by the southern Dem strongholds finally coming in? Another anomoly.
Nope. I watched these returns. Bush was up 170,000 votes before those big Dem counties came in. They knocked his lead down to 20,000 at one point, by which times those counties were 88-95% reported. The networks had *not* called the state for Bush at that point. Then, the narrowing trend reversed, and Bush's lead widdened out to +52,000 votes again. The southern Dem counties were already in, and the only places left in the 50-80% reported range (or lower) were smaller rural counties where Bush was leading. I was watching the VNS news feed county breakdowns as published on CNN's website, and I saw so myself.
When Bush's lead began widdening again and I noticed this about the counties that had and hadn't reported, I predicted to the folks I was with (they were on the TV, I was on the computer) that Florida was going to go to Bush and the networks could now call it, since the southern Dem counties were in. And shortly after I made that prediction to them, the networks did indeed call Florida for Bush.
The disappearing 51,000 vote act happened *after* this. Many commentators and election-watchers had gone to bed. For example, NRO had been counting down the % of FL vote and stopped. Gore was in his limo on his way to give his speech, and we heard it announced that he had called Bush and conceeded - none of which would have happened if the southern counties weren't already reported. The % reported climbed up to 99%, and Bush was still ahead by 34,000 votes at that point.
Then the weird stuff happened. The votes changed while the % reporting sat dead still, and while everyone was expecting Gore to get out and give his concession speech. There was one middle figure on the % reporting, 99.6% was listed, when the Bush lead had dropped to 14000 votes. Then it was 6000, 100% reporting. It sat there and the vote dropped to +2000, and the networks put Florida back in the undecided column and told us about the retracted concession.
They tried to steal the election earlier with pre-planned amounts of fraud in every district. My personal estimate of the stolen vote country-wide is around 4 million votes, and there are lots of different statistics and anomolies that all point to that same conclusion. The excess double-punches are a trail left by their manner of robbing votes, by invalidating votes for Bush / not for Gore.
This shows up in a statistical cross relationship - Dem turnout is boosted, and Republican turnout is depressed, in exactly the areas with the most double punches. This does not fit the Dem spin that it is there vote that is depressed by the doubles. It does fit the fraudulent double-punch theory, perfectly.
I think they tried to steal an extra 40,000 votes at the last minute, screwed up and only stole about 30,000. They may have cast some extra votes for Buchanan in the process, owing to a mistake. They were trying added stealing means after the concession. In my opinion, Gore could not go through with the concession speech, and the additional theft attempt was a last gasp alternative.
Since election night, they have been playing by the following script. If they stole it successfully, that's it, it stays stolen. When they didn't, they pointed to their own frauds as evidence of irregularities. Heads, they win the election, tails, they demand a new vote or that a judge give them the race or whatever. For sheer chutzpa, it is pretty amazing.
This has all been extensively discussed here, especially so in the first week after the election. The closer the statistical evidence is examined, the more it points to the same widespread fraud, on a massive scale. If this election had been conducted honestly, it is likely W would have won by 2-4% of the vote nationwide, and solidly in the electoral college. Fraud made it razor thin, and we are darn lucky they didn't plan for enough of it to swing Florida, or Gore would already by on his way to being a fraudulent President.
Contributions from others -
Here is a very simple experiment that would show the whole country that dimpled chads are forensic evidence of voter fraud. Call a press conference and on live TV perform a ballot experiment. The news people are swarming around Florida and would come.
Tools required:
1. A voting machine
2. About 100 blank ballot cards clearly marked VOID.
3. A small hammer.
4. A small nail with the point filed flat.
The Experiment:
Part A. Use the voting machine and try to make a dimpled chad on a SINGLE ballot. It can't be done. It either punches completely through or nothing. This would show that counting a dimpled chad as "voter intent" is not valid.
Part B. Take a stack of cards about 100 thick and place the nail over a hole and tap it through the cards till it will go no farther. The top cards will be clearly punched. The bottom cards will be dimpled. Show the dimpled cards on the TV camera. Let them zoom in close. This would show one way dimpled chad could be manufactured.
This simple experiment would put an end to the dimpled chad count and expose the associated vote fraud.
- Varmint Al (varmint@DittosRush.com)
On the whole theory, consult also -
- Robert A. Cook, PE (cook.r@csaatl.com) reagan.com
I hope this is interesting.
Sincerely,
Jason Cawley
Chicago, IL
Supporting data...
geocities.com |
