|
|
|
|
|
|
|
Click here to see our elections web page
MIM has done some independent investigation into whether Republicans stole the elections in Florida; even though our candidate was "none of the above." An impressive analysis titled "Surprising Pattern of Florida Electoral Results" shows that optical scan voting machine counties seem to have increased the vote for Republicans while touch-screen voting systems did a more accurate job. We went over this evidence meticulously and concluded that it is more likely that Florida whites did vote for Bush though registered as Democrats, especially in the rural areas and areas up near Georgia, Alabama and Mississippi. There was no fraud, except in whites claiming to be Democrats when they are really Bush voters. We can't say we blame the less intelligent rural whites for registering Democrat and then voting Bush when there is so little difference between Democrats and Republicans anyway. Nonetheless, we will now go over the evidence of polling fraud versus white politics fraud.
See, http://ustogether.org/Florida_Election.htm
Currently, there are a number of sweeping analyses circulating that indicate Bush stole the election in 2004. MIM starts the process of digging in in the Florida result that MIM found most impressive so far.
"Surprising Pattern of Florida Electoral Results" shows a number of counties with heavy Democratic Party registrations voting for Bush. Since we know that overall over 90% of Democrats voted for Kerry, this seems odd and suspicious.
When we see that the touch-screen counties had results well predicted by voter registration and the optical scan counties did not, our blood should already be boiling about a stolen election.
To start with, MIM looked at the optical scan counties showing registered voters that are over 70% Democratic, because it is there that the potential for a gap between what we would expect and what happened is largest.
Among the optical scan voting counties supposedly registered over 70% Democratic, Bush won Calhoun, Dixie, Franklin, Hamilton, Holmes, Jackson, Liberty, Lafayette, Madison, Taylor and Union. Kerry won only Gadsden and Jefferson among the optical-scan counties over 70% registered Democratic according to the Florida results. (With Gadsden 57% Black and Jefferson 38% Black, we're not surprised the Democrats won.) Based on this, our blood is still boiling, even hotter. It seems that Kerry does not even win the predominantly Democratic counties in the optical scan counties and we wonder how that could be true.
So then MIM checked the data on the counties to see if the registration information is correct. Are these really lopsidedly Democratic counties? By checking the Republican- controlled state government, we find that yes, these are heavily Democratic counties as the analysis shows. So our blood boils hotter. It looks very suspicious.
http://election.dos.state.fl.us/voterreg/pdf/2004/2004genParty.pdf
Next, we checked on elections results in the past. If we check the 1996 results when Bob Dole and Ross Perot opposed Clinton, we see that Clinton won many more counties, but he still lost Holmes, Jackson, Liberty, Lafayette and Union. If there is a fix on, the fix was on in 1996 in those counties too.
In 2000, Bush also won Calhoun, Dixie, Franklin, Hamilton, Holmes, Jackson, Liberty, Lafayette, Madison, Taylor and Union. In Bush's favor, in 2000, Calhoun had manual tabulation, so if there were a fix in 2004, it would have had to have been different than the one in 2000 (if we assume that's how Bush won in 2000.) Likewise, Gore won Gadsden and Jefferson like Kerry did in 2004. OK, so now we turn to being perplexed, because in 2000, Florida was famous for the "hanging chad" punch card system, a paper ballot system.
So now we turn to the question of punch cards. If Bush won punch card counties over 70% Democratic in 2000, then we can no longer surmise that there was an across-the-board programming conspiracy by optical scan programmers alone that accounted for the problem.
It turns out that Bush won only Dixie by punch card in 2000, but it was a big margin 2697 to 1827, so we cannot say it is impossible for an overwhelmingly Democratic county to go big for Bush. If the whites fixed Dixie, they fixed it in a different way in 2000. The other counties Bush won generally had a different version of optical scan than used in 2004.
http://www.rose-hulman.edu/~bremmer/professional/ballot_florida_election_paper.htm
The question remains, whatever differences there were, was there some kind of fix put in in 2000 and 2004, perhaps by slightly varying optical scan methods?
Something we would not be surprised to see is whites in control of a county with a heavy Black voting bloc. Another reason that Democratic registrations could be wrong is that we could be looking at counties where people long ago registered as Democrats and then drifted Republican as much of the South did--no fraud explanation necessary. So we checked into the racial and age composition of these counties.
Calhoun--79.9% white, 14.0% over 65, 22.9 people per square mile
Dixie--88.8% white, 17.1% over 65, 19.6 people per square mile
Franklin--81.2% white, 15.7% over 65, 20.3 people per square mile
Hamilton--58.8% white (42.5% female!), 11.2% over 65, 25.9 people per square mile
Holmes--89.8% white, 14.8% over 65, 38.5 people per square mile
Jackson--70.2% white, 14.6% over 65, 51.1 people per square mile
Liberty--76.4% white, 10.2% over 65, 8.4 people per square mile
Lafayette--79.3% white, (40.2% female!), 12.4% over 65, 12.9 people per square mile
Madison--57.5% white, 14.6% over 65, 27.1 people per square mile (74.7% voting age, 18,766 people)
Taylor--77.8% white, 14.1% over 65, 18.5 people per square mile
Union--73.6% white, 7.5% over 65, 55.9 people per square mile
Source: http://quickfacts.census.gov/qfd/states/12/12125.html
The average number of people per square mile in Florida is 296.4. The average percentage of people over 65 years of age in 2000 was 17.6%.
When we look at these worst-case counties of heavily Democratic counties supposedly won by Bush, we can conclude a few things now. 1) They are all rural counties, less than one fifth the average population density of Florida. 2) Not a single county is a retirement community relative to the rest of Florida. By contrast, ALL the touch-screen counties were more populous than all of the worst-case optical scan counties over 70% Democratic but voting for Bush.
Of these counties, we are most suspicious that Bush won Madison. Clinton won Madison. Bush barely won Madison in 2000 and again in 2004. Nonetheless, the Madison numbers are believable in a race and party sense. It would only mean that two-thirds of registered white voters voted for Bu$h, a completely believable result.
See, http://election.dos.state.fl.us/voterreg/pdf/2004/2004genRace.pdf
Hamilton has a suspiciously low proportion of Black voters registered relative to population. However, that is a problem unrelated to optical scanning, but it could point to a different kind of intimidation and Democratic Party infiltration by unscrupulous whites; although registration can go through the state level, also controlled by Republicans. The solution for this problem is to send Florida city Democrats to the rural areas to monitor what is happening and not to trust rural whites to monitor rural whites. Each machine used should be tested several thousand times before the real vote to see how it tallies Bush, Kerry etc. The results of those tests should be public knowledge and we should not have to rely on uneducated, possibly senior citizens to tell us that there was a statistical fix on.
MIM checked these counties on the ground, and we are not able to say we had enough people observing politics in Florida to contribute decisively but that nothing in this article that we have said is out of line with what we saw as individuals.
In the most suspicious looking counties, Madison has a believable vote total and Hamilton simply did not have Blacks registered to vote. That would be a problem other than optical scan.
If we look at the most populous optical scan counties in Florida--Duval and Brevard--Brevard is 86.8% white, so although the proportion of Republican voters went up several percentage points compared with previous elections, even Bob Dole won Brevard before, so a 42% election result for Kerry is not too incredible.
Plus, Duval was punch card in 2000 and it went 3 to 2 for Bush then too. In fact, since the Bush advantage lessened to 4 to 3 in 2004, we might have to thank optical scan for the single biggest Kerry gain over Gore in the optical scan counties.
The suspicions placed by Kathy Dopp are definitely justified. Any time we see a county over 70% Democrats go overwhelmingly for Bush, we have to wonder about fraud. It just so happens that these counties in the optical scan group were the rural counties and the big urban concentrations were in the touch-screen voters. The real Democrats are in the cities. The "Dixie Democrats" live in the rural areas and that accounts for the difference between touch-screen and optical scan voting. If there were a fix in the voting, we should look at registration and also the fixes on in 2000 and 1996.
In response to this sort of point, the analysts showed that even in the populous counties, the optical scan machines resulted in a small gain for Republicans relative to what registrations would predict.
See, http://ustogether.org/election04/FloridaDataStats.htm
Simultaneously the analysts admit that there is just no relationship between registration and voting in smaller counties. The rural Dixiecrat phenomenon is the explanation for that. So then the question becomes whether the more urban counties prove that there was a slight fix averaging perhaps 6% of the vote on optical screens. The analysts present good evidence for that without giving us a reason to think we should toss the rural results. In other words, MIM still has one theory for the data: the more rural the white, the more phony the Democrat. The districts using optical scan are smaller population-wise than the real cities using touch-screens. That is true even in the category over 100,000. The average touch-screen county in that category is 358,620 votes in 2004 while the optical screen over 100,000 voter category is still smaller at 209,005 on average.
| Average number of voters | %Black population | Population per square mile | |
| Touch screen | 358,620 | 10.4 | 924.7 |
| Optical scan | 209,005 | 16.0 | 534.6 |
The above table shows again our assertion that there is one theory to explain the whole thing. The more urban the white, the more like a Democrat. The more rural the white, the more likely to vote for Bush, whether or not registered as Democrat. This holds true even in counties with over 100,000 voters, because in Florida the touch-screen counties are simply more densely urban than the optical scan counties. The only counter to our argument would be that the optical scan counties are a higher percentage Black, so there is a lower proportion of whites to have that Dixiecrat effect. On the other hand, we should note that the above data is an arithmetic mean, which just means that each county counted equally, not based on its population when it came to calculating percent Black. The arithmetic mean is appropriate because we are criticizing a paper that used the arithmetic mean.
Since some counties are much bigger than others, it would be nice to be able weight them appropriately or take each individual voter separately. Unfortunately, the present reviewer does not have the exit polling data to be able to analyze that further by taking the characteristics of voters to predict how they vote. Nonetheless, to oppose what we are saying potential critics would have to take a patchwork approach, while we stick with our one theory about whites which explains even the counties over 100,000 in population.
If we look further, we will see that none of the touch screens had use in the most conservative parts of the state that ex-New Yorkers do not move to, namely the western and northern parts of Florida excluding Jacksonville. Of course those counties closer to Georgia and Alabama are more like those states with a few exceptions. Those areas have been like that long before there was such a thing as optical scan voting and so we are not surprised Bush got his highest vote there and his lowest in southern Florida near Miami, which is a disproportionate share of touch-screen voters.
((http://www.cnn.com/ELECTION/2004/pages/results/states/FL/P/00/epolls.0.html) (which somewhat contradicts our view of the suburbs but without explaining how they decided what is a big city. It probably means that affluent New York liberals moved to nice southern suburbs.))
In conclusion, if we look at the most suspicious looking counties, they are rural counties with a history of voting Republican. The rural white nature of these counties raises the question whether hardcore Democratic activists are deluded about the nature of whites in Amerika, and that is why they are claiming another stolen election. Democrats are now claiming exit polls may be more accurate than fraudulent voting machines, but one thing the polls showed before and after the election was that 57% of whites were going to vote for Bush. That percentage decreases in the cities and increases in the rural areas. Needless to say, a 57% white vote for Bush is tough to overcome. In the South, it's even harder.
There were widespread problems that voters had in waiting in line and having their votes counted. The reports throughout the South of Blacks not being able to vote are outrageous. In this article, we only examined the idea that it was a fix in optical-scan voting that produced crazy looking results in Florida. While the ole' boy whites may have a reason to fix the ballot box, there is no denying that academic research shows whites and rural people more inclined to Republicans. Academic and journalistic polls cannot be fixed in the same manner, so we suggest to hard-core Democrats to take a serious look at alleged Democrats in the rural areas, and not to give too much credit to the majority of whites.
We're not bragging too much, because the vast majority of people calling themselves "Marxist" has the same delusion about Amerikkka as these hard-core Democrats who think Kerry won. It comes from not having the overall picture on whites. We do not live in an America of civics theory. It's Amerikkka and everyone seeking change should get used to it and adjust strategies.
You can send comments on this article to mim3@mim.org
[HC123 raises doubts:
A quick glance at the figures shows higher turnout in the counties using optical scanning. In those counties, 72.4% of the voters voted; in those with e-touch, 69.3%. Could it be that the former salted their ballot boxes with a few extra Republikan ballots? ]
mim3@mim.org replies:
A three percent voter turnout difference is not that big. We could expect it sheerly on the basis of age differences
or the urban/rural distinction once again. Here is an even bigger gap in Pennsylvania in 2000:
"The number one county for voter turnout in 2000 was rural Forest County at 76 percent, and the two counties with the
lowest turnout were urban Philadelphia and Centre counties with 55 and 56 percent, respectively. The top 12 counties for voter
turnout were rural." (Source: http://www.ruralpa.org/news0501.html )
Unless we are in a rural area that is much younger in age than the urban area, we are going to expect higher rural turnout.
On the other hand, the rural districts in Pennsylvania also used optical scan in 2000 and set a record vote for Bush with a 65% turnout.
http://www.prea.com/pennlines/august%202001.pdf
The rural vote turnout is higher also in Japan:
http://countrystudies.us/japan/117.htm
Also, Ireland: http://www.ucd.ie/dempart/workingpapers/dublin.pdf, page 35
HC123 comments again:
You're right about the votes in Florida. It's just a matter of urban and rural. If the two voting systems had been assigned randomly to counties, these data might be suspicious. But they were assigned on the basis of population: the one system to urban counties, presumably because it was more efficient at counting the large number of votes; the other system to rural counties, presumably because it was cheaper.
As for the difference in correlation coefficients, again, it's incorrectly attributed to the voting systems. There's little evidence that a difference in voting techniques is responsible for the discrepancy. The data sets--one rural, the other urban--are not comparable, so the analysis of correlation coefficients is incorrect. Although at first it seems suspicious that counties with a large majority of registered Demokrats would vote for a Republikan in an election that was 90% along party lines, the most dramatic discrepancies turn out to appear in small counties, and they could easily be explained by obsolete data on party affiliation or other factors. All that Dopp has demonstrated is that rural populations are more likely to vote Republikan than urban ones. But we knew that already. That's why North KKKarolina and MiSSiSSippi voted for Bu$h while Massachusetts and Illinois voted for Kerry.
The best analysis is MIM's: rural crackers will vote Republikan; exiles from the Frost Belt will vote Demokratic. That suffices to explain the discrepancy that Dopp noticed. The weakest point in MIM's analysis is the omission of Latinos, an important population in Florida, especially in the southern part. Unfortunately, it might be difficult to assess the Latino vote, which is likely to be mixed, with anti-Castro émigrés from Cuba tending to favour Bu$h. Black voters, on the other hand, strongly supported Kerry in this sham of an election, a fact that facilitates analysis.
mim3 for MIM replies: The comrade is correct. It's tough to factor in the Cubans. We'd have to go by precinct. In only one other state did Bu$h win the Latino vote. That could be because of the Cuban influence or a special connection to the evangelical churches. At the very least, we know Bu$h was going to rack up some heavy Cuban votes around Miami regardless.
On an overall basis of the Latino vote, it would be too hard to know, because the vote was too close to 50/50 in Florida in both the overall vote and the Latino vote. In most of the rest of the country, if we saw a heavily Latino precinct vote for Bu$h, we could be suspicious. When we see counties that are heavily Black, then we have better bases for questioning pro-Republikan results.