Sunday, July 09, 2006
The Things We Think We Know, II: How our media successfully mislead even the wisest among us and how to see through the shadows to understand Mexico
I suspect that if a Dem wins in 2008, the threshold for dodginess will suddenly become a whole lot lower...
Huge demonstrations are supposed to take place on Wednesday. We'll see if AMLO can keep a lid on it, and keep them peaceful.
Will we have another Ukraine, 2004? Stay tuned...
However, I wouldn't have put it up if I wasn't reasonably certain that one actually doesn't need to see it. First, the dropoff of AMLO's margin was perfectly linear as the vote count rose. This is shown in the graph from El Universal in a post below. Granted, it's not a mathematical proof, but it's a strong indication that the shift is independent of the performance of other candidates.
But Giordano tells us what the end point of Madrazo's vote was: "During this count of the final 35 percent of the tallies, interestingly, PRI candidate Roberto Madrazo’s percentage remained steadily the same as it had all day (within half-a-percentage point, landing at 22.26 percent)." The endpoint is a verifiable fact.
If one examines the graph carefully, one sees that the data at 87% is essentially identical to the endpoint. The extrapolated point shown on the grapg is approximately 34.1 or 34.2, representing a final count of 22.1 or 22.3%. This is so close to 12+ 22.26 = 34.26 as to be indistinguishable.
Taking the average of the performance in the two northern districts, as 23.5% or so, Giordano argues that in the equation 0.87 x + (0.13)(22.3) =22.26, x isn't and can't be consistent with Madrazo's totals in the rest of Mexico.
It's a h--l of a complicated argument. He should not only show the data, he should show what the limits on x are. But for me, all I need to see is that Obrador's losses are perfectly correlated with Calderon's gains. It's just too d--n neat.
Just substitute "Gore" for "Lopez Obrador" and "Bush" for "Calderon". Any effort by the lefty candidate to assert his rights is howled down by the US media as "contributing to the atmosphere of burgeoning chaos that threatens the stability of the nation."
Most responses have not been polite or rational. I'm not sure whether Daisukechul was speaking about me or about AMLO when he said, "What a piece of garbage this person is. Besides being a fool, he's stubborn, ignorant, and a liar. How is it possible with all the proof, he refuses to accept what he said in front of millions of us who saw these television watchers?" Other comments were similar.
Those PANistas, they got no respect.
WaPo disgraces itself, and sends a very strong signal about the nature of its future coverage.
First, the endpoint of the green line at 100 percent is supposed to represent the actual PRI percentage (plus 12); however, the endpoints of the red and blue lines at 100 percent cannot represent the actual percentages of Lopez Obrador and Calderon. This appears inconsistent. The main problem is that I am still not clear exactly what the solid lines are intended to be.
Second, the slope of the green line seems too small to fit the data between about 62 and 87 percent, although the line does appear to fit the whole of the data in a least-squares sense. Again, it would help to know exactly what the lines are.
In any event, the allegations of vote fraud do not depend on one single graph; I fear that I've belabored the issue far more than it's worth. Thanks for your patience.
Not at all, Andy. These are good questions, and if there's an error or even unclarity, I'm sure Al Giordano would much prefer to learn about it and correct it on his own.
Andy: "the endpoints of the red and blue lines at 100 percent cannot represent the actual percentages of Lopez Obrador and Calderon."
Those endpoints look by eye to be 35.3 and 35.7%. In the link I gave above, as of 8AM on the morning of the count, it says, "With all of the 41 million votes counted, Calderon of President Vicente Fox's National Action Party had 35.88 percent to 35.31 percent for Lopez Obrador, of the Democratic Revolution Party." So I think the endpoints are correct.
Andy: "The main problem is that I am still not clear exactly what the solid lines are intended to be."
They're trend lines, drawn to help the eye see variation around the trends. As far as I can tell, no data were issued before ca. 25%; this is seen in the detailed figures given in a post I presented on the night of the election below. Giordano says, "in the final stretch only Obrador and Calderón percentages diverged from the consistency of the first two-thirds of the tallies..." so we can assume these lines are drawn using data from the 25- 66% of the vote count, probably as least squares, but it's not important.
What the graph does show is that the data in the 25-66% range were drawn from a very homogenous set. They don't show the large seesaws that we see in our presidential elections as, say, the very red Mountain states come in and then California. indeed, the shifts in, say, Madrazo's data are in the tight range of about 21.7-22.3% from 25-100% vote count. The graph also shows that the data for AMLO and FeCal in the range above 66% count come from a statistically different set of data.
Unraveling what happened in the count isn't that important. It's theoretically possible to select precincts to produce the result observed. (The techical phrase, I beleieve, is "an underdetermined system.") In such a case, IFE would simply be guilty of news manipulation, of arranging to have the results change in the dead of night specifically to dismay AMLO's supporters and get the shocking outcome on the early morning broadcasts, for example. What will expose the fraud is the testimony of witnesses and the examination of the ballots-- assuming FeCal can't manage to substitute fakes in the worst precincts.
But if the IFE is engaged in news manipulation to benefit one party, are they really the squeaky clean organization that we keep getting told they are? And if they're not, what other hijinks might they have committed?
What we need to do is keep asking questions until we find one they can't answer. But to do that, we also need to challenge our own assumptions. As you so commendably have done. And are more than welcome to continue doing.
In their "balanced coverage," they leave out that Calderon brother-in-law Hilebrando's software was already been hacked.
They print IFE's denial that Hildrebrando had a contract at all without commment, even though the contracts are a matter of public record.
Good work, Charles, and all who have pursued this issue.
Mexican politics have been corrupt for generations, my family comes from there and immigrated out.
It's never been a great place to live because the political scene shifted from month to month. However if I remember correctly wasn't everyone happy when Vincente Fox won and there was a peaceful transition to his party, wasn't everyone really happy? Now you are worried that there might have been election fraud. I don't remember anyone worrying about that when he came in, but now you are worried about it? concern yourself with your own country and stop trying to stir up a civil war funded by drugs and leftists
If you wanna talk drugs and dictators, there's evidence linking your favorite right-wing killer (next of course to Adolf Hitler), Augusto Pinochet, to the cocaine trade.
More blogs about politics.