Bai-a-lie

At the end of this year, the mainstream media narrative regarding exit polls has been all but sealed: they are not the reliable gauges they once were. In fact, mainstream efforts have centered around discrediting exit polls when their results indicated that something untoward might have taken place in a given election, especially in an election in which the US government and the corporate establishment have vested interests and where the results have been -- or might have been -- less than desirable.

Lately the move has been to keep exit polls results entirely hidden from public scrutiny. We saw this in the recent mid-terms elections. No commercial networks discussed their exit polls on air, or even that they had exit poll data. It turned out the CNN did publish exit poll data on their website, which were slurped up by the Election Defense Alliance. The data, as one might expect, showed significant discrepancies with the votes totals -- some 3 million Democratic votes went missing just as they had in 2004 -- but this was never mentioned. Warren Mitofsky, inventor of the exit poll, explained his rationale for having Mexican television networks keep exit polls a secret: things were "too close to call." Mitofsky, of course, had an interest in protecting the validity of his exit polls while refusing to acknowledge that they might indicate something amiss with certain elections -- at least with elections in the US and Mexico. No one had any trouble citing exit polls in the Ukraine when those polls indicated election fraud. But in North America, exit polls transmogrified into statistical trouble-makers. Americans must be a strange lot, with Republicans especially skittish about telling pollsters what their preferences were. After the 2004 elections, elections that left 3.6 million votes uncounted, any number of ad hoc explanations surfaced about why the exit polls were "wrong." In light of the huge number of uncounted votes -- 240,000 in Ohio alone -- these explanations were shown to be entirely specious, but they were marshalled onto the pages of the mainstream press while the same outlets never bothered to report the number of uncounted votes.

Which is why I found Matt Bai's brief yet entirely mendacious load of bullshit in the pages of The Times Magazine a fitting year-end wrap to the mainstream media's enduring campaign against any discussion of the obvious problems of fraud seen recent elections in the US and Mexico. Bai, long a "liberal" apologist for the corporate establishment, delivers a paean to Warren Mitofsky while he further adopts and reinforces the mainstream electoral narrative we have all grown to loathe:

Of all the political customs that came under scrutiny after the calamitous election night of 2000, perhaps none were as widely discredited as the mystical science of exit polling. Furious Republicans screamed that inaccurate exit polls had dissuaded some Florida conservatives from bothering to vote. Confused Democrats, meanwhile, demanded to know how polls that proclaimed Al Gore to be the winner in that state could have been so disastrously wrong.

In the world of the liberal apologist, exit polling entered the realm of mysticism in 2000. I doubt that Bai is as clueless as this passage would indicate and I suspect he has curled up into a comfortable ball of credulity that would have toadies like Bai believe that "it can't happen here." Bai appears to be unaware of the fact that the media organisation recount of the Florida 2000 ballots -- beginning to show that Gore won the vote -- was suspended as the uncomfortable truth of the vote began to emerge. He also appears to be unaware of the tens of thousands of Democrats illegally purged from Florida voter rolls. I say "appears" because we have to know that Bai simply cannot be this ignorant. If he is, he has no business writing about elections anywhere.

What I am saying? Of course he has business writing about elections, no matter how disastrous they were. It is Bai's job, one among the throngs of mainstream hacks, to impart legitimacy upon our deeply flawed elections. Despite the vast body of evidence, admission of fraud in American elections is tantamount to heresy. It was exit polling that was "misleading" and not the armies of political operatives diligently suppressing voter turnout if not simply throwing votes away.

It is the last paragraph, however, with which I take the greatest issue in this utterly ridiculous, factually challenged article:

The network Televisa, acting on Mitofsky’s advice, refused to project a winner on the night of the election, saying that the race was too close to call. The exit polling, however, predicted that Felipe Calderón had narrowly beaten his opponent, Andrés Manuel López Obrador. And, much to Mitofksy’s relief, that was precisely what had happened.

There is almost nothing in this paragraph that is true. Not only is this not "precisely what had happened," it barely resembles the circumstances of Mexico's election at all.

While Mitofsky may have advised Televisa, Interior Ministry officials approached the television networks and told them to keep exit polls off the air. But the greater sin in this dreadful bit of disinformation is the statement that the exit polls matched the final vote count. This is a stinking, reeking lie. As Reuters reported on July 3, exit polls, at least ones that were not massaged after the fact, showed Obrador with a 2+% lead throughout the day. This was matched by the final tally until about 75% of the votes had been counted, when suddenly the last remaining precincts reported voting ratios of, in some cases, 100 to 1 in favour of Calderón. Video of election workers stuffing ballot boxes sailed across the internet. While Mitofsky himself claimed that his exit poll showed that Calderón would win by a margin of 200,000 votes and that the final margin was 200,000 votes, this appears to be nothing but a post hoc lie. Calderón's initial margin started out at 4.5% and decreased steadily throughout the PREP count until the gap steadied at 400,000+ votes. This would have been the final winning margin had it not been for Obrador and his supporters demanding the IFE release and count some 3 million votes, which the IFE initially claimed were "lost," that the gap narrowed to 250,000+ votes, which narrowed yet again in the post-election partial recount. I take it from this that, had the initial 400,000+ vote margin remained, Mitofsky would have been more than willing to adjust his exit poll data to match that count as well. In secret, post hoc exit poll manipulation, you can do anything damn thing you like. Mitofsky also claimed that that the ballots were "counted" twice. This, too, is simply wrong.*

So what the hell is Mitofsky talking about? Exit polls did not show Calderón in lead, ever, but that wouldn't have been convenient to admit. The PREP count did, but this did not match the exit polls, at least it didn't match the ones that weren't kept secret. But then, when you keep everything secret, you can pretty much say any shit you want about the accuracy of your polls. Before he died, Mitofsky set out to vindicate his methodology despite the fact that it was not the methodology of the exit polls that was the problem. Fixing the facts around a policy of election fraud was the problem and happy idiot Matt Bai spills this drivel across the pages of the New York Times once again. But then, Bai is an apologist for Wal-Mart, so chances are he's not too concerned about the legitimacy of elections in Mexico. Nor, would it appear, does he have much concern about the reality of elections in this country.



* If you want to find out what really happened during the vote counting in the Mexican elections, read What went wrong in Mexico, The Humanist Magazine, Nov-Dec, 2006. (.pdf file)