Evidence of Election Fraud Grows in México
August 02, 2006 | 12:00AM ET
A month after more than 41 million Mexicans went to the polls to elect their next president, the country is still awaiting a result. A preliminary count of polling station tally sheets put conservative Felipe Calderón of the National Action Party (PAN) ahead with a slight lead over left-populist Andres Manuel López Obrador of the Democratic Revolution Party (PRD). Both candidates have claimed victory, with López Obrador and his supporters holding vigils and protests across the country and calling for a vote-by-vote recount.
That hasn't kept a consensus from emerging in the commercial media that Calderón won by a small margin in a squeaky-clean election. In a hyperbolic editorial on July 30 -- one that bordered on the ridiculous -- the Washington Post accused López Obrador, known as AMLO to his supporters, of taking "a lesson from Joseph Stalin" and launching an "anti-democracy campaign" by demanding a manual recount and urging his supporters to take to the streets in peaceful protests. Calling the vote "a success story and a model for other nations," the editors concluded that it's "difficult to overstate the irresponsibility of Mr. López Obrador's actions."
Days after the election, the New York Times irresponsibly declared candidate Calderón the winner, even though no victor had been declared under Mexican law, and just this week, in an article about López Obrador's protests, the Times reported that López Obrador had "escalated his campaign to undo official results."
But there are no "official" results and probably won't be until after Sept. 1. Under Mexican law, the Federal Electoral Institute (IFE) is charged with running the elections and counting the vote. But only the country's Election Tribunal, known by its Mexican nickname as the "TRIFE," has the power to declare a victor (See here for background on the TRIFE). They have until Sept. 6 to rule on the election.
It appears that the U.S. media has become so enamored with the construct of the "anti-democratic" left in Latin America -- the ubiquitous "fiery populists" (a term that has described everyone from the centrist Lula da Silva to Hugo Chávez) -- that they are incapable of fulfilling their basic mandate to inform their readers when it comes to the political landscape south of the border. It's nothing short of journalistic malpractice.
But back in the real world, a growing body of credible evidence from mainstream Mexican journalists, independent election observers and respected scholars indicates that an attempt was made to deliver the presidency to Calderón. It includes a pattern of irregularities at the polls, interference by the ruling party and some very suspicious statistical patterns in the "official" results.
The TRIFE is now sifting through 900 pages of formal complaints lodged by López Obrador. Their ruling on those challenges will indicate how well México's electoral process holds up in a closely fought and highly polarized race.
Growing evidence of irregularities and fraud
México has a history of the party in power's using its clout to tip the election in its favor, and strict laws prohibiting ruling party interference were enacted in the 1990s. Election law prevented Vicente Fox, the outgoing PAN president, from making public statements of a partisan or political nature. But he overstepped this line many times in the 2006 campaign, including dozens of speeches reinforcing candidate Felipe Calderón's basic message that López Obrador was a "danger to México." In a well-publicized speech, candidate López Obrador responded, "With all respect, Mr. President, shut up. You sound like a chattering bird." Fox continued with these speeches until election authorities and public commentators warned Fox he was violating election laws.
The Fox administration also ran public service announcements touting government programs and services and promoting the vote. PAN saturated the television airwaves with "swift-boat" style attack ads against López Obrador, comparing him to Venezuela's Hugo Chávez and calling him a "danger to México." Election authorities eventually ordered these commercials off the air on the grounds that they were untrue and maligned the candidate's character, but critics believe they moved too slowly.
Under Mexican law, ruling party interference is a serious charge and grounds for annulling an election. In the last ten years, the same Electoral Tribunal judges that are reviewing AMLO's complaints annulled governors' races in Tabasco and Colima, based on ruling party interference. The Institutional Revolution Party (PRI), which ruled México for seven decades before the system was reformed in the 1990s, made vote buying and voter coercion into a high art form, and there is strong evidence that they were up to their old tricks in the 2006 election. With PRI governors in 17 of México's 31 states, election observers documented a significant number of examples of voters being offered money or receiving food or building materials in exchange for their PRI vote. In a country where half the citizens live in poverty and rely on different forms of government assistance, voters are often told that their public assistance is dependent on voting for the party in power. There are examples of PAN using similar practices, especially a well-documented case of funds diverted from a San Luis Potosi building program into PAN electoral races.
The Mexican electoral system has come a long way in two decades in implementing anti-fraud systems. But there are still several ways that results can be tampered with on election day. López Obrador's campaign and hundreds of independent election observers documented several hundred cases of "old fashioned" election-day fraud in making their case for a recount.
Here's how the system was supposed to work. On July 2, Mexicans voted at over 130,000 different polling stations, casting separate ballots for president, senator and federal deputy. Each political party was encouraged to have registered poll watchers at every polling station to observe the voting process and count at the end of the day. As international and Mexican election observers noted, however, problems emerged when there weren't enough independent and party observers to go around. In regions where one party was dominant, this created opportunities for vote shaving, ballot stuffing, lost ballots and other forms of fraud.
The PRD's strongest case for a recount comes from the fact that ballots in almost one-third of the country were not counted in the presence of independent observers. One analysis of IFE results found that there were 2,366 polling places where only a PAN observer was present. In these districts, Calderón beat López Obrador by a whopping 71-21 margin.
Other elements of PRD's legal challenge include documentation of several ballot boxes found in dumps in the PRD stronghold of México City. They also point to evidence such as the nonpartisan Civic Alliance's report documenting 17 polling sites in PAN-dominated Nuevo León, Michoacan and Querétaro, where the number of votes cast vastly exceeded the number of registered voters at a site.
Reports by international and domestic election observers affiliated with the Civic Alliance and Global Exchange stop short of claiming fraud in the elections. They laud the dedication of most poll workers they monitored and the preparations for the vote in most of the polling places, as well as the orderly and peaceful process overall. But the cumulative evidence is damning in such a closely contested race.
In the weeks after the election, PRD observers again sounded the alarm as sealed ballot packets were being illegally opened at IFE district offices in several PAN-dominated regions. PRD officials accused IFE officials of possibly tampering with ballots or attempting to cover up fraud in the event of a recount. The TRIFE ordered these offices to stop opening vote packets.
While the López Obrador campaign has not made major charges of "cyber fraud," there is an emerging controversy over the IFE's role in reporting who was ahead in the vote count. For the 2006 election, the IFE had developed a sophisticated system to provide preliminary results called the PREP. Relying on results being phoned in from a sample of precincts, the IFE could compile a credible picture of the vote. If the PREP showed one candidate with a clear majority, the system would have allowed Mexicans to go to sleep on election night knowing who their next president would be. But because of the razor close results, the PREP proved to be an inadequate measure.
Now research is emerging to suggest that the PREP results were cooked to create the appearance of a Calderón victory. Physicist Jorge López at the University of Texas, El Paso, conducted a statistical analysis of the PREP results and found that, as the results came in, the differential between the candidates' totals remained almost constant. One would expect that, as results from each party's geographic strongholds were counted, the gap between their totals would rise and would fall. In such a tight election, one would even expect the lead to change back and forth as the count progressed. None of that happened. The results of a third candidate, Roberto Madrazo of the Institutional Revolutionary Party (PRI), fluctuated as expected.
He also noted that there was very little deviation between the actual results as they came in and the average results; in a normal, natural distribution, one would expect significant differences between the two (it should look something like a squashed bell-shaped curve). Dr. López concluded the pattern was "a clear indication that the data was manufactured by an algorithm and does not stand a chance at passing as data originated at the actual voting."
Luis Mochan, a physicist at the Universidad Nacional Autónoma de México, did similar work. He noted that the PREP data was posted after the first 10,000 reports had been processed, and looked at whether those first 10,000 reports were consistent with the statistical trends for the rest of the day. When he plotted the data backwards, Calderón's vote total originated at zero, as is normal, but López Obrador began the day 126,000 votes in the hole.
Mochan and López both point out that the Calderón began the day building a large percentage lead -- seven points -- that decreased steadily throughout the day. The large early lead would have been handy from a psychological and political perspective, allowing Calderón to claim that he led all day long, but the results had to end in a close result given that polls conducted a week before the tally showed a statistical dead heat.
Mochan also notes gross discrepancies in the number of votes processed late in the evening: "At the end of the plot, we find intervals with more than 1,200 votes per [voting] booth. I understand that no booth was to receive more than 750 votes. Even more worrisome, some data points indicate a negative number of votes per booth."
Mochan notes that these statistical anomalies aren't definitive proof of anything. But economist James Galbraith, reviewing Mochan's data, speculated about a likely scenario that would fit the discrepancies seen that night:
Felipe Calderón started the night with an advantage in total votes, a gift from the authorities.
As the count progressed, this advantage was maintained by misreporting of the actual results. This enabled Calderón to claim that he had led through the entire process -- an argument greatly repeated but spurious in any case because it is only the final count that matters.
Toward the end of the count, further adjustments were made to support the appearance of a victory by Calderón.Critics suggest that the IFE may have aggressively pushed to swiftly declare Calderón a victor, obviating the need for a poll-by-poll vote recount.