By Paula Bourges-Waldegg

They say that Kafka would have been a realist writer if he had been born in Mexico. The events following the 2006 presidential election seem to confirm this theory.

Last September, in the most contended presidential election in the history of modern Mexico, Felipe Calderón, the candidate of the conservative National Action Party (PAN), was declared winner by a narrow margin after a long and complicated judicial process. The runner-up, Andres Manuel López Obrador (AMLO) of the left coalition led by the Democratic Revolution Party (PRD), challenged the results, accused his opponents of cheating, led huge demonstrations, and proclaimed himself Mexico’s “legitimate” president before a large crowd at El Zocalo, the capital’s central plaza. Thus, Mexico has now not one, but two presidents ruling over the same territory in parallel universes.

AMLO supporters still claim that the state’s democratic institutions have been completely shattered and that a new order should be re-erected with their leader’s guidance. Clearly, the fear that AMLO still generates among a large portion of the population has more to do with his scorn for institutions and the law, than for any of his economic views. He certainly cannot be defined as center-left. He is a populist whose political compass seems to be closer to an authoritarian revolutionary left than to a liberal democratic one. In fact, it’s possible that a large number of people who voted for Calderón did so not because they liked him, but because they suffered from AMLO-phobia.

I don’t support Calderon, or his party, or any other conservative for that matter. But I respect Mexico’s long and arduous effort to build true democratic institutions. That’s why it’s sad to see how these fraud allegations, mostly based on conjectures, speculations and extrapolations are undermining this effort and damaging Mexico’s image internationally. Thus, the purpose of the following lines is to show why the real fraud in Mexico is to claim there was a fraud.

Back to the Past

To compare the 2006 presidential elections with the fraud in 1988 reveals a complete ignorance about how the Mexican electoral system actually works. Back in 1988, when the official candidate Salinas de Gortari “won” the presidency, the Institutional Revolutionary Party (PRI) controlled everything. There were no independent electoral institutions, no human rights commission and no free press. Since the fraud in 1988, a new electoral system was slowly and painfully built thanks greatly to the ostensible winner of that election, Cuauhtémoc Cárdenas, founder of the PRD.

With the approval of all the political parties, dozens of security devices were implemented expressly to prevent a fraud such as the one in 1988. Some examples include, an autonomous Federal Electoral Institute (IFE), an independent Electoral Tribunal (TEPJF), reliable and regularly updated name registers containing photos, and voting photo-IDs. Also, tally sheets and ballot papers have several security features such as folio, watermarks and fine-line printing patterns, and the latter are handed out only to voters who have an ID that matches their precinct’s register.

This system is uniform across the country and one of its main features is that the election process is almost entirely conducted by citizens. They are chosen randomly to chair the precincts corresponding to their address. Their responsibilities include, verifying the identity of voters, counting manually the votes and filling the tally sheets of the precincts they chair. To pull off a fraud in 2006 it would have been necessary to coordinate, buy and coerce thousands, if not millions, of people. That makes this theory not impossible, but highly improbable.

However, some things never change. The culture of distrust that exists among Mexicans is one of them. A poll made in September revealed that 39% of the population believed there was a fraud (1). A long history of corruption cannot be easily left behind. Thus, no matter how improbable, a fraud seems to be always the most believable explanation.

Voto por Voto

“Vote by vote, precinct by precinct”, was AMLO’s campaign slogan against the supposed fraud. At first glance, the petition for a complete recount sounded perfectly reasonable. But the request was misleading for several reasons. In the first place, AMLO’s party coalition only presented cases formally in court for 11,839 of the 130,477 precincts. So, while on the streets AMLO and his supporters were demanding a total recount, legally they could only justify a partial one. Clearly, they knew that the former would be rejected and that this would make the electoral institutions look as if they were hiding something.

In second place, this slogan makes everything look as if votes were never counted one by one when, in fact, every single vote in every single precinct was manually counted by citizens the same day of the election. What's more, AMLO’s party coalition had delegates watching them in almost 80% of the precincts (3) and only a tiny proportion of the tally sheets were signed under protest.

The ballots were added up three times during the whole electoral process, always with virtually the same result: The first, the day of the election in a preliminary tally; the second, three days later in the official District Computations where ballots from disputed precincts were recounted before members of all parties; and the third, in the final and definitive computation conducted by the Electoral Tribunal after a partial recount ordered to attend the indictments made by AMLO’s coalition. A hundred and ninety two judges -not electoral officials- carried out this partial recount. Party representatives also attended these sessions and were allowed to challenge the results and put down their objections in the official records, which were then sent to the Electoral Tribunal for their review. Rulings over specific precincts are available in the search section of the Tribunal’s webpage.

So, it appears that AMLO’s strategy wasn’t aimed at obtaining a judicial order for a total recount, which he knew was legally impossible, but at spreading doubts and gaining popular support through a cause that seemed fair so that he could force the electoral institutions and his opponents to negotiate the results.

Summer Trends

The IFE has two different calculation tools that are commonly mistaken. One is the “Rapid Count”, which is basically an exit poll, and the other is the “Preliminary Results Program” (PREP), which is mainly a public access tally. The aim of the former is to anticipate who’s the winner, while the purpose of the latter is to make the computation process transparent and provide quick preliminary results. The PREP is available through the Internet and constantly updated so that anyone can see how the votes are processed and counted as they arrive in real time to the data center from each electoral district. While the PREP counts the actual votes, the Rapid Count is based on a representative sample and makes projections from trends.

It has been argued that trends in the PREP clearly showed that AMLO was the real winner. Felipe Calderón, initially led by five points, but with each new update the tally of votes for AMLO rose steadily, while Calderón’s sank in equal proportion. By 1.20 am, the difference was 1.4 %. AMLO supporters claim that according to those trends, his candidate should have been in the lead by 4 am (4).

However, since the PREP is not a calculation based on statistical estimates or a projection from a given sample, it is illogical to talk about trends in this sense. As the PREP records the real set of votes, the data flux doesn’t affect the final result (5). That is, no matter the order in which you capture the information you’ll have basically the same sum at the end. Moreover, because the vote distribution in each precinct is different, it is impossible to frame a representative sample at any given moment of the progression (5). Therefore, trends are not stable and cannot be projected.

Nevertheless, the behavior of the data can be analyzed and explained. If we consider that tally sheets coming from urban places typically arrive faster than those from rural areas with poor communications, it’s not surprising that Calderón, who had a larger urban vote, led at the beginning. The margin started to shrink as the results from rural places, where AMLO came first, arrived (6). The official District Computations, performed three days later, yielded mainly the same results even though the progression of the data occurred in reverse order. That is, AMLO led at the beginning because the ballots of the states where he won were registered first, and Calderón took the lead later as the states where he won were added up (6).

However, the most controversial issue about the PREP is, perhaps, that some tally sheets were kept apart from the main database (2,800 in total). AMLO’s party coalition accused the IFE of hiding information. A very dishonest allegation as a few months before the election all parties agreed to put precincts with anomalies in a special file to be corrected later during the District Computations. The IFE never announced publicly that this file existed and was available in their website, a mistake that gave AMLO the perfect ammunition to sow the seeds of mistrust.

The Silence of The Networks

Another factor that contributed to spread doubt among the population was that the winner wasn’t immediately announced as in other occasions. The day of the election the two major TV networks, Televisa and TV Azteca, announced they would not reveal their polls. Later, the chairman of the IFE, Luis Carlos Ugalde, appeared on national TV to say he would be withholding the Rapid Count results.

Ghosts of 1988 combined with ignorance about statistics conspired against their credibility. It seems that the vast majority of the population in Mexico didn’t understand why polls are not a reliable way to predict results when the margin between two candidates is so narrow. The margin of error in polls, depending on the size of the sample, is between 1% and 3%, and the difference between the candidates in this election was less than 1%. That’s why, as it would have occurred in any other democracy, the winner couldn’t be announced right after the election. Ugalde and an independent technical committee explained all this on national TV to an already skeptical audience. But the IFE and the media acted responsibly as their credibility would have been much more affected by having to rectify the announcement of a winner.

Even though days later the Rapid Count was made available and revealed mainly the same results as the PREP, the District Computations, and other exit polls from diverse independent sources (7), the harm was already done and the initial silence was perceived as “evidence of the conspiracy”.

Law and Disorder

It has been said that given the widespread doubts about the validity of the results, the Electoral Tribunal (TEPJF), an independent branch of the federation’s judicial power, had enough legal grounds to take any measures necessary to restore public confidence in the election’s outcome and avoid further disorder. Of course, according to AMLO and his supporters, the only thing that could have restored public confidence was a total recount followed by the annulment of the whole election. They claimed that it was not just an option for the magistrates but their constitutional duty.

  

The members of the TEPJF are well-respected judges approved in congress by all parties, including the PRD. Their impartiality has been more than proven along the ten years they’ve served in this court, during which, they have applied principles consistently and ruled equally against and in favor of all parties. They couldn’t have initially ordered a total recount, because, as mentioned earlier, AMLO’s party coalition presented cases for only 11,839 of the 130,477 precincts. However, they did ordered a partial recount to rule over these disputed precincts.

According to AMLO supporters, this partial recount showed that out of 11,839 precincts, 7,442 either had ballots missing or ballots above the number of people who had voted there (4). Nevertheless, assuming these numbers are real, both types of incidents are not necessarily evidence of fraud. Considering that the average schooling in Mexico is 8 years, human error is a simpler explanation. In fact, some common mistakes include introducing votes for congressmen in the president ballot box and vice versa, adding up the votes incorrectly, leaving spaces in blank in the tally sheets or misspelling the candidates’ names. Furthermore, the inconsistency between number of votes and voters in a precinct is very easy to detect and correct within the current electoral system. Actually, precincts with these characteristics are not even counted in the PREP. So, if AMLO supporters were right, this would be an extremely lousy method to commit fraud.

Of course, they were wrong because after the partial recount all candidates were almost equally affected, (minus 81,080 votes for Calderón, minus 76,897 votes for López Obrador). Thus, most irregularities were attributed to random errors and the judges found no reason to suspect any widespread manipulation of votes. With no clear pattern of fraud, the TEPJF couldn’t annul the 11,839 precincts and the result stayed mainly the same.

But then, AMLO supporters argued that extrapolating nationwide the votes that Calderón lost in the recount, their candidate would have won the election by more than one million votes (of course, they didn’t extrapolate the votes lost by AMLO as well). This argument is absurd because without a clear pattern of fraud these numbers cannot be generalized. Besides, only the precincts where Calderón won were recounted, so it is impossible to know if there was any bias in favor of AMLO in those where he came first. In fact, there were many precincts in Chiapas where AMLO won every single ballot, a typical indicator of coercion (8).

AMLO’s party coalition also made a petition for the annulment of the entire election based on irregularities that occurred during the campaign, which they claim, affected their candidate. President Fox’s sloppy attempts to promote his candidate, negative advertising portraying AMLO as dangerous, supposed unbalanced media coverage and the private sector’s endorsement for Calderón, are some examples.

The TEPJF recognized that some of these incidents were not acceptable, but concluded that there were no objective measures of their impact, and that all parties indulged in this kind of behavior (9). The attacks AMLO received during the campaign were as nasty as those he perpetrated against his opponents, his party spent more money in the media than Calderón ’s (57.1% of a total budget of 628,882,943 pesos of which 99.6% was used in TV and radio vs. 39.8% of a total budget of 555,866,538 pesos of which 98.9% was used in TV and radio) (10) and, according to a report on media coverage (11), AMLO had more exposure than Calderón in the news (TV 21% vs. 19% and radio 25% vs. 18%). In addition, AMLO used his position as major of Mexico City to campaign for nearly six years, an advantage that no other candidate had.

Thus, the total recount and the annulment petition didn’t prosper because AMLO’s party coalition couldn’t provide the necessary evidence to prove a fraud beyond reasonable doubt. The accusing part is the one obliged to prove their accusations, not the judges. Since the presumption of validity is a principle that applies in these cases, the TEPJF needed more than mere conjectures, speculations and extrapolations to invalidate a popular mandate of such relevance. So, the magistrates didn’t disobey their constitutional duty. On the contrary, by refusing to cancel the votes of millions of citizens without a justified cause, they provided the legal certitude that should have restored public confidence in the election.

---

It’s not yet clear if the Mexican left is going to choose reform rather than revolution as a means of change. The escalation of local conflicts, like the one in Oaxaca, is not a good sign. However, as AMLO’s fellow party members take charge of their newly conquered places in congress and local governments, it looks like they will have no option but to recognize the legitimacy of the institutions they serve and consequently accept the results of the election.

Not long ago in Mexico, the left was completely marginalized and political conflicts could only be solved through backroom deals or by the use of force. There’s still a long way to go, but if this difficult election has proven something it is that, despite their many weaknesses, Mexican institutions actually work. And that is exactly what gives legitimacy to a president, not a big crowd packed in a plaza. So, if AMLO’s followers decide to leave for good their leader’s parallel universe and start building bridges with the new government, Mexico might not be that Kafkaesque after all.

References

(1) Zavala. D. (2006) “Opinan que Calderón debe llamar al diálogo” Nota periodística, El Universal, Lunes 04 de septiembre de 2006

(2) Aparicio J., (2006) Representantes de casilla y voto presidencial: un análisis preliminar, División de Estudios Políticos, CIDE, Julio 30, 2006.

(3) Distribución de casillas reportadas como instaladas y que contaron con la presencia de representantes de partidos políticos o coaliciones, por entidad federativa PROCESO ELECTORAL FEDERAL 2005-2006, Reporte IFE

(4) Giordano, A. (2006) Mexico’s Presidential Swindle, New Left Review 41, September-October 2006

(5) Ragasol, E., Cárdenas, C., Muedano, L., Espinosa De Los Monteros, E. (2006) Elecciones Presidenciales 2006: Análisis De Los Resultados Electorales Del PREP.

(6) Aparicio, J. (2006) El Algoritmo Improbable. División de Estudios Políticos, CIDE.

(7) Aparicio, J. (2006) Public Economics, Ideas In Political Economy, Applied Econometrics, And Public Finance (blog).

(8) Riveros, H. & Cabrera, E. (2006) Análisis De Los Datos De Las Elecciones Para Presidente 2006.

(9) Dictamen Relativo Al Cómputo Final De La Elección De Presidente De Los Estados Unidos Mexicanos, Declaración De Validez De La Elección y De Presidente Electo.

Comisión Dictaminadora: Magistrados Alfonsina Berta Navarro Hidalgo Y Mauro Miguel Reyes Zapata, TEPJF, septiembre 2006

(10) Herrera, J. (2006) “IFE: Gasto De Campañas En Medios Creció Más De 100%”, Nota, El Universal, Domingo 08 de Octubre de 2006.

(11) Análisis General De Los Resultados Del Monitoreo De Noticias Del 19 De Enero Al 30 De Junio De 2006, Reporte IFE

Subscribe to ChiliConDarwin weekly updates