Controversies of the 2006 Mexican general election


The Mexican general election of July 2, 2006, was the most hotly contested election in Mexican history and as such, the results were controversial. According to the Federal Electoral Institute, the initial "Quick Count" determined the race was too close to call, and when the "Official Count" was complete, Felipe Calderón of the right-of-center National Action Party had won by a difference of 243,934 votes. The runner-up, Andrés Manuel López Obrador of the left-of-center Coalition for the Good of All, immediately challenged the results and led massive marches, protests, and acts of civil resistance in Mexico City. On August 9, while protests continued to expand, a partial recount was undertaken by election officials after being ordered to do so by the country's Federal Electoral Tribunal. The tribunal ordered the recount of the polling stations that were ruled to have evidence of irregularities, which were about nine percent of the total.
On September 5 the tribunal declared that Felipe Calderón met all the constitutional requirements in order to be elected, and was declared president-elect. Some civil resistance acts led by Andrés Manuel López Obrador were maintained in an attempt to encourage a change in the country's opinion, as well as other activities such as a documentary by Mexican filmmaker Luis Mandoki.

Elections and results

On Sunday July 2, 2006, Mexico held its first presidential election of the post-PRI era. Media polls had predicted a very tight race.
To understand some of the allegations and controversies, it will be useful to know some of the details of how the votes were counted and how the partial results were reported.

Vote counting

Since the 1997 federal elections, elections in Mexico had been overseen by the Federal Electoral Institute, known by its Spanish acronym IFE. Prior to 1997, elections were run by the federal government's Secretaría de Gobernación. The law governing elections was the Federal Code for Electoral Institutions and Procedures, the COFIPE. The COFIPE was amended after the election of 2006; references to articles in this section refer to the COFIPE as it stood during the election in 2006, unless explicitly noted otherwise.
Under the law then in force, each polling place was manned by randomly selected citizens from the corresponding district, who formed the Mesa Directiva. They received a training session from the IFE prior to the election. In addition to the polling place officers, each party, coalition, or alliance participating in the election could name up to two observers and one alternate for each polling place.
In order to vote, citizens were required to present a valid credencial para votar con fotografía. The card was checked against the official voter list for the polling place, which contained facsimiles of each card. The citizen was then given the blank ballots and directed to the booth so he or she could mark them secretly; if a person could not read or was physically impaired, they could be accompanied by a single person they themselves designated. The citizen then folded the ballots, and deposited them in the corresponding ballot boxes. Then, the Secretary of the Directing Committee would write "voted" in the entry corresponding to the citizen in the official voter list, the identification card was marked and returned to the voter, and their right thumb marked with indelible ink. The party and coalition representatives voted in the booth to which they are assigned, and were logged in special entries in the voter list. During the voting, the party representatives could file complaints in writing, which were received by the Chair and logged by the Secretary. A specified number of "special voting places" were also installed for people voting outside their polling place. Their thumbs were inspected before allowing them to vote, and the data logged. They were allowed to cast votes in all elections corresponding to both the location of the special voting place and their own registry; for example, if they were voting in their state but outside their district they were allowed to cast votes for senator and president, but not for district deputy.
Voting stopped at 6 pm or whenever anyone who was in line to vote at 6 pm was done voting ; the poll could be closed earlier only if all members of the Directing Committee agreed that all registered voters had cast their vote. Once the polling place closed, the Directing Committee was in charge of counting and recording the votes, in the presence of the party representatives. After the counting was over, the totals for each candidate, the total of write-in votes, the total of annulled votes, and the total of left-over blank ballots were recorded in the Acta de Escrutinio ; in addition, any incidents that occurred during the voting or counting would be described in writing. Finally, every member of the Directing Committee and every party representative present were required to sign the tally sheet; the latter were allowed to sign under protest and explain their reasons, if they so desired.
A file was then made for the polling place. The file contained a copy of the log, a copy of the signed tally sheet, and a copy of any protests filed. Separate packets contained the unused ballots, the annulled ballots, and the valid votes. A separate envelope contained the list of voters. Finally, all files were placed together in a sealed packet; the sealed packet was then signed by every member of the Directing Committee, and any party representative who so desired.
Each party representative then received a copy of the official tally sheet; the representative was then required to sign a proof of receipt. On the outside of the sealed electoral packet a transparent envelope containing a signed copy of the official tally sheet was placed. The chair of the Directing Committee then placed an enlarged copy of the final tallies in a visible location outside of the polling place; this copy was supposed to be signed by the Chair, and could be signed by any other members of the Committee and party representatives who so desired. Finally, the polling place was then officially closed, and the Chair of the Directing Committee delivered the sealed packet and documentation to the District Office, where the official count began some days later. Polling places next to the District Office were required to deliver it "immediately" upon closure of the polling place. Urban polling places located away from the District Office were required to deliver the packets within 12 hours, and rural polling places within 24 hours of closure.

PREP

The IFE was mandated by law to provide quick information and a fast estimate of the electoral results, but it had no authority to determine the outcome of the election.
The IFE had two systems in place. One is the Programa de Resultados Electorales Preliminares ; the second was the Conteo Rápido . Although the two were often confused, they were in fact two separate systems.
The PREP is a mechanism to make public the results of the election. As soon as the polls closed and the votes were tallied in each polling place, a copy of the acta de escrutinio was sealed in a special envelope with a transparent window. It was then sent to a Center for Information Collation and Transmission that was located in the corresponding electoral district. There, the information was captured and transmitted to the central offices of the IFE for public dissemination. The information was then backed up and made public on the IFE website and a number of mirrors. The PREP is not an estimate or a statistical exercise, but rather the full unofficial returns from each polling place, as they were computed at the polls on the day of the election. They are not official returns, since the tallies must be certified by the District's Electoral Council before they become official. The PREP would continue to collate data until all polling places reported or until the early morning hours of Monday July 3. If the tally sheets contained certain inconsistencies, then they were reported separately, under the heading of Irregular, as per prior agreement between the parties, and were not added to the totals as they were received. Where possible, their results were added when the final report was made on Monday morning.
In the PREP, election results were posted as they were being received by the IFE. Calderón quickly overtook a very early lead by López Obrador and kept a diminishing lead until the end.

Quick count

The Conteo Rápido, on the other hand, was a statistical attempt at predicting the trend of the election.
On the night of the election, a random selection out of a representative sampling of 7,636 polling places was to be selected, and statistical analysis would then be performed in an attempt to predict the winner.
The agreement regarding the Conteo Rápido was as follows: Using the results that had been obtained by 10:00 pm on July 2, the technical committee would perform a statistical analysis. Once it was finished, the committee would give Luis Carlos Ugalde, the president of the IFE, a report with one of four recommendations/conclusions: that there was sufficient clarity and assurances as to which political party would get the most votes; that more time was needed to obtain further data that would clarify the conclusions; that the race was too close to call; or that there was an insufficient scientific basis to apply the tests on the samples obtained. Ugalde would then make the conclusions of the report public; again, as per prior agreement, in case the winner would be announced; in case, an announcement would be made to that effect and the announcement of the conclusion postponed for no more than two hours; and in case the race would be labeled "too close to call" and the precise results would not be revealed. According to the official report it had been decided in advance that a winner would be announced and the election called only if the separation between the intervals corresponding to the first two places was greater than 0.6%.
At 10:15 pm on July 2, a random sampling of 7,263 polling places was selected. Three statistical tests were performed on the sampling. A "robust" test, meant to provide conservative estimates as controls for the other two methods; a classic test; and a Bayesian test. The robust test predicted Calderón would get between 35.25% and 37.4% of the vote, versus 34.24% to 36.38% for López Obrador. The classic test predicted 35.68% to 36.53% for Calderón, versus 34.97% to 35.7% for López Obrador. The Bayesian test predicted Calderón with 35.77% to 36.40% versus López Obrador with 35.07% to 35.63%. Only the Bayesian test gave definitive results, but with a distance between intervals of less than 0.15%. Around 11 pm Luis Carlos Ugalde announced on live television that the technical committee had concluded the difference too small and that the race was too close to call. As per the original prior agreement between IFE and all political parties, Ugalde did not divulge the results of the Quick Count. Later that night, the representatives of the PAN and the Coalition For the Good of All requested that the results from the Quick Count be made public, while the representative of the Alliance for Mexico requested that the prior agreement of confidentiality be respected. Contradictory rumors of whom the count had favored were beginning to surface, and the official report was released a few days later to the press and through the IFE's website to quash them.
The final official count gave Calderón 35.89% of the vote, while López Obrador received 35.31%, confirming the prediction of the Quick Count.

Official count

In accordance with the COFIPE, the Official Count began on Wednesday July 5, 2006. The sealed ballot packages together with their signed actas de escrutinio y cómputo were collected by the District's Electoral Council. Each tally sheet was examined for possible errors or alterations; Council members from any party are allowed to raise certain objections, and under a specified set of circumstances packages could be opened and examined; for example, the package could be opened and recounted in the case of an illegible tally sheet. The District Electoral Council would then certify the totals from each package and polling place, and remit the results to the IFE.
The process took approximately 30 hours of continuous uninterrupted work. The IFE made public the results as they were being received and tabulated from all 300 of the country's Electoral Councils, one per congressional district. The results were tabulated live; an early lead by López Obrador dwindled slowly until Calderón overtook him in the early hours of Thursday morning. Unlike the PREP, the Official Count tallied votes in a more organized manner. Information received from the District Councils was added to the running count in the order of the district number, as they arrived to the office. The Councils themselves examined the electoral packages in numerical order. Returns from several northern states which had heavily favored the PAN according to the PREP were in particular slow to arrive; the delay was at least partially attributable to double checking of the tally sheets. Under Mexican Law, any political party could ask for the double-check, and raise objections. The campaign coordinator for Calderón blamed the delay on dilatory tactics by the PRD representatives in those districts.
The final vote count reported on July 6 showed that Calderón obtained 35.89%, while López Obrador received 35.31%. The difference between the two was 243,934 votes.

Recount and appeals

On Saturday, August 5, the TEPJF met in public session to rule on López Obrador's filings alleging irregularities and requesting recounts. The seven magistrates voted unanimously that there was sufficient legal justification to order a recount in only 11,839 ballot boxes in 155 districts, thus rejecting López Obrador's public demand that all votes and ballot boxes be recounted.
The Tribunal based its decision of a partial recount on the fact that, despite publicly demanding a "vote-by-vote" general recount, López Obrador's party only presented legal claims for fewer than 44,000 polling stations, or less than 34%. Therefore, legally, only those 44,000 polling stations were deemed controversial by the TEPJF. The Tribunal ruled that the non-controversial votes should not be recounted, because "the certainty asked by the Coalition is tied to the respect of the citizen counts in non-controversial polling stations". However, the Tribunal did certify that principles of certainty in the elections called for a justified recount in some of the controversial stations, as irregularities were found there.
The partial recount of votes in the disputed election began on August 9 amid escalating protests against alleged electoral irregularities in the close election.
On August 28, the TEPJF announced the results of the partial recount, subtracting 81,080 votes for Calderón, 76,897 votes for López Obrador, 63,114 for Roberto Madrazo, 5,962 for Patricia Mercado, 2,743 for Roberto Campa, and 7,940 for the remaining candidates. A total of 237,736 votes were annulled out of the approximately 4 million votes recounted. Than means around 6% of the recounted votes were annulled.

Election observers

There were national and international election observers.
There was a mission of election observers from the European Union. From their June 15, 2006 press release:
El Universal reported that the mission was present in all the states of Mexico and the Federal District since June 22. After the election, José Ignacio Salafranca, chief of the EU delegation, said he considered the election clean and the decision of the Electoral Tribunal as fair.
Mark Almond, an election observer in several countries, questioned the objectivity of the mass media in general and Salafranca in particular, citing the latter's political affiliation with Spain's Partido Popular, considered to be ideologically close to the PAN. Almond writes:
The Global Exchange group of 24 international observers claims it found electoral fraud or irregularities at all of the 60 polling places it observed, and called for a full recount.
The Mexican election watchdog group Alianza Cívica also claimed that many irregularities occurred.
Sophie McNeill, an Australian TV reporter for SBS Dateline Australia, reported:
Sophie McNeill also reported in that article that there were 673 international election observers, and that they "only oversaw a fraction of the country's 130 thousand polling stations."

Alleged irregularities

The day after the July 2, 2006 elections, the left-leaning newspaper La Jornada published reports of supposed electoral irregularities and López Obrador's claims that the ruling party had won the election through a supposed "fraud" escalated.

Illegal campaigning

During the run-up to the election, José María Aznar, the former Prime Minister and honorary chairman of the Popular Party in Spain, was alleged to have illegally interfered with the election, as described in Article 33 of the constitution of Mexico. John Ross, in an April 6, 2006 article, writes:
While Hugo Chávez was President of Venezuela at the time he made his comments, José María Aznar was not head of government of any nation when he made his comments. Additionally, Hugo Chávez made his comments in public and through broadcasting media, while Aznar's comments were made in a private meeting with PAN members and were not intended for broadcasting.

Zavala accusations

On June 6, during the second presidential debate, López Obrador alleged that Calderón's brother-in-law, Diego Heriberto Zavala Gómez, had failed to pay taxes on earnings of two and a half billion pesos that Zavala had obtained in favorable government contracts obtained while Calderón was Energy Secretary. López Obrador called Zavala the cuñado incómodo, a play on the nickname inconvenient brother of Raúl Salinas de Gortari.
The Calderón campaign and Zavala denied the accusations. El Universal reported on June 8 that according to Compranet, the government's procurement database, Zavala had three contracts related to the energy sector, all of them after Calderón had left office; and that his companies had a total of 18 contracts since 2003, totaling 129 million pesos.
On Friday June 9, 2006, López Obrador promised to deliver documentation that would prove the allegations. Later that day, Claudia Sheinbaum from the López Obrador campaign arrived at Calderón campaign headquarters with three filing boxes. They were met outside by the vice-coordinator of Calderón's campaign, and by the press and a notary public who had been called by the Calderón campaign. The vice-coordinator of the campaign insisted that the boxes be opened and the contents catalogued and witnessed by the notary public before they were brought inside. A verbal confrontation ensued; during this confrontation, the boxes were opened and it was found that they were almost empty and contained only a handful of copies of documents the PRD had made public previously, and three CDs. Sheinbaum attempted to explain what the contents were, but a shouting and shoving match began among PAN and PRD members present. The accusations against Zavala were discontinued after the confrontation, though they were never withdrawn.

Federal Electoral Institute corruption allegations

In April 2006, amid allegations that one of Felipe Calderón's brothers-in-law, Diego Hildebrando Zavala Gómez del Campo, was involved in the elections, the IFE declared that Hildebrando company did not participate in the implementation of the PREP. However, it was discovered later that year that Sagem Défense Sécurité, a French company of which Hildebrando is a partner, did provide biometric verification of the database of voters for the IFE.
An October 12, 2006 article by journalist John Ross claimed that Elba Esther Gordillo, the head of the National Education Workers Union, was aligned with the PAN. The article also stated, "Prior to July 2nd, Gordillo had corrupted the governing circles of the IFE—its president, the now-reviled Luis Carlos Ugalde, and three members of the governing counsel were Gordillo appointees and the teachers union czarina used her muscle to have 22,000 IFE polling place and district officials replaced by her people during the final weeks of the campaign."

Vote buying

The Global Exchange final report states: "...since vote buyers can know the results at each polling place, they can threaten to cut off benefits unless a community produces a determined number or percentage of votes for their party." The report also states: "According to research conducted by Alianza Cívica in the 2006 electoral cycle, a significant percentage of social program beneficiaries have been approached by persons who would condition receipt of program benefits upon electoral outcomes."
Julia Murphy, a University of Calgary anthropology professor, was a Global Exchange observer. A University of Calgary newspaper reported:
In his book El mito del fraude electoral en México, Fernando Pliego Carrasco made a study of the beneficiaries of several government social programs, sorted by the marginalization index of the electoral district in which they lived. The programs he considered were the "Desarrollo Local Microrregiones" program, and the "Desarrollo Humano Oportunidades" programs of the federal government, and to a lesser extent the program of "Empleo Temporal", "Vivienda Rural", and "Atención a los Adultos Mayores en Zonas Rurales". The first two were specifically mentioned by the Alianza por el Bien de Todos in their judicial filings. Pliego Carrasco is a Ph.D. in Social Sciences by the Colegio de Mexico, specializing in Sociology, and is a researcher at the Institute of Social Research of the Universidad Nacional Autónoma de México.
Pliego Carrasco criticizes the Alianza's filings by stating that they were based only on individual situations rather than a widespread study, that the information was based almost invariably on newspaper clippings, and that they only studied 26 districts. For his own study, Pliego Carrasco contrasted the growth in votes from the 2000 to the 2006 election for all parties. In all cases, the Alianza increased substantially more than the PAN, which even lost votes in the two categories of lowest marginalization. His conclusion is that there was no evidence that the PAN benefited from the expenditures. He found no statistical correlation between expenditures and the winning candidate. He derived similar conclusions for the other four programs. His conclusion is that the vote distribution does not support the hypothesis that social programs were used wholesale to influence the vote.

Ballot packets opening

's July 11, 2006 article has a La Jornada photo of a July 10 protest outside the office of the Federal Electoral Institute in the Tabasco town of Comalcalco. Giordano writes: "Trapped inside are at least ten IFE officials who, according to eyewitnesses, illegally entered the building, brought sealed ballot boxes out into the patio, and began to open them, breaking the official seals. They were seen revising anew the 'actas' with the vote tallies and recounting the ballots, without, as the law requires, the presence of representatives from all the political parties."
On July 16 La Jornada claimed that 40% of the vote packets nationwide had been reopened illegally by the IFE. IFE officials later explained that the opening of the packets was not illegal, and that it only served to satisfy a petition by López Obrador's Coalition.

Ballots found in trash

There were various reports of ballots found in the trash.
In Ciudad Nezahualcóyotl and in Xalapa, used ballots and other electoral materials were found in rubbish dumps.
Al Giordano of Narco News wrote on July 5, 2006 of ballots discovered in the Nezahuacoyotl garbage dump. The article also has an El Universal photo of the garbage dump. His July 8 article has a La Jornada photo of three completed ballots found in a Mexico City trash can.
Reforma later reported that the Federal Prosecutor in Electoral Crimes stated that no one had judicially made an accusation regarding the illegal use or appropriation of electoral material, and that the alleged ballots in the Nezahualcoyotl garbage dump were in fact photocopies of electoral reports, and not ballots. She also added that the PRD had not filed a suit or a judicial accusation to any of their public remarks.

Videos

In a July 10, 2006 press conference, López Obrador showed a video in which a man was seen taking several ballots and depositing them in a ballot box. López Obrador alleged the video, and a second one he also distributed, was proof of "pregnant ballot boxes", a euphemism for ballot stuffing.
Within a few hours, the IFE and the polling station officials explained that between 8 and 10 ballots for the election to the Lower House had been incorrectly placed in the presidential ballot box. Article 231 of the COFIPE states that ballots placed in the incorrect box should be moved to the correct box. The president of the polling station ordered the ballots moved to the correct box, and all representatives present agreed to this. The video had been recorded by the representative of the Coalition, who also signed the report that explained it in the official tally sheet. On July 11, both the president of the polling place and the observer representing López Obrador's Coalition confirmed that this was indeed what had occurred.
In response, López Obrador accused his own Coalition observer of corruption, saying "Not every one of our observers acted properly" and that "there was a lot of money involved, and we know some of our representatives were offered money." When asked if they had any proof of this, he said they did not.

Annulled votes

In a master complaint of around 900 pages López Obrador's party claimed that 900,000 annulled votes were from areas that voted more strongly for him.

PREP

There were various claims of the possibility of fraud concerning the PREP results, and also the official count that followed. Many people confused the two and were further thrown by several differences between them. For example, the PREP listed polling booth-by-polling booth, district-by-district, and state-by-state running totals in addition to a full running total by candidate. The report of the Official Count, on the other hand, only gave the full running totals by candidate, obtained by adding results as they were reported at the district council level. The order in which tallies were added was also different. Finally, the behavior of the running tallies for the candidates was also very different: in the unofficial PREP count Calderón quickly overtook a very early lead by López Obrador and kept a diminishing lead until the end, while in the official count López Obrador led from the first report on, until Calderon overtook him near the end of the count.

Statistical anomalies

Physics professor Jorge A. López did a statistical analysis of the PREP results. There was some discussion among statistics experts, and other non-expert readers, about that analysis at a Columbia University blog on "Statistical Modeling, Causal Inference, and Social Science."
In a July 19, 2006 comment to his July 17 article Professor James K. Galbraith writes: "For the PREP data, the issue of whether we are seeing 'cyberfraud' or a more traditional kind is a red herring, in my view. I did not use the term 'cyberfraud' in my essay. In principle, the anomalies observed could be generated electronically, or manually. Or, to be complete, they could have innocent explanations; however so far these have not been provided." Professor Galbraith holds the Lloyd M Bentsen Jr chair of government/business relations at the Lyndon B Johnson school of public affairs, the University of Texas.
Physics professor Luis Mochán also did several analyses of the PREP results; one involved the evolution over time of the PREP results. An August 2, 2006 AlterNet article writes: "Mochán notes that these statistical anomalies aren't definitive proof of anything."
In his book El mito del fraude electoral en México Fernando Pliego Carrasco considered the timing of the results and related them to the "marginalization degree" of the electoral districts, as determined by the Consejo Nacional de Población's 2005 census. The marginalization degree of a district is a measure of the affluence of the district, its infrastructure, population centers size, and educational level. Very low marginalization represents an affluent district with good infrastructure and large, educated population. Very high marginalization represents a poor district with little or very poor infrastructure, and small, largely uneducated population.
The table below shows the results of the PREP, sorted by marginalization degree and candidate.
Marginalization degreeCalderón% of totalLópez Obrador% of totalMadrazo% of total
Very low9,944,22624.18%8,793,37921.38%4,560,13711.09%
Low2,037,1854.95%2,090,6755.08%1,365,3483.32%
Medium1,420,3403.45%1,637,6753.98%1,274,6693.10%
High1,146,9282.79%1,532,2113.73%1,424,2963.46%
Very High222,3300.54%459,5371.12%503,4391.22%
Total14,771,00935.91%14,513,47735.29%9,127,88922.19%

The marginalization degree of the district was directly related to the time it took for the results to be carried to the corresponding CEDAT. Those of very low marginalization took an average of 4:31 hours; those of low marginalization an average of 6:14 hours; medium marginalization an average of 6:54 hours; high marginalization 8:45 hours; and very high marginalization an average of 10:32 hours. The time it took to get results from a polling place to the corresponding CEDAT was also directly related to the average distance between the polling place and the CEDAT. Almost two thirds of Calderón's votes came from districts with very low marginalization, arriving first, with almost three percentage points advantage over López Obrador. Moreover, it was only on those districts that Calderón obtained victories, while López Obrador was ahead mainly in districts with low, medium, and high marginalization; Madrazo only won in some districts with very high marginalization. Pliego Carrasco states that taking all these factors into consideration completely explains the behavior of the partial tallies, in particular the changes in the rates of growth that began occurring around 9 pm, and the further changes between 2 and 3 am. Pliego Carrasco criticizes Mochán and others for not considering the marginalization and the distance between districts and the corresponding CEDAT, and for considering only a very rough urban/rural sorting of the districts and using methods and techniques that are valid for the natural sciences and engineering but that he considers unjustified in their application to social sciences.
In a letter published on August 31, 2006 in La Jornada, José Woldenberg wrote in part:
José Woldenberg was a founding member of the Unified Socialist Party of Mexico, the Mexican Socialist Party and the PRD, which he left in April 1991. He was president of the IFE from 1996 to 2003 and oversaw the 2000 general election.

Allegations of added or missing votes (overvotes and undervotes)

On July 10, 2006, Frontera NorteSur, reporting on an initial analysis by Civic Alliance, wrote: "In the Lopez Obrador strongholds, 312,450 more votes for senators than president were tallied, while in the pro-Calderon zones, 403,740 more votes for president than senators were tabulated." However, they do not specify which candidate seems to benefit from this. For example, in the PAN stronghold of Guanajuato, both candidates received almost exactly the same number of votes above their corresponding senatorial candidate, with López Obrador edging Calderón by 51,498 to 51,283
The Civic Alliance claimed that in some Calderón strongholds the number of votes cast greatly exceeded the number of registered voters. Pliego Carrasco, however, found that the totals were similar for polling places carried by López Obrador.
López Obrador also claimed that there were 1,621,187 votes added or missing from 72,197 polling stations. See article and chart with breakdown of numbers: It also has links in the sidebar to various scientific studies on the presidential election results.
Luis Mochán analyzed the overvotes and undervotes for those polling places that had sufficient data intact to do so, basing both totals and controls on the PREP. He used four different methods and came up with between 1.5 and 2.35 million overvotes and undervotes; these are calculated by comparing the number of votes plus the number of blank ballots remaining as reported in the PREP with sundry other numbers also taken from the PREP: the total number of voters signed in, the total number of ballots originally allotted to the polling place, and so on. Note that this vote count analysis is separate from his analysis of statistical anomalies in the PREP voting timeline. His September 7, 2006 paper discusses both of these analyses.
Some, including Mochán in his article note that undervotes may result from citizens who decide not to cast their ballot; overvotes may result from poll workers failing to register a citizen in the log; and so on.
According to Mochán others have speculated in the press that the majority of overvotes and undervotes come from so-called "contiguous polling places" and are the result of confusion on the part of the voters, who deposited their ballots in the wrong box. Mochán neither confirms nor denies the provenance of the overvotes and undervotes, and only reports aggregated totals.
The reason for this possible confusion is the following: the law places a hard limit of 750 voters per polling place. For districts with a high population density, usually a single location will contain several polling places in order to split the voters among them. These polling places are usually assigned a single number; one is designated as the "basic" polling place, and the others are designated with a second number identifying them as "contiguous 1", "contiguous 2", etc. Each one has its own officials and ballot boxes. However, being in close proximity it is not uncommon for voters to be confused as to which box corresponds to which polling place.
This confusion results in a ballot for one polling place being deposited in the ballot box corresponding to an officially different polling place, giving both an undervote for the correct polling place and an overvote for the incorrect one. According to the COFIPE, article 231 a ballot which is placed in the wrong box should be moved to the correct box if possible, but should in any case be counted and reported. A single such vote would appear as both an undervote and an overvote in the analyses and reports mentioned above.
Mochán's article reports almost three times as many undervotes as overvotes in the aggregated totals.
Mochán notes that "it is quite probable that many of the inconsistencies has its origin in errors made without malice", listing several possible such errors. Mochán strongly stresses that in his opinion these and other explanations considered are "not enough to explain the magnitude of errors."
Mochán's analyses are about the PREP numbers. The PREP totals are ignored when the official count begins, and do not have any legal validity for the purposes of deciding the election. On the other hand, the existence of overvotes and of undervotes is one of the grounds for requesting that an electoral packet be opened and recounted. Mochán notes that the official totals does not include the information on leftover blank ballots, so that a verification of undervotes and overvotes in the case of adding up errors is impossible.
The Center for Economic and Policy Research reported that "61,192 of 126,139 ballot boxes contained 'adding up' errors. Election workers received a fixed and recorded number of blank ballots for each ballot box, and were instructed to keep track of them. Yet for nearly half of all ballot boxes, the total votes plus leftover blank ballots did not add up to the number of ballots received."
Pliego Carrasco compared the size and type of the errors in the different polling places, sorted according to who won the polling place. For example, for polling places in which the number of ballots issued to the polling place is different from the number of votes tallied plus the number of leftover ballots, he gives the table below. The FCH column gives the number of polling places where Calderón received the most votes that exhibited that type of error, while the AMLO column shows the number where López Obrador received the most votes. Undervotes refers to the case where the number of votes tallied plus the number of leftover ballots is less than the number of blank ballots allocated to the polling place, while Overvotes is the case where the number of allocated blank ballots is less than the votes tallied plus leftover ballots. The numbers are also taken from the PREP.
Type of errorRangeFCH%AMLO%
Undervotes101 or more2920.5%2520.5%
21 to 1008291.4%9231.8%
11 to 201,1762.1%1,3282.5%
6 to 102,1813.8%2,3504.5%
1 to 512,12321.2%11,10021.2%
No error29,45551.5%25,40448.4%
Overvotes1 to 57,24612.7%7,10513.5%
6 to 101,4182.5%1,4392.7%
11 to 207831.4%7321.4%
21 to 1007681.3%8521.6%
101 or more9441.6%9821.9%

Pliego Carrasco notes that both the absolute and relative numbers are very close for both candidates. He found similar results for errors where the number of votes tallied is different from the number of voters registered as having voted in the official rolls; where the tallied total is different from the number of ballots in the box; and where the number of ballots in the box differs from the number of voters registered as having voted in the official votes. Taking into account the number of errors per polling place, again the results are very similar for all candidates.
His conclusion is that there was no bias in the arithmetic errors ; he concludes that the errors were likely caused by the normal distribution of mathematical and organizational abilities of the randomly selected citizens in the polling place, and in the sometimes confusing design of the tally sheets and inadequate training of polling place workers.

Counting votes in the official count

In the official count López Obrador was ahead until late in the count. The official count does not use computers to count the votes. It tallies the totals from the actas as they are certified by the district boards. Every party has physical copies of every certified acta, and a representative in every district board.
The votes were not counted randomly, but in order of the districts, with Mexico City, a López Obrador stronghold, coming first, and the pro-Calderón North coming last. See the PREP, Quick Count, and Official count sections of the article farther up.
The IFE did not put up the official count results by precinct. The report of the Official Count online only gave the full running totals by candidate, obtained by adding results as they were reported at the district council level.
The Official Count does not necessarily reopen the ballot packets and do a full vote-by-vote recount. It mostly only tabulates the numbers from the 130,000 sets of official paperwork for the hand counts previously done for all of the 130,000 precincts. Because the packets usually remain closed, some claim that much of the fraud and irregularities allegedly done at the precinct level would not be discovered by the official count. Although the representatives from any political parties in the district board may request that a packet be opened and recounted, this can only be done under specific sets of circumstances listed in the COFIPE. For example, the package could be opened and recounted in the case of an illegible, missing, or inconsistent tally sheet.
"A copy of the 'acta' was supposed to be given to the IFE officer and each party representative, one copy had to be displayed outside of the polling place, and another copy had to be included with the ballots in the packages that everyone there had to seal and send to their respective electoral district." There are several locations on the web that have copies of the actas.
The only result that counts legally for a precinct is the "acta" signed by the election officials for that precinct. There were many claims of fraud and discrepancies between the actas and IFE's PREP results. There are photos comparing actas to PREP results.
On July 7, 2006 John Ross wrote concerning the official count: "As the day progressed, PAN and PRI electoral officials, charging Lopez Obrador's people with trying to obstruct the process, repeatedly rejected PRD demands to open the ballot boxes and recount the votes inside one by one in instances where Lopez Obrador's tally sheets did not coincide with numbers in the PREP or were different from the sheets attached to the ballot box. When a recount was allowed such as in one Veracruz district, Lopez Obrador sometimes recouped as many as a thousand votes."

"Old fashioned fraud"

Charges that there had been a "cybernetic fraud" began to surface on Monday, July 3, 2006. Questions were raised about "lost votes", differences between results reported in the PREP and tally sheets, and other alleged anomalies. The charges and questions multiplied after July 5 when the running totals of the Official Count behaved differently from the ones in the PREP.
During the next week and a half, PRD's president Leonel Cota and some of the main figures in López Obrador's campaign, including Ricardo Monreal, Claudia Sheinbaum, and Manuel Camacho Solís, gave numerous interviews in which they explained how the cybernetic fraud had allegedly occurred, through the use of "an algorithm" in the IFEs computers, and demanded an audit of the computer code used in the PREP. Several reports quoted earlier in this article also referred to these problems and allegations.
However, on July 17, 2006, in a radio interview with Miguel Ángel Granados Chapa on Radio UNAM, López Obrador said:
Despite this statement, the allegations of "cybernetic fraud" continued from sources other than the López Obrador campaign. The report on the interview in La Jornada the next day quoted only the new charge of old-fashioned fraud and did not mention the disavowal of a cybernetic fraud. On August 10, 2006, the PRD filed suit against Calderón, Ugalde, and all nine councilors of the IFE for cybernetic fraud, "given the results that were given in the vote counts on the day of the election, with a tendency in favor of the presidential candidate of the National Action Party."

Annulled votes

In a master complaint of around 900 pages López Obrador's party claimed that 900,000 annulled votes were from areas that voted more strongly for him.
The Coalition's complaint was itself not free of inconsistencies. For example, according to José Contreras:
he coalition says of polling place 5085 contiguous 2 in Miguel Hidalgo: "the number of ballots received by the president of the polling place, which was 404, does not agree with the 293 ballots deposited in the ballot box plus the 141 blank ballots left over, giving a variance of more than 30." That polling place does not exist. Another nonexistent polling place is number 4940 contiguous 2, of which the Coalition states: "the number of ballots received by the president of the polling place, which was 405, does not agree with the 309 ballots deposited in the ballot box plus the 175 blank ballots left over, giving a variance of more 11." The list of nonexistent polling places that were challenged has no end. Another example is number 2164, of which the coalition notes: "the number of ballots received by the president of the polling place, which was 455, does not agree with the 314 ballots deposited in the ballot box and the 167 unused ballots, giving a variance of more than 24." In addition, there is a good number of polling places that were challenged despite the fact that the data in the tally sheet agrees fully, and that the tally sheet is signed by the representatives of the coalition For the Good of All, and others that were already opened during the district council counts."

Opinion of Mexican intellectuals

On August 2, before the TEPJF recount hearing on August 5, a group of 136 intellectuals signed a letter in which they invited everyone to stay calm and respect the TEPJF decision. They stated that even though the campaigns were full of disqualifications by all sides, the actual voting was "exemplary in its orderly participation of over 42 million voters." They also restated their belief in the IFE's independence and integrity, and complimented it on the organization of the vote, modulo what they called some "specific/isolated errors". Then they went on to address the controversy regarding the validity and cleanliness of the election. They wrote:
"We who sign this document have followed the arguments and the evidence presented. We find no solid evidence to sustain the existence of a fabricated fraud against or in favor of any candidate. In an election in which the citizens count the votes there can be errors and irregularities, but no fraud."

They called on all sides to respect the final decision of the TEPJF, calling it "the path designed to attend to doubts, complaints, or electoral conflicts. It is impossible to completely eradicate differences on electoral matters." The signers included many political observers from different political persuasions, including José Woldenberg, one of the founders of the PRD and the former president of the IFE who oversaw the 2000 presidential election.
Another group of intellectuals who identified themselves as "members of the artistic and cultural community" issued a different letter on the same day. They said the numerous reports of irregularities "fan, to different degrees, the suspicion that the voting could have suffered decisive alterations." Citing the polarizing campaign, the number of alleged irregularities, and the closeness of the race, they called on the authorities to order a full recount to "erase any possible doubt" about the outcome of the election. They called on the electoral authorities to act in accordance with both the law and the "exceptional circumstances" the country was going through, while rejecting what they called "arguments based on legal technicalities". Among the signers of the second missive were Carlos Monsiváis, Juan Villoro, Sergio Pitol, Elena Poniatowska, and Luis Mandoki.

Polls

According to a poll of Mexico City residents by the newspaper Reforma, 65% of those polled believed that there was fraud and that there should be a full recount. A poll of Mexico City residents on August 9 by the newspaper El Universal found that 59% believed that fraud had occurred, and 63% believed there should be a full recount of all the votes. Mexico City was one of López Obrador's main bases of support, as he was mayor of the city before he ran for the presidency.
Nationwide polls: A poll released on July 27 by El Universal found that 48 percent wanted a full recount, and 28 percent were against it. 39 percent of Mexicans believe fraud occurred according to a nationwide poll of registered voters taken August 25 through 28, 2006 by the newspaper El Universal. 51 percent believed the election was clean.

Recount

Al Giordano wrote: "The partial recount did show that, out of 11,839 precincts recounted, 7,442 either had ballots missing or ballots above the number of people who had voted there. Had the trife annulled those precincts—a precedent set in its review of past state and municipal elections—López Obrador would have been declared president-elect. Instead, on August 28th the trife announced that it had annulled 237,736 votes, without specifying which or how many ballot boxes these came from. The result was to reduce Calderón's margin of victory by a mere 4,183 votes."
The TEPJF is sometimes referred to in the media by the acronym of its predecessor, the TRIFE.
The chief electoral officials and judges refused to release details about the partial recount of 4 million votes, and moved to destroy all 41 million ballots. Media articles cited the lack of reasons, principles, legal precedents, etc. in the decision of the Federal Electoral Tribunal to declare Felipe Calderón as President-elect.
Both López Obrador and Calderón called for the ballots to be saved, as did a petition from 800 organizations, businesses, and individuals. The Federal Electoral Institute said that the law required them to destroy the ballots and electoral documentation before December 1, 2006. But media articles noted that the IFE's governing council could have delayed the destruction.
TEPFJ released some recount information. By reading thousands of those pages the Center for Economic and Policy Research was able to find specific recount data for 1,706 ballot boxes of the 11,839 ballot boxes that were recounted by the TEPJF. Calderón lost 1362 votes. López Obrador gained 77 votes. CEPR: "The 1,362 votes lost by Calderón represent 0.54 percent of his votes in these ballot boxes."
Calderón only officially won the election by 0.58 percent of the nationwide vote.
CEPR: "The result for the whole group of recounted ballot boxes would likely show a similar percentage, since the above ballot box totals were chosen randomly from the documents posted on the Electoral Tribunal of the Federal Judiciary web site." In a study also by the CEPR, the asymmetry in the changes were further discussed: "The PAN has claimed that any such discrepancies represent merely counting or other clerical errors,
rather than fraud. If this were true, it certainly is a high level of error; given the rules and
instructions, it is not immediately obvious why local election officials would fail to keep track of the
ballots assigned to them at a majority of the ballot boxes. This is by itself a strong argument for the
electoral authorities to conduct a full recount, if the public is to have confidence in the vote tallies."
The Federal Electoral Tribunal as a whole, and/or its individual members, has been accused of bias, one-sided political stacking, conflicts of interest, PAN business partnership, revolving-door political nepotism, Supreme Court judgeship ambitions, secret post-electoral meetings with the Chief Justice of the Supreme Court, ruling against election annulment due to exaggerated fears of anarchy, etc.. Also, Fox administration interior secretary Carlos Abascal has been accused of pressuring the tribunal.

Similarities to other elections

Several articles claim similarities between the alleged irregularities in this election and those claimed in the 2000 United States presidential election.
In the August 20, 2006 Chiapas state election for governor claims of fraud and irregularities were made by the parties opposing Juan José Sabines Guerrero who represented the Coalition for the Good of All. He won in the preliminary counts. Media articles at that time noted that final results would not be known until the state Electoral Institute had resolved the large number of result review filings that were lodged, chiefly by the PRI-led Alliance for Chiapas. Media articles noted that this election mirrored the presidential election in that it was a close election where fraud and irregularities were claimed, but that the roles of the right- and left-wing parties were reversed.

Organized opposition

Protests

With the result in doubt, the fear of civil unrest rippled across the nation. López Obrador and his supporters quickly began to organize a campaign to challenge the results of the election, including mass protests, marches, and civil disobedience, culminating in a massive rally in Mexico City's historic Zócalo on July 30, 2006. Estimates of the crowd at the rally range from 500,000 to 3,000,000 supporters.
López Obrador has consistently used Mexico City's Zócalo as the focal point for his movement. A constant location for celebrations and spontaneous gatherings, López Obrador has called together at least three "Informative Assemblies" to be held in the Zócalo and Mexico City's historic centre. He has also called his party's state and municipal organizations for nationwide protests.
Protests and acts of civil disobedience against the alleged irregularities included:
More than 800 photographs of the rallies that have been attended by millions of participants in Mexico City can be seen at the following locations:
First Informative Assembly. July 8, 2006:

Second Informative Assembly. July 16, 2006:

Third Informative Assembly. July 30, 2006:
Some photos below of the Third Assembly. Click the photos to enlarge them. Click them again to further enlarge them.

Results of opposition

Due to financing issues, the PRD has decided to continue financing the opposition with PRD legislators' diets. The diets are financial subsidies, and public resources, that the government grants to legislators to aid them in communicating with their constituents.
Some polls indicate that López Obrador's popularity has declined as a result of the protests. Some political analysts have stated that López Obrador's actions constitute sedition against the constitution and Mexican republic.
On the other hand, some say that the actions of the peaceful civil resistance had not been adequately covered by the mass media outside of Mexico City. They say that the importance of the Mexico City rallies called by López Obrador can be substantiated by the photographic evidence. In Mexico City according to an El Universal poll the majority of residents supported a full recount, but a larger majority there were against the closing of streets by the protesters.
According to a Sept. 8-30, 2006 Ipsos/AP poll of citizens of nine nations, Mexicans had the lowest confidence that their votes are counted accurately. Angus Reid Global Scan, Canada, writes: "87 per cent of Canadian respondents are very or somewhat confident that votes in their elections are counted accurately. France was next on the list with 85 per cent, followed by Germany with 84 per cent, South Korea with 83 per cent, Britain with 79 per cent, and Spain with 75 per cent. The lowest level of trust was registered in Mexico with 60 per cent, Italy with 65 per cent and the United States with 66 per cent."

Calls for electoral reform

On Thursday November 23, 2006 the European Union election observer mission presented its final report. The report was generally positive and defended the official result of the July 2 presidential election saying they "reflect the legitimate will of the Mexican people." However, the report also expressed concerns and raised several criticisms, both general and specific, of the electoral process and the election. The report includes 48 specific recommendations for future reform to "strengthen the electoral process," such as introducing random automatic recounts in the case of a close election, shortening the campaign period, considering the introduction of a two round run-off system for presidential elections, and general clarifications in the electoral law.
A November 24, 2006 article by The Herald Mexico reports: "All of Mexico's major parties have called for reforms to electoral laws, including clearer limits on funding, greater transparency on campaign spending and a shorter campaign period."
The article also reports: "The EU mission, headed by Spanish Deputy José Ignacio Salafranca, said on Thursday that a runoff election would help the nation's electoral system, especially following results as close as this year's, when Felipe Calderón beat Andrés Manuel López Obrador by less than a percentage point. In a news conference, Salafranca said that while a second round would be expensive, it would give 'greater democratic legitimacy' to the result."

2006 Inauguration

According to the Constitution, the President Elect must take the Oath of Office on December 1, before a joint session of Congress; the Constitution also states that the presidential term begins on December 1. Members of Congress belonging to the parties in the Coalition had threatened to disrupt the ceremony, with actions similar to those that prevented Vicente Fox from delivering his State of the Union address on September 1. According to the Constitution, if the elected president does not take the Oath of Office on 1 December before a joint session, the presidency is declared vacant and an interim president must be named and new elections called.
Anticipating a confrontation, legislators began to stake claims to key physical positions on the stage of the Legislative Palace in the days leading to December 1. On Tuesday, November 28, 2006, the members of the lower Chamber engaged in a confrontation as members of Calderón's PAN and López Obrador's PRD shoved each other off the stage. Later in the week, the impasse appeared to take a less hostile tone. As Fox's term came to an end at midnight on Thursday night, he held an unprecedented and largely symbolic ceremony in which he handed the presidential sash and presidential residence over to Calderón.
On Friday, hours before the scheduled Oath of Office ceremony in the Legislative Palace, the legislature erupted in a brawl. The incident was broadcast on live television. In spite of such events the ceremony took place. Calderón entered the Congress chamber through a back door directly onto the podium, and in a quick ceremony took the Oath of Office. Then, rather than deliver his inaugural address to Congress, he left through the back immediately after the National Anthem was sung, and delivered the address in the National Auditorium, filled with supporters.