THE LIMITS OF TRANSPARENCY: THE CASE OF MEXICO'S ELECTORAL BALLOTSJohn M. ACKERMAN* SUMMARYI. Introduction. II. Reasonable doubts and uncertainties. III. Transparency and electoral ballots. IV. Implications.
I. INTRODUCTION
We do not know with certainty who actually won Mexico's 2006 presidential election. On September 5th, 2006 the Federal Electoral Tribunal declared Felipe Calderón president-elect and on December 1st he took office. But doubts remain about whether Calderón actually received more votes than his principal contender, Andrés Manuel López Obrador, on Election Day. If Calderón did not receive more votes, López Obrador is correct to claim that the sitting president is illegitimate since his election would be the result of a violation of the democratic principle of majority rule. If Calderón did win the popular vote, López Obrador is wrong to make such statements.
There is one easy way to resolve this issue. The Federal Electoral Institute (IFE) could clear up the doubts by opening up the electoral ballots to public scrutiny. Inspired by the Florida newspaper recount after the 2000 elections in the U.S., major media outlets such as Proceso magazine, W Radio and El Universal newspaper have requested access to the ballots to carry out an independent review of the electoral results.1 Unfortunately, the IFE and the Federal Electoral Tribunal have denied access using highly questionable legal argumentation. In an attempt to skirt Mexico's powerful new freedom of information law, the Institute declared that the ballots are not "documents" at all and that allowing access to them would "destabilize" the county.2 The Tribunal has recently ratified this decision by claiming that although the ballots are indeed documents they are simply "unavailable" to public scrutiny, thereby illegally inserting a new reserve category into Mexico´s freedom of information law.3
This case is a crucial test for Mexico's fledgling access to information legislation and may mark the beginning of the end of the country's significant advances in the area of transparency. Other government agencies undoubtedly will learn from the example of the IFE and the Electoral Tribunal to construct equally ad hoc arguments to close down access. In a country where the typical attitude towards the law has been "obey but do not comply" ("obedezca pero no cumpla"), Mexico's freedom of access law could quickly be hollowed out as the country moves down the slippery slope of permitting "access" without achieving real "transparency".4
The present article is divided into three parts. First, I review the 2006 electoral results and the actions of the electoral authorities in order to demonstrate that there is reasonable doubt with regard to the results of the past presidential elections. Although there is no smoking gun, there are enough strange things about the results to merit further investigation. Second, I critique the IFE's and the Tribunal's responses to the request for access and demonstrate that there is no question that according to Mexican law the ballots are public documents once the election has been completed. Third, I draw out the implications the IFE's rejection of access has for the future of transparency in the country.
II. REASONABLE DOUBTS AND UNCERTAINTIES
The 2006 election was a mess at the very least. Even scholars who defend the behavior of the electoral authorities admit that at least 46.7% of the tally sheets from the presidential election had some sort of counting error.5 López Obrador's Coalición por el Bien de Todos (CBT) claims that there were irregularities in 72,197 of the total of 130,477 ballot boxes, or 62 percent of the total.6 Independent reviews have brought this number down to about 61,192 of the boxes, almost 50% of the total.7 But this is still an enormous number.
Many of these mistakes are apparently minor, with one or two extra or missing ballots per box. But we should remember that the difference between Calderón and López Obrador was only 233,000 out of a total of more than 41 million votes. Only 0.58 percentage points separated first and second place, amounting to approximately 1.8 votes per ballot box. To alter the results it would have been enough for a single ballot to be changed incorrectly from the López Obrador column to the Calderón column.
Many of the "mistakes" were not so minor at all. Hundreds of ballot boxes showed differences of dozens or even hundreds between the number of ballots received, votes issued and blank ballots left over.8 The limited partial recounts conducted by the IFE and the Electoral Tribunal only strengthened suspicions that something had gone wrong. For instance, the first partial recount of less than 2.2% of the ballot boxes (2,864 in total) ordered by the IFE on July 5th led to a change of 4,594 votes in only 116 of the ballot boxes.9 Some specific examples include ballot box 740B in the State of Querétaro where the number of votes for Calderón went from 431 to 224, a loss of 207 votes at a single polling station. In Mexico City, the IFE recounted 12 ballot boxes in the 27th district yielding a result of 120 additional votes for López Obrador and 26 fewer for Calderón.
These are only a couple of examples that the author has been able to document on his own. To date, the IFE has not issued a full report on the results of this preliminary recount. In its final report of the events of the 2006 elections published this past January, 2007,10 the institute states that it opened 2,864 ballot boxes and presents a summary table which indicates how many boxes were opened in each state. But it does not provide any information at all on how the results changed as a result of this recount nor a ballot box by ballot box summary of what occurred during the recount.
The IFE waited weeks before releasing preliminary data on its recount. For two weeks no one even knew how many boxes had been opened or where those boxes were located. On July 20th the IFE finally posted on its web site the locations of the ballot boxes where the recounts had taken place. But it was not until a week later, on July 28th, that the IFE announced the total change in the vote which had resulted from the recount. López Obrador apparently lost 14,350 votes and Calderón 12,854, a net gain of approximately 1,500 votes for Calderón in this initial recount.11
But we should keep in mind that the boxes opened during this process were only those which the IFE formally authorized opening. This is by no means a statistical sample and it is possible that the IFE may have only opened those boxes which promised to not affect the overall result. For instance, José Antonio Crespo has recently documented12 using data from the IFE itself that in District 15 of the PAN dominant State of Jalisco the IFE did not authorize the opening of a single box even though 150 of the sheets had inconsistencies, some of which were quite large including cases of discrepancies of 322, 117,106, 93, 83 and 70 ballots. In District 1 of PAN stronghold Aguascalientes the IFE failed to open a hundred boxes with tally sheets containing inconsistencies, some of them quite large including cases of discrepancies of 167, 135, 96, 59, and 56 ballots. In District 3 of Oaxaca the IFE only recounted two of the seventy boxes with inconsistencies, failing to open boxes in which the discrepancies were in the order of 1,469 ballots in one case and 131, 107, 53 and 41 in other cases.
The total amount of change in votes in this initial recount (over 27,000) is a worrisome number in itself given that only 2,864 of the 117,287 boxes (less than 2.2%) were actually reviewed by the IFE. If the same number of irregularities were present in the other 97.8% of the ballot boxes the total amount of irregular votes would amount to 1,200,272 votes, more than five times the difference between the first and the second place candidates. But most importantly, the fact that the IFE to date has not released a ballot box by ballot box summary of the results of this recount only raises suspicion as to the micro-foundations of the vote changes that occurred.
On August 5th, the Federal Electoral Tribunal ordered a second partial recount in 11,839 ballot boxes, approximately 9% of the total.13 In this decision the Tribunal applied questionable legal reasoning to categorically rejected the possibility of a full recount. The decision states that:
…each voting center is unique, made up of different actors, located in a different place and surrounded by a different context, (for that reason) the events and occurrences which take place in one, do not have interconnection with the others, in particular if the irregularities are due to the citizens who form part of the tables for the reception of the vote (SUP-JIN-212/2006-Inc.1, my translation).
The justices here state that each voting booth is a totally isolated universe without any "interconnection" with the other booths or universes. According to this logic, it is never valid to interpret a specific piece of information as representative of a larger phenomenon. The Tribunal therefore decided to order a recount only in those specific booths where according to its own strict criteria the irregularities in the acts were sufficiently large so as to merit recounting. Importantly, to date the Tribunal has failed to make public the specific criteria it used to decide which boxes to open and which ones to leave untouched.
This "interconnection thesis" breaks with centuries of research on social phenomenon, bringing us back to a radical empiricism which not even the positivists of the Vienna Circle would have dared to defend. Any social scientist who knows the first thing about statistical sampling knows that a small sample is enough to gain trustworthy information about a much larger universe. This is particularly true in the case of social phenomenon like the organization of elections where the poll watchers for each one of the electoral booths had been trained by the very same government institution and subject to the same political, social and media pressures.
If the Tribunal had ordered a recount in a statistical sample of the voting booths this would have significantly strengthened the results of its partial recount. Otherwise, we are left with and entirely slanted sample which cannot say much about the rest of the voting boxes. Some analysts have argued that since the boxes which were opened in this partial recount were the ones which had been challenged by López Obrador, the rest of them must have fewer irregularities than the ones which were opened. But this does not necessarily evident to the naked eye. In the absence of a scientific sample, judgments about the content of the unopened boxes are grounded in pure speculation.
Regardless, the Tribunal's decision to recount 9% of the boxes was a step in the right direction and made a clear public statement about the deficiencies of the IFE's own partial recount. The Tribunal's argumentation led it to only order a recount in those districts where according to its analysis the IFE should have done so in the first place during the July 5th district counts. This was therefore equivalent to a public reprimand to the IFE for having only recounted a quarter of the boxes it should have investigated and especially for having failed to open boxes which had been questioned by Lopez Obrador's CBT.
The Tribunal's partial recount also fed suspicions that something might be wrong with the election results. After the recount was completed the CBT released its own analysis of the results.14 According to the López Obrador camp the recount revealed that 3,873, or 33%, of the recounted booths had extra votes and 3,659, or 31%, had missing votes. In total 7,532 voting booths, or 65%, presented irregularities which amounted to a total of 119,744 altered votes (including both extra and missing ballots). In addition, the CBT claimed that in "practically all of the districts where the recount took place there were tens of thousands of boxes without seals and hundreds of thousands of envelopes which contain the electoral votes with violated seals and without signatures and even missing electoral boxes".15
Calderon's Party of National Action (PAN) downplayed these supposed irregularities. It pointed out that the recount had been ordered precisely to correct for the problem of extra and missing votes and that the voting boxes and envelopes had been opened previously by an explicit judicial order.16 The PAN has also emphasized that the recount apparently led to a minimal change in the difference between Calderón and López Obrador. A comparison between the final numbers presented by the Tribunal on September 5th and the original IFE numbers from the July 5th district count reveals that Calderón lost at least 2,756 votes and López Obrador gained at least 2,101 votes as a result of the Tribunal's recount, a total change of almost 5,000 votes.17
Nevertheless, the numbers are surely significantly higher since this estimate does not include the ballot boxes which were recounted and then annulled entirely by the Tribunal because of serious irregularities. The Tribunal annulled 80,601 votes for Calderón and 75,355 votes for López Obrador.18 Since almost all of the boxes annulled were from the recounted districts, it appears that López Obrador gained an additional 5, 246 votes from the recount which are not taken into account in the above calculations. The Tribunal's partial recount therefore reduced the difference between Calderón and López Obrador by approximately 10,000 votes.
If we were to extrapolate these changes in 9% of the ballot boxes to the rest of the election Calderón's margin of victory would have reduced significantly, from 233,000 votes to only about 122,000 or approximately 0.30% of the total vote. This would not have been enough to change the final result but it certainly would have put the election results under even closer scrutiny.
But all of this is speculation. Given the widespread prevalence of inconsistencies, it is by no means out of the question that a total recount could have given the victory to López Obrador. The problem is that, just as with the IFE's partial recount, to this day the Electoral Tribunal has still not released a final ballot box by ballot box account of the results of its own recount. In its final decision issued on September 5th, 200619 the Tribunal offered the vote totals for each one of the country's 300 voting districts but did not provide any information at the voting booth level nor any information on how the vote had changed through the two partial recounts. One can wade through thousands of pages of legal documents which lay out the Tribunal's previous decisions, but even then one does not get the entire picture. For instance, while some of the Tribunal's August 28th decisions with regard to the annulment or no of hundreds of ballot boxes include specific information on the recount, others simply do not.
The electoral authorities have not provided enough information for outsiders to independently verify the final vote count. In the face of this ignorance, the independent observer is left with a stark choice: trust or skepticism. Perhaps if elections had been clean in Mexico throughout the entire 20th century and there were absolutely no questions about the independence of the electoral officials we could blindly trust the results. But the fact is that Mexico is known internationally for its expertise in electoral fraud and all of the sitting IFE councilors were nominated and appointed by political forces (PAN, plus a fraction of the PRI) who are now aligned with Felipe Calderón. A healthy dose of skepticism is therefore the best service we can pay to democracy.
Additional considerations discourage blind trust in the electoral results. One of the CBT's most serious accusations was that the seals on the storage rooms, ballot boxes and envelopes containing the votes had been systematically violated. This could indicate possible vote tampering by the IFE itself or external actors during the days and weeks after the elections. This is precisely what happened after the openly fraudulent 1988 elections. According to Mexican law, only the Electoral Tribunal can require the seals to be broken to extract specific documents from the electoral boxes or, in special cases, to conduct a recount of the ballot. Not even the IFE is authorized to open up the boxes on its own. The Tribunal did authorize such actions on some occasions but by no means in such a broad manner as was apparently the case.20
The Tribunal's total silence on the issue of the conditions in which it found the ballots during the partial recount is quite worrisome. They do not mention a word about the topic either in their August 28th resolutions or their final September 5th decision. We are therefore left with the CBT's accusations and the PAN's abstract responses, once again without sufficient information to independently evaluate what actually happened to the electoral boxes.
There is an additional detail which clinches the case for suspicion. In many of the states where López Obrador came out ahead there were tens of thousands of votes fewer counted for President than for Senator, while in many of the states where Calderón won there were tens of thousands of votes more for president.21 In total, there were over 250,000 votes fewer votes for President than for Senator in six states (Tabasco, Michoacán, Veracruz, Oaxaca, Chiapas and Mexico City), all of these won by López Obrador. There were over 80,000 votes extra for President than for Senator in seven states won by Calderón (Nuevo León, Sinaloa, Baja California, Morelos, Sonora and Chihuahua). It is very difficult to find a rational explanation for this discrepancy. Why would citizens in some states reject all presidential candidates while those of other states reject all senatorial candidates? The suspicion of course is that somehow presidential votes were removed in the pro-López Obrador states and added in the pro-Calderón states.
In anticipation of possible difficulties with the electoral results a group of independent academics and civil society leaders created before the elections a special Citizen's Commission ("Comité Conciudadano de Seguimiento del Proceso Electoral" -CCSPE-) designed to keep a watchful eye on the electoral authorities. The committee included over twenty well known experts from a diversity of political backgrounds, some closer to Calderón, some closer to López Obrador and most entirely independent of all political parties.22 This committee issued a series of reports which also belie the idea that the 2006 election was perfect and that the results should be trusted blindly.
The Committee's initial report on the elections23 clearly states that it "shares with other citizens a reasonable doubt about who was the winner of the presidential elections". This reasonable doubt arises because of the inadequate information offered by the IFE about the electoral results, the pressure placed by the General Council on the Local IFE councils not to recount electoral boxes, the absence of party representatives in a high percentage of electoral booths,24 the deficiencies in the training of poll watchers, human errors in vote counting, as well as possible problems with the nullification of valid votes. In addition, the committee received reports about the deficient quality of the specialized ink used to mark voters' fingers and the relaxation of the criteria for replacing poll watchers. The report also cites an Alianza Cívica study on the election which states that approximately 30% of the voting booths observed presented cases of citizens who had been inexplicably removed from the voters list. The report concludes with the following statement: "We are concerned about the triumphalism of the IFE's General Council, which devalues the seriousness of the situation which is a product of its problematic and insufficient actions to control the electoral process and the exercise of its authority".
Was there fraud? We don't know. Was the election a mess? Definitely. Might Mexico's political actors have been intelligent and experienced enough to take advantage of this situation to manipulate the electoral results? Quite likely. Can we be absolutely certain that Felipe Calderón received more votes than López Obrador on election day? Definitely not.
III. TRANSPARENCY AND ELECTORAL BALLOTS
An independent examination of the electoral ballots used during the past presidential elections in Mexico is merited. There is a reasonable doubt about the electoral results and such an exploration could help clear up many of the questions. Indeed, it could easily end up legitimating Calderón's victory. The newspaper study conducted by the New York Times, the Washington Post and the Tribune Company after the 2000 elections in the United States concluded that Bush would have won anyway even if a full recount had taken place in Florida.25 This study also provided important information on election inequality since it demonstrated that if all Florida citizens had equal access to effective voting machines and clear ballots then Gore would have won. Faulty voting machines and complicated ballots in predominantly African American districts led to the incorrect nullification of countless numbers of pro-Gore votes.26 A comparable study of the Mexican elections of 2006 would undoubtedly provide a great deal of important information and help improve elections in the future.
But would such a study be legal? Both the IFE and the Electoral Tribunal have argued that the law does not allow them to give the public access to the ballots. In this section I show that there is no question about the legality of providing access. The IFE's rejection of the numerous FOIA requests which have been brought by citizens and journalists is based on an overtly politicized and highly questionable interpretation of the law. The Tribunal's ratification of the IFE's decision corrects some of the deficiencies of the original response, but simultaneously opens up a dangerous new escape hatch for authorities who wish to avoid freedom of information requests.
Mexico stands out internationally for the strength and effectiveness of its new freedom of information law (Ley Federal de Transparencia y Acceso a la Información Pública Gubernamental -LFTAIPG-), passed in 2002,27 It offers an excellent example of the new breed of FOI legislation which recently has grown up in new democracies throughout the developing world. Mexico along with South Africa, South Korea, Thailand and over twenty countries of Eastern Europe are setting the standard for access to information throughout the globe. These laws tend to be much stronger than both the historic laws of Sweden, Colombia and the U.S. and recently passed laws in older, wealthier democracies such as Germany, the U.K. and Switzerland. In a new democracy, access to information is not just a good idea which may help combat corruption and make government more effective, but an essential right of the people which will allow citizens to bring democracy to its ultimate consequences. After decades of authoritarian rule, social actors are tired of being treated as passive subjects and eager to take up their roles as active citizens. The design and implementation of a new FOI law is therefore inspired by idealistic hope and hard hitting societal demands.
Mexico's FOI law is firmly grounded in the principle of full disclosure. The law states in no uncertain terms that the interpretation of the legislation and the corresponding rules should be based on the "principle of maximum publicity and accessibility of the information possessed by the obligated agencies" (LFTAIPG, Article 6). What this means is that all government held information is public (LFTAIPG, Article 2), except for that which is explicitly reserved or classified by law. This is, after all, the point of freedom of information legislation. Such laws turn the tables from a principle of "need to know" to one of "right to know". Before the global transparency revolution, citizens needed to justify why they wanted to gain access to a particular document. Now it is the government that needs to justify why it cannot provide access.28
The Mexican law creates three different categories of information: 1) Public information to which access must be granted, 2) "Reserved" information which the government has the right to keep secret for up to 12 years (Articles 13, 14), and 3) "Confidential" information which can never be released (Article 18).
The only confidential information is "personal data" ("datos personales") such as addresses, sexual preferences, weight, height, etc. In addition, other documents, such as commercial secrets, can be confidential if explicitly provided for by the corresponding law and if individuals report the information as confidential when they hand it in to the authorities.
The electoral ballots therefore could not possibly be categorized as confidential documents. The act of voting is secret. No one should be able to tell for whom a particular citizen voted. But precisely because they are secret, electoral ballots do not contain any marks which could allow a third party to identify the person who filled out a particular ballot. Electoral ballots, qua documents, do not contain any personal data at all.
Nor would it be possible for citizens to declare that they wanted their ballots to be secret upon voting. Voting is a quintessentially public act. The very point of voting is because you want your preference to be known. With confidential ballots, officials and citizens would have no way of adding up voters preferences.
The law stipulates a broader set of criteria for determining whether information can be "reserved" from public view for a period of time. Article 13 indicates that the government can reserve information for up to twelve years if its release might significantly harm national security, public safety, international relations, economic stability, a person's life or causes serious harm to the government's ability to assure the rule of law, the investigation of crimes, or tax collection, etc. Article 14 then states that commercial secrets, judicial files and administrative procedures which have not been completed, and deliberative opinions also can be reserved for the same time period. Finally, agencies can reserve information when another law "explicitly" considers the information confidential, reserved, "commercial reserved" or "government confidential".
Significantly, information cannot be reserved simply because some part of it might affect some element of the reserve areas. The principle of maximum publicity requires the government to fully justify the lack of disclosure based on the grounds that publicizing the information would significantly harm the reserve areas and that this harm would be larger than the damage which would result from keeping the information secret. The law therefore requires the application of a "harm test" in the application of the reserve areas.
In its recent response to various requests for access to the ballots,29 the IFE has argued that disclosure of the ballots would significantly damage "the State´s role of organizing the federal elections and the other constitutional attributions which are assigned to the Federal Electoral Institute… because by giving the complainant access to the requested documentation it would harm the central principles of the electoral function…especially those of certainty and legality" (p. 62). The IFE's central argument is that the Constitution establishes the "definiteness" of each stage of the electoral process. For instance, Article 41, subsection IV of the Constitution states: "In order to guarantee the principles of constitutionality and legality of electoral acts and resolutions, there will be established a complaints system… which will give definiteness to the different stages of the electoral process and will guarantee the protection of the political rights of citizens to vote, be voted and association…".
The idea here is that the actions of the IFE can be appealed to the Judiciary, specifically the Electoral Tribunal, and that the Tribunal's decisions are final. Once the Tribunal has decided on a particular case or stage of the electoral process (campaigns, organization of the elections, reviewing voting booths, conducting the final count, pronouncing the winner, etc.) political actors cannot return to a previous stage to question its validity. This makes common sense. If, for instance, parties could continually present complaints against the results of each voting booth the election would literally never end. The losing party would go through each of the 130,000 booths one by one and then start again from the beginning, preventing a final decision from being reached.
The IFE argues that in the interest of assuring the "definiteness" of the electoral process Article 254 of the electoral law (Código Federal de Instituciones y Procedimientos Electorales -COFIPE-) states that the IFE must destroy all of the electoral documents once the final decision of the election is pronounced by the Electoral Tribunal. The idea is that permitting direct citizen access to the electoral ballots would confuse the people and would bring the electoral process back to the stage of counting up the ballots which has already been completed. This in turn would threaten the central task of the IFE of assuring the "certainty" of the electoral results.
The problem with this argument is that Mexico's freedom of information law does not stipulate anywhere that harm to "the State's role of organizing elections" or harm to "the central principles of the electoral function" are valid reasons for reserving information. The list of reasons for reserving information included in Articles 13 and 14 is a restricted, exclusive list, not simply an open list of some possible reasons. In order for the IFE to justify reserving the information it would have to offer a specific argument that access to the ballots would cause significant and direct harm to one of the areas listed. It cannot simply make up new ad hoc grounds for reserving information based on its particular interpretation of the law.
In addition, Article 254 of the electoral code does not require that the IFE destroy the electoral ballots immediately after the elections are decided by the Tribunal. Indeed, after the 2000 presidential elections the IFE waited over six months, until January of 2001, to destroy them. To this day, the IFE still holds the ballots used in the past 2006 presidential elections.
Apparently aware of the fact that their general defense of the principle of "definiteness" is insufficient to formally reserve the electoral ballots according to the access to information law, the IFE also claims that national security and "public order" is at stake.
This fundamental right (of access to information), as in any democratic state, is subject to restrictions, which not only have the purpose of protecting the intimacy of people, but also of guaranteeing public order, social interest, as well as primordial operations of the Mexican state, including, for instance, national security… In this way, and by way of example, one can mention the fact that the Fundamental Law of Bonn and the German Constitutional Court recognize the concept of "militant democracy", which restricts the freedom of expression or information when their exercise threatens certain public values which are of fundamental importance for the collectivity, including all of the procedures for deciding upon the makeup of the institutions of government (pp. 52 and 53).
The Institute here suggests that the request for access to the ballots used in the past presidential elections is an "attack against fundamental public values" equivalent to openly supporting a fascist takeover of government. The Fundamental Law of Bonn and the decisions of the German Constitutional Court cited by the IFE refer to norms designed to prevent the resurgence of Nazism or other extremist political movements in the country. German institutions have decided that it is acceptable for the government to silence such violent and anti-democratic expressions even though this implies the violation of basic civil rights like freedom of speech and expression. In parallel fashion, the IFE argues that the journalists who have requested access to the ballots threaten the very foundations of democratic politics and should have their basic civil rights stripped from them.
Such a leap of logic actually amounts to an implicit confession that something might be seriously wrong with the way in which the votes were counted during the past presidential elections. If everything were in order, an independent review of the ballots would only strengthen Calderón's legitimacy and ratify the IFE's defense of the cleanliness of the elections. The supposed threat to "public order" could only raise its head if the study revealed that the elections were indeed a mess and López Obrador might in fact have been the victor.
But even then it would be wrong to blame the disturbances which would most likely result from this discovery on the act of permitting access to the ballots. The real culprits would be the IFE and the Electoral Tribunal for not having cleared up the electoral results in time. They would be the actors responsible for eroding the confidence in Mexico's democracy, not the journalists who have requested access. The very point of a FOI law is to facilitate independent review of government actions.30 Shooting the messenger never resolves the problem.
Beyond these statements concerning "definiteness" and "public order", the IFE also tries to make the argument that, formally speaking, the electoral ballots are not "government documents" at all and therefore not subject to the access to information law. The IFE cites article 3, subsection III of the FOI law which defines "documents" as:
The files, reports, studies, acts, resolutions, oficios, correspondence, agreements, orders, directions, circulars, contracts, convenios, guides, notes, memos, statistics or any other registry which documents the exercise of the faculties or the activity of the obligated subjects and its public servants, independently of its source or date of elaboration. The documents can be in any form, written, printed, sound, visual, electronic, computerized or holographic.
According to the IFE, the electoral ballots do not document the activity of the electoral authority or its public servants since they are not filled out by IFE employees but by independent citizens. The ballots therefore "are not documents legally speaking, they are only the material expression of the electoral preference of the voting citizenry. The ballots register the popular expression of the vote and this expression cannot be categorized as an activity or faculty of the IFE and its public servants" (p. 11).
Here the IFE conveniently ignores the fact that the ballots are designed, printed, guarded and distributed by the IFE itself and are the basic input for the institution to carry out its central function of deciding who has won the elections. As discussed above, the IFE itself argues in other parts of its response that the ballots are central the "State's role in organizing elections". The fact that citizens and not public servants are the ones who actually fill out the ballots does imply that the documents themselves are private material. Indeed, the transparency law's definition of "information" as "that contained in the documents which the obligated subjects generate, obtain, acquire, transform or conserve in any manner" (article 3, subsection V) is explicitly designed to avoid such a limitation of the reach of the law. The law is an access to "information" law not an access to "documents" law.
In any case, if the electoral ballots are not "documents", what then might they be? The IFE argues that they are simply the "material expression of electoral preferences". But the "materialization of an expression" is, precisely, a "document". Merriam-Webster's Dictionary offers the following definition of the word "document":
1 a archaic: PROOF, EVIDENCE b: an original or official paper relied on as the basis, proof, or support of something c: something (as a photograph or a recording) that serves as evidence or proof.
2 a: a writing conveying information b: a material substance (as a coin or stone) having on it a representation of thoughts by means of some conventional mark or symbol c: DOCUMENTARY.
3: a computer file containing information input by a computer user and usually created with an application (as a word processor).
The Dictionary of the Spanish Royal Academy of Language offers almost exactly the same definition of the word "documento" in Spanish. In its efforts to bend the law, the IFE has gone to the extreme of implying that the electoral ballots are some sort a metaphysical material which we simply cannot define.
This reinterpretation of the nature of the ballots is in direct contradiction to the electoral code as well. Article 254 which speaks of the destruction of the ballots explicitly states that the material to be destroyed is "documentation". In an important decision the Electoral Tribunal has also recently categorized electoral ballots as "documents", distinguishing them from other electoral "material" such as the voting booths, boxes, crayons, etc.31
Finally, the institute also claims that Article 254 implies that the Institute is not permitted to do anything else with the ballots except destroy them once the elections are over. According to the "principle of legality", government agencies are only allowed to do what they are explicitly authorized to do by law. They may not violate a law simply because they think it is inappropriate or unconstitutional. Only if the judiciary explicitly intervenes are they permitted to do so.
Here the IFE cites an important case resolved by the Electoral Tribunal in 1998.32 In this case the Tribunal decided that the IFE is not permitted by law to conduct independent investigations of the content of the electoral packages.
The Federal Electoral Institute is prohibited from establishing different procedures from those contemplated by the Federal Code of Electoral Institutions and Procedures. Ordering the practice of studies or investigations, to be carried out by external parties or by the organs of the Institute, as well as by its own personnel, involving actions different from those set out by law, of the documentation contained in the electoral packages and the information inscribed in the electoral documents of which they are made up, constitutes a practice which is incorrect (viciado) in its origins, because of the lack of faculties in this direction.
But this case was decided in 1998, four years before the passage of Mexico's access to information law. At that time it was indeed the case that there was no law which explicitly ordered the IFE to provide access to the electoral boxes after the elections were over. But now there is such a law. The access to information law passed in 2002 explicitly states that all documents in the hands of government agencies are public unless they fall under one of the categories of reserved or confidential information.
This past April 25th, 2007 the Electoral Tribunal ratified the electoral institute's decision to deny access to the ballots (SUP-JDC-10/2007 y SUP-JDC-88/2007 accrued). To its credit, the Tribunal flatly rejected the institute's arguments with regard to the nature of the ballots and the possible threat to national security. According to the justices, the ballots are without question "documents" and they do not fall into any of the reserve areas, including a possible risk to national security, listed in article 13 of the law.
The Tribunal grounds its denial of access on two different clauses in the transparency law. First, article 14, subsection I states that information can be withheld if there is an "explicit indication" in another law which states that the requested documents are "confidential, reserved, commercial reserved or government confidential". The justices argue that since access to the ballots is strictly protected by law and then destroyed once the Tribunal has decided on the election, it is as if the ballots were declared "reserved" or "confidential" by the electoral law.
The problem with this argument is that the law states that there must be an explicit indication ("disposición expresa") of the confidentiality or reservation of the information. There is no room to construct analogies. Indeed, as argued above, the principle of maximum publicity demands that all information be declared public unless it explicitly falls into one of the reserve areas. In their decision the justices do their best to construct a case for the reservation of the ballots, but they are simply unable to find an explicit mention that prohibits access once the elections have been decided.
Perhaps aware of the weakness of this first argument, the Tribunal offers a second one to strengthen its case. The justices point out that article 42 of the transparency law only allows certain forms of access to information:
The obligation of access to information will be considered completed when the requestor is allowed to consult the documents in the place where they are located; or by issuing simple or certified copies or through any other medium. Access will only be given in the form in which the document permits it, but it will be given totally or partially, upon request.
The crucial phrase here is that "access will only be given in the form in which the document permits it" ("El acceso se dará solamente en la forma en que lo permita el documento de que se trate"). Indeed, the Tribunal conveniently does not cite the previous sentence which lists the different forms of access. According to the justices there is no way to access the ballots, since the electoral packages which contain them are sealed and the law only permits the electoral institute to open them on judicial order during the phase of judgment and evaluation of the election.
Theoretically, this would not prohibit access after the election has been decided when the ballots are no longer of any use to the judicial authorities. Today, the ballots are still sitting inside the electoral packages simply waiting to be destroyed. Nevertheless, the Tribunal argues that even though the ballots still exist and are no longer useful, citizens still cannot access them because the law does not enable the institute to open the packages.
The ballots are therefore simply "unavailable" ("indisponible") to the public. In this way the Tribunal avoids having to find an explicit clause which declares the ballots reserved or confidential and also escapes from the sophistry of pretending that they are not documents or might affect national security. The ballots are public documents but not available. Access would be permitted, but it is unfortunately just not possible.
But this argument fails to take into account the fact that article 42 explicitly refers to forms of access, not to forms of reservation. Title I, Chapter III (articles 13-19) is the section of the law which refers to "reserved and confidential information" and contains all of the stipulations which can apply in order to reject a freedom of information request. Article 42 belongs to Title II, Chapter III (articles 40-48) on "the Procedure of Access to an Agency or Entity". This article therefore assumes that access has been granted and regulates the way in which that access can be provided. It is an inappropriate leap of logic to imagine that this clause can give a government agency reason to deny access entirely.33
The Tribunal's arguments are much more sophisticated than the electoral institute's. But this should not distract us from the fact that they are equally aggressive to the freedom of information law. With this decision, the Electoral Tribunal takes on full legislative powers to inscribe an entirely new category of documents into the law, that of "unavailable documents". By reinterpreting Article 42 as a cause for withholding information, the Tribunal has opened up a massive new black hole into which government agencies throughout the county can throw uncomfortable documents they would like to protect from the public eye.
IV. IMPLICATIONS
The case of Mexico's electoral ballots is a landmark case in the international struggle to expand and exercise the right to know. It has already been highlighted by international NGOs such as the National Security Archive in Washington, freedominfo.org, and Global Exchange in San Francisco. Now that the Mexican electoral authorities have been allowed to get around the law using such ad hoc argumentation, other agencies will surely follow suit. Mexico's international leadership in the area of transparency could be at risk.
In its response to the ballot request the IFE makes a particularly revealing statement:
There is no necessity for sacramental phrases or absolute and overarching dispositions, which literally establish each and every one of the conditions for the exercise of a public subjective faculty (of access to information), since these conditions of exercise can be deduced validly from the systematic and functional reading of the norms applicable to a concrete case, given that the authorities do not proceed in an arbitrary manner, deciding what ever they think is convenient to do (p. 47).
Such a statement amounts to the unilateral erasure of the principle of publicity which gives force to FOI laws. Here the IFE is saying that we should not conceive of the access to information law as being grounded in fundamental, "sacramental" or "overarching" principles such as maximum publicity or the need to hold government officials accountable. Instead, we should proceed on a case by case basis to see whether access is merited in each particular situation.
A case by case analysis is of course always necessary, but such a procedure is not in conflict with the overall principles and exemptions of a FOI law. It is a false dichotomy to state that either we follow overall principles in the application of a law or we look at each individual case. The point of any law is precisely to set out broad principles and clear rules that should be followed when deciding specific cases. Mexico's access to information law clearly states that all information in the hands of government is public, except for that information which is explicitly reserved or classified by the law itself. As we have seen above, the ballots do not fall into either of these categories and therefore must be declared public documents. The law allows no room for special cases.
Now that the Electoral Tribunal has ratified the IFE's refusal to grant access to the ballots the door is left wide open for other agencies to also make up their own "special cases" in the interest of closing down access. For instance, one of the areas which has been particularly resistant to the new transparency law in Mexico is the area of government trusts (fideicomisos).34 These trusts hold government funds, such as customs duties, or a mixture of public and private funds, as in the case of the banking trust. Parallel to the case of the electoral ballots, the Secretary of Finance and Public Credit has used the fact that these trusts are not formally "government agencies" to avoid the transparency law, even though they clearly manage public monies. Mexico's Institute for Access to Public Information (IFAI) has waged an important battle to gain access to information on the trusts. The ratification of the IFE's decision by the Electoral Tribunal sets an extremely negative precedent in this regard.
Now that the IFE's refusal has been left standing this could lead to a massive reclassification of information in the hands of government. Officials could claim that any document which is eventually destroyed or requires special handling is simply "unavailable". In the oral arguments of the case justice Salvador Nava set out the road for this eventual spread of opacity. Nava argues that the electoral ballots are like the bills which the Central Bank holds in its safe rooms: citizens can know how many there are but they cannot go into the Bank to count the bills themselves.
Such an argument is a non-sequitor. First of all, central banks do not hold their reserves in bills, but in government bonds or other securities. Access to copies of these securities would have to be permitted. Otherwise, suspicions would arise with regard to the validity of the central bank reports since investors would have no way of independently verifying them. We would then be in a parallel, or perhaps even more dangerous, situation to the widespread suspicion about the results of the 2006 Mexican elections. Indeed, even if the central bank did hold its reserves in peso bills, access to certified copies would also have to be permitted.
The Tribunal's message is that citizens should be satisfied with government reports, that they should not be given the raw information and documents necessary to verify these reports. As a result, when a citizen or a journalist now requests the receipts of the cement purchases of the Secretary of Public Works the agency will be able to satisfy the request with a document which supposedly summarizes its purchases. There will be no need for it to hand over copies of the actual pieces of paper which document these purchases.
Democracy implies much more than free and fair elections. It also requires a fundamental sea change in the way that government relates to society. In order to fully institutionalize democratic politics, the governing class needs to reconceptualize its role and establish a truly open process of respectful dialogue with social actors.35 Unfortunately, the response of Mexico's electoral authorities to the requests to access the electoral ballots used in the past elections reveals that a profound faith in Machiavellian "reason of the state" and a rejection of citizen demands for transparency remain solidly entrenched among Mexico's top public officials. If this attitude does not change soon Mexico's democracy may be at risk.
I. INTRODUCTION
We do not know with certainty who actually won Mexico's 2006 presidential election. On September 5th, 2006 the Federal Electoral Tribunal declared Felipe Calderón president-elect and on December 1st he took office. But doubts remain about whether Calderón actually received more votes than his principal contender, Andrés Manuel López Obrador, on Election Day. If Calderón did not receive more votes, López Obrador is correct to claim that the sitting president is illegitimate since his election would be the result of a violation of the democratic principle of majority rule. If Calderón did win the popular vote, López Obrador is wrong to make such statements.
There is one easy way to resolve this issue. The Federal Electoral Institute (IFE) could clear up the doubts by opening up the electoral ballots to public scrutiny. Inspired by the Florida newspaper recount after the 2000 elections in the U.S., major media outlets such as Proceso magazine, W Radio and El Universal newspaper have requested access to the ballots to carry out an independent review of the electoral results.1 Unfortunately, the IFE and the Federal Electoral Tribunal have denied access using highly questionable legal argumentation. In an attempt to skirt Mexico's powerful new freedom of information law, the Institute declared that the ballots are not "documents" at all and that allowing access to them would "destabilize" the county.2 The Tribunal has recently ratified this decision by claiming that although the ballots are indeed documents they are simply "unavailable" to public scrutiny, thereby illegally inserting a new reserve category into Mexico´s freedom of information law.3
This case is a crucial test for Mexico's fledgling access to information legislation and may mark the beginning of the end of the country's significant advances in the area of transparency. Other government agencies undoubtedly will learn from the example of the IFE and the Electoral Tribunal to construct equally ad hoc arguments to close down access. In a country where the typical attitude towards the law has been "obey but do not comply" ("obedezca pero no cumpla"), Mexico's freedom of access law could quickly be hollowed out as the country moves down the slippery slope of permitting "access" without achieving real "transparency".4
The present article is divided into three parts. First, I review the 2006 electoral results and the actions of the electoral authorities in order to demonstrate that there is reasonable doubt with regard to the results of the past presidential elections. Although there is no smoking gun, there are enough strange things about the results to merit further investigation. Second, I critique the IFE's and the Tribunal's responses to the request for access and demonstrate that there is no question that according to Mexican law the ballots are public documents once the election has been completed. Third, I draw out the implications the IFE's rejection of access has for the future of transparency in the country.
II. REASONABLE DOUBTS AND UNCERTAINTIES
The 2006 election was a mess at the very least. Even scholars who defend the behavior of the electoral authorities admit that at least 46.7% of the tally sheets from the presidential election had some sort of counting error.5 López Obrador's Coalición por el Bien de Todos (CBT) claims that there were irregularities in 72,197 of the total of 130,477 ballot boxes, or 62 percent of the total.6 Independent reviews have brought this number down to about 61,192 of the boxes, almost 50% of the total.7 But this is still an enormous number.
Many of these mistakes are apparently minor, with one or two extra or missing ballots per box. But we should remember that the difference between Calderón and López Obrador was only 233,000 out of a total of more than 41 million votes. Only 0.58 percentage points separated first and second place, amounting to approximately 1.8 votes per ballot box. To alter the results it would have been enough for a single ballot to be changed incorrectly from the López Obrador column to the Calderón column.
Many of the "mistakes" were not so minor at all. Hundreds of ballot boxes showed differences of dozens or even hundreds between the number of ballots received, votes issued and blank ballots left over.8 The limited partial recounts conducted by the IFE and the Electoral Tribunal only strengthened suspicions that something had gone wrong. For instance, the first partial recount of less than 2.2% of the ballot boxes (2,864 in total) ordered by the IFE on July 5th led to a change of 4,594 votes in only 116 of the ballot boxes.9 Some specific examples include ballot box 740B in the State of Querétaro where the number of votes for Calderón went from 431 to 224, a loss of 207 votes at a single polling station. In Mexico City, the IFE recounted 12 ballot boxes in the 27th district yielding a result of 120 additional votes for López Obrador and 26 fewer for Calderón.
These are only a couple of examples that the author has been able to document on his own. To date, the IFE has not issued a full report on the results of this preliminary recount. In its final report of the events of the 2006 elections published this past January, 2007,10 the institute states that it opened 2,864 ballot boxes and presents a summary table which indicates how many boxes were opened in each state. But it does not provide any information at all on how the results changed as a result of this recount nor a ballot box by ballot box summary of what occurred during the recount.
The IFE waited weeks before releasing preliminary data on its recount. For two weeks no one even knew how many boxes had been opened or where those boxes were located. On July 20th the IFE finally posted on its web site the locations of the ballot boxes where the recounts had taken place. But it was not until a week later, on July 28th, that the IFE announced the total change in the vote which had resulted from the recount. López Obrador apparently lost 14,350 votes and Calderón 12,854, a net gain of approximately 1,500 votes for Calderón in this initial recount.11
But we should keep in mind that the boxes opened during this process were only those which the IFE formally authorized opening. This is by no means a statistical sample and it is possible that the IFE may have only opened those boxes which promised to not affect the overall result. For instance, José Antonio Crespo has recently documented12 using data from the IFE itself that in District 15 of the PAN dominant State of Jalisco the IFE did not authorize the opening of a single box even though 150 of the sheets had inconsistencies, some of which were quite large including cases of discrepancies of 322, 117,106, 93, 83 and 70 ballots. In District 1 of PAN stronghold Aguascalientes the IFE failed to open a hundred boxes with tally sheets containing inconsistencies, some of them quite large including cases of discrepancies of 167, 135, 96, 59, and 56 ballots. In District 3 of Oaxaca the IFE only recounted two of the seventy boxes with inconsistencies, failing to open boxes in which the discrepancies were in the order of 1,469 ballots in one case and 131, 107, 53 and 41 in other cases.
The total amount of change in votes in this initial recount (over 27,000) is a worrisome number in itself given that only 2,864 of the 117,287 boxes (less than 2.2%) were actually reviewed by the IFE. If the same number of irregularities were present in the other 97.8% of the ballot boxes the total amount of irregular votes would amount to 1,200,272 votes, more than five times the difference between the first and the second place candidates. But most importantly, the fact that the IFE to date has not released a ballot box by ballot box summary of the results of this recount only raises suspicion as to the micro-foundations of the vote changes that occurred.
On August 5th, the Federal Electoral Tribunal ordered a second partial recount in 11,839 ballot boxes, approximately 9% of the total.13 In this decision the Tribunal applied questionable legal reasoning to categorically rejected the possibility of a full recount. The decision states that:
…each voting center is unique, made up of different actors, located in a different place and surrounded by a different context, (for that reason) the events and occurrences which take place in one, do not have interconnection with the others, in particular if the irregularities are due to the citizens who form part of the tables for the reception of the vote (SUP-JIN-212/2006-Inc.1, my translation).
The justices here state that each voting booth is a totally isolated universe without any "interconnection" with the other booths or universes. According to this logic, it is never valid to interpret a specific piece of information as representative of a larger phenomenon. The Tribunal therefore decided to order a recount only in those specific booths where according to its own strict criteria the irregularities in the acts were sufficiently large so as to merit recounting. Importantly, to date the Tribunal has failed to make public the specific criteria it used to decide which boxes to open and which ones to leave untouched.
This "interconnection thesis" breaks with centuries of research on social phenomenon, bringing us back to a radical empiricism which not even the positivists of the Vienna Circle would have dared to defend. Any social scientist who knows the first thing about statistical sampling knows that a small sample is enough to gain trustworthy information about a much larger universe. This is particularly true in the case of social phenomenon like the organization of elections where the poll watchers for each one of the electoral booths had been trained by the very same government institution and subject to the same political, social and media pressures.
If the Tribunal had ordered a recount in a statistical sample of the voting booths this would have significantly strengthened the results of its partial recount. Otherwise, we are left with and entirely slanted sample which cannot say much about the rest of the voting boxes. Some analysts have argued that since the boxes which were opened in this partial recount were the ones which had been challenged by López Obrador, the rest of them must have fewer irregularities than the ones which were opened. But this does not necessarily evident to the naked eye. In the absence of a scientific sample, judgments about the content of the unopened boxes are grounded in pure speculation.
Regardless, the Tribunal's decision to recount 9% of the boxes was a step in the right direction and made a clear public statement about the deficiencies of the IFE's own partial recount. The Tribunal's argumentation led it to only order a recount in those districts where according to its analysis the IFE should have done so in the first place during the July 5th district counts. This was therefore equivalent to a public reprimand to the IFE for having only recounted a quarter of the boxes it should have investigated and especially for having failed to open boxes which had been questioned by Lopez Obrador's CBT.
The Tribunal's partial recount also fed suspicions that something might be wrong with the election results. After the recount was completed the CBT released its own analysis of the results.14 According to the López Obrador camp the recount revealed that 3,873, or 33%, of the recounted booths had extra votes and 3,659, or 31%, had missing votes. In total 7,532 voting booths, or 65%, presented irregularities which amounted to a total of 119,744 altered votes (including both extra and missing ballots). In addition, the CBT claimed that in "practically all of the districts where the recount took place there were tens of thousands of boxes without seals and hundreds of thousands of envelopes which contain the electoral votes with violated seals and without signatures and even missing electoral boxes".15
Calderon's Party of National Action (PAN) downplayed these supposed irregularities. It pointed out that the recount had been ordered precisely to correct for the problem of extra and missing votes and that the voting boxes and envelopes had been opened previously by an explicit judicial order.16 The PAN has also emphasized that the recount apparently led to a minimal change in the difference between Calderón and López Obrador. A comparison between the final numbers presented by the Tribunal on September 5th and the original IFE numbers from the July 5th district count reveals that Calderón lost at least 2,756 votes and López Obrador gained at least 2,101 votes as a result of the Tribunal's recount, a total change of almost 5,000 votes.17
Nevertheless, the numbers are surely significantly higher since this estimate does not include the ballot boxes which were recounted and then annulled entirely by the Tribunal because of serious irregularities. The Tribunal annulled 80,601 votes for Calderón and 75,355 votes for López Obrador.18 Since almost all of the boxes annulled were from the recounted districts, it appears that López Obrador gained an additional 5, 246 votes from the recount which are not taken into account in the above calculations. The Tribunal's partial recount therefore reduced the difference between Calderón and López Obrador by approximately 10,000 votes.
If we were to extrapolate these changes in 9% of the ballot boxes to the rest of the election Calderón's margin of victory would have reduced significantly, from 233,000 votes to only about 122,000 or approximately 0.30% of the total vote. This would not have been enough to change the final result but it certainly would have put the election results under even closer scrutiny.
But all of this is speculation. Given the widespread prevalence of inconsistencies, it is by no means out of the question that a total recount could have given the victory to López Obrador. The problem is that, just as with the IFE's partial recount, to this day the Electoral Tribunal has still not released a final ballot box by ballot box account of the results of its own recount. In its final decision issued on September 5th, 200619 the Tribunal offered the vote totals for each one of the country's 300 voting districts but did not provide any information at the voting booth level nor any information on how the vote had changed through the two partial recounts. One can wade through thousands of pages of legal documents which lay out the Tribunal's previous decisions, but even then one does not get the entire picture. For instance, while some of the Tribunal's August 28th decisions with regard to the annulment or no of hundreds of ballot boxes include specific information on the recount, others simply do not.
The electoral authorities have not provided enough information for outsiders to independently verify the final vote count. In the face of this ignorance, the independent observer is left with a stark choice: trust or skepticism. Perhaps if elections had been clean in Mexico throughout the entire 20th century and there were absolutely no questions about the independence of the electoral officials we could blindly trust the results. But the fact is that Mexico is known internationally for its expertise in electoral fraud and all of the sitting IFE councilors were nominated and appointed by political forces (PAN, plus a fraction of the PRI) who are now aligned with Felipe Calderón. A healthy dose of skepticism is therefore the best service we can pay to democracy.
Additional considerations discourage blind trust in the electoral results. One of the CBT's most serious accusations was that the seals on the storage rooms, ballot boxes and envelopes containing the votes had been systematically violated. This could indicate possible vote tampering by the IFE itself or external actors during the days and weeks after the elections. This is precisely what happened after the openly fraudulent 1988 elections. According to Mexican law, only the Electoral Tribunal can require the seals to be broken to extract specific documents from the electoral boxes or, in special cases, to conduct a recount of the ballot. Not even the IFE is authorized to open up the boxes on its own. The Tribunal did authorize such actions on some occasions but by no means in such a broad manner as was apparently the case.20
The Tribunal's total silence on the issue of the conditions in which it found the ballots during the partial recount is quite worrisome. They do not mention a word about the topic either in their August 28th resolutions or their final September 5th decision. We are therefore left with the CBT's accusations and the PAN's abstract responses, once again without sufficient information to independently evaluate what actually happened to the electoral boxes.
There is an additional detail which clinches the case for suspicion. In many of the states where López Obrador came out ahead there were tens of thousands of votes fewer counted for President than for Senator, while in many of the states where Calderón won there were tens of thousands of votes more for president.21 In total, there were over 250,000 votes fewer votes for President than for Senator in six states (Tabasco, Michoacán, Veracruz, Oaxaca, Chiapas and Mexico City), all of these won by López Obrador. There were over 80,000 votes extra for President than for Senator in seven states won by Calderón (Nuevo León, Sinaloa, Baja California, Morelos, Sonora and Chihuahua). It is very difficult to find a rational explanation for this discrepancy. Why would citizens in some states reject all presidential candidates while those of other states reject all senatorial candidates? The suspicion of course is that somehow presidential votes were removed in the pro-López Obrador states and added in the pro-Calderón states.
In anticipation of possible difficulties with the electoral results a group of independent academics and civil society leaders created before the elections a special Citizen's Commission ("Comité Conciudadano de Seguimiento del Proceso Electoral" -CCSPE-) designed to keep a watchful eye on the electoral authorities. The committee included over twenty well known experts from a diversity of political backgrounds, some closer to Calderón, some closer to López Obrador and most entirely independent of all political parties.22 This committee issued a series of reports which also belie the idea that the 2006 election was perfect and that the results should be trusted blindly.
The Committee's initial report on the elections23 clearly states that it "shares with other citizens a reasonable doubt about who was the winner of the presidential elections". This reasonable doubt arises because of the inadequate information offered by the IFE about the electoral results, the pressure placed by the General Council on the Local IFE councils not to recount electoral boxes, the absence of party representatives in a high percentage of electoral booths,24 the deficiencies in the training of poll watchers, human errors in vote counting, as well as possible problems with the nullification of valid votes. In addition, the committee received reports about the deficient quality of the specialized ink used to mark voters' fingers and the relaxation of the criteria for replacing poll watchers. The report also cites an Alianza Cívica study on the election which states that approximately 30% of the voting booths observed presented cases of citizens who had been inexplicably removed from the voters list. The report concludes with the following statement: "We are concerned about the triumphalism of the IFE's General Council, which devalues the seriousness of the situation which is a product of its problematic and insufficient actions to control the electoral process and the exercise of its authority".
Was there fraud? We don't know. Was the election a mess? Definitely. Might Mexico's political actors have been intelligent and experienced enough to take advantage of this situation to manipulate the electoral results? Quite likely. Can we be absolutely certain that Felipe Calderón received more votes than López Obrador on election day? Definitely not.
III. TRANSPARENCY AND ELECTORAL BALLOTS
An independent examination of the electoral ballots used during the past presidential elections in Mexico is merited. There is a reasonable doubt about the electoral results and such an exploration could help clear up many of the questions. Indeed, it could easily end up legitimating Calderón's victory. The newspaper study conducted by the New York Times, the Washington Post and the Tribune Company after the 2000 elections in the United States concluded that Bush would have won anyway even if a full recount had taken place in Florida.25 This study also provided important information on election inequality since it demonstrated that if all Florida citizens had equal access to effective voting machines and clear ballots then Gore would have won. Faulty voting machines and complicated ballots in predominantly African American districts led to the incorrect nullification of countless numbers of pro-Gore votes.26 A comparable study of the Mexican elections of 2006 would undoubtedly provide a great deal of important information and help improve elections in the future.
But would such a study be legal? Both the IFE and the Electoral Tribunal have argued that the law does not allow them to give the public access to the ballots. In this section I show that there is no question about the legality of providing access. The IFE's rejection of the numerous FOIA requests which have been brought by citizens and journalists is based on an overtly politicized and highly questionable interpretation of the law. The Tribunal's ratification of the IFE's decision corrects some of the deficiencies of the original response, but simultaneously opens up a dangerous new escape hatch for authorities who wish to avoid freedom of information requests.
Mexico stands out internationally for the strength and effectiveness of its new freedom of information law (Ley Federal de Transparencia y Acceso a la Información Pública Gubernamental -LFTAIPG-), passed in 2002,27 It offers an excellent example of the new breed of FOI legislation which recently has grown up in new democracies throughout the developing world. Mexico along with South Africa, South Korea, Thailand and over twenty countries of Eastern Europe are setting the standard for access to information throughout the globe. These laws tend to be much stronger than both the historic laws of Sweden, Colombia and the U.S. and recently passed laws in older, wealthier democracies such as Germany, the U.K. and Switzerland. In a new democracy, access to information is not just a good idea which may help combat corruption and make government more effective, but an essential right of the people which will allow citizens to bring democracy to its ultimate consequences. After decades of authoritarian rule, social actors are tired of being treated as passive subjects and eager to take up their roles as active citizens. The design and implementation of a new FOI law is therefore inspired by idealistic hope and hard hitting societal demands.
Mexico's FOI law is firmly grounded in the principle of full disclosure. The law states in no uncertain terms that the interpretation of the legislation and the corresponding rules should be based on the "principle of maximum publicity and accessibility of the information possessed by the obligated agencies" (LFTAIPG, Article 6). What this means is that all government held information is public (LFTAIPG, Article 2), except for that which is explicitly reserved or classified by law. This is, after all, the point of freedom of information legislation. Such laws turn the tables from a principle of "need to know" to one of "right to know". Before the global transparency revolution, citizens needed to justify why they wanted to gain access to a particular document. Now it is the government that needs to justify why it cannot provide access.28
The Mexican law creates three different categories of information: 1) Public information to which access must be granted, 2) "Reserved" information which the government has the right to keep secret for up to 12 years (Articles 13, 14), and 3) "Confidential" information which can never be released (Article 18).
The only confidential information is "personal data" ("datos personales") such as addresses, sexual preferences, weight, height, etc. In addition, other documents, such as commercial secrets, can be confidential if explicitly provided for by the corresponding law and if individuals report the information as confidential when they hand it in to the authorities.
The electoral ballots therefore could not possibly be categorized as confidential documents. The act of voting is secret. No one should be able to tell for whom a particular citizen voted. But precisely because they are secret, electoral ballots do not contain any marks which could allow a third party to identify the person who filled out a particular ballot. Electoral ballots, qua documents, do not contain any personal data at all.
Nor would it be possible for citizens to declare that they wanted their ballots to be secret upon voting. Voting is a quintessentially public act. The very point of voting is because you want your preference to be known. With confidential ballots, officials and citizens would have no way of adding up voters preferences.
The law stipulates a broader set of criteria for determining whether information can be "reserved" from public view for a period of time. Article 13 indicates that the government can reserve information for up to twelve years if its release might significantly harm national security, public safety, international relations, economic stability, a person's life or causes serious harm to the government's ability to assure the rule of law, the investigation of crimes, or tax collection, etc. Article 14 then states that commercial secrets, judicial files and administrative procedures which have not been completed, and deliberative opinions also can be reserved for the same time period. Finally, agencies can reserve information when another law "explicitly" considers the information confidential, reserved, "commercial reserved" or "government confidential".
Significantly, information cannot be reserved simply because some part of it might affect some element of the reserve areas. The principle of maximum publicity requires the government to fully justify the lack of disclosure based on the grounds that publicizing the information would significantly harm the reserve areas and that this harm would be larger than the damage which would result from keeping the information secret. The law therefore requires the application of a "harm test" in the application of the reserve areas.
In its recent response to various requests for access to the ballots,29 the IFE has argued that disclosure of the ballots would significantly damage "the State´s role of organizing the federal elections and the other constitutional attributions which are assigned to the Federal Electoral Institute… because by giving the complainant access to the requested documentation it would harm the central principles of the electoral function…especially those of certainty and legality" (p. 62). The IFE's central argument is that the Constitution establishes the "definiteness" of each stage of the electoral process. For instance, Article 41, subsection IV of the Constitution states: "In order to guarantee the principles of constitutionality and legality of electoral acts and resolutions, there will be established a complaints system… which will give definiteness to the different stages of the electoral process and will guarantee the protection of the political rights of citizens to vote, be voted and association…".
The idea here is that the actions of the IFE can be appealed to the Judiciary, specifically the Electoral Tribunal, and that the Tribunal's decisions are final. Once the Tribunal has decided on a particular case or stage of the electoral process (campaigns, organization of the elections, reviewing voting booths, conducting the final count, pronouncing the winner, etc.) political actors cannot return to a previous stage to question its validity. This makes common sense. If, for instance, parties could continually present complaints against the results of each voting booth the election would literally never end. The losing party would go through each of the 130,000 booths one by one and then start again from the beginning, preventing a final decision from being reached.
The IFE argues that in the interest of assuring the "definiteness" of the electoral process Article 254 of the electoral law (Código Federal de Instituciones y Procedimientos Electorales -COFIPE-) states that the IFE must destroy all of the electoral documents once the final decision of the election is pronounced by the Electoral Tribunal. The idea is that permitting direct citizen access to the electoral ballots would confuse the people and would bring the electoral process back to the stage of counting up the ballots which has already been completed. This in turn would threaten the central task of the IFE of assuring the "certainty" of the electoral results.
The problem with this argument is that Mexico's freedom of information law does not stipulate anywhere that harm to "the State's role of organizing elections" or harm to "the central principles of the electoral function" are valid reasons for reserving information. The list of reasons for reserving information included in Articles 13 and 14 is a restricted, exclusive list, not simply an open list of some possible reasons. In order for the IFE to justify reserving the information it would have to offer a specific argument that access to the ballots would cause significant and direct harm to one of the areas listed. It cannot simply make up new ad hoc grounds for reserving information based on its particular interpretation of the law.
In addition, Article 254 of the electoral code does not require that the IFE destroy the electoral ballots immediately after the elections are decided by the Tribunal. Indeed, after the 2000 presidential elections the IFE waited over six months, until January of 2001, to destroy them. To this day, the IFE still holds the ballots used in the past 2006 presidential elections.
Apparently aware of the fact that their general defense of the principle of "definiteness" is insufficient to formally reserve the electoral ballots according to the access to information law, the IFE also claims that national security and "public order" is at stake.
This fundamental right (of access to information), as in any democratic state, is subject to restrictions, which not only have the purpose of protecting the intimacy of people, but also of guaranteeing public order, social interest, as well as primordial operations of the Mexican state, including, for instance, national security… In this way, and by way of example, one can mention the fact that the Fundamental Law of Bonn and the German Constitutional Court recognize the concept of "militant democracy", which restricts the freedom of expression or information when their exercise threatens certain public values which are of fundamental importance for the collectivity, including all of the procedures for deciding upon the makeup of the institutions of government (pp. 52 and 53).
The Institute here suggests that the request for access to the ballots used in the past presidential elections is an "attack against fundamental public values" equivalent to openly supporting a fascist takeover of government. The Fundamental Law of Bonn and the decisions of the German Constitutional Court cited by the IFE refer to norms designed to prevent the resurgence of Nazism or other extremist political movements in the country. German institutions have decided that it is acceptable for the government to silence such violent and anti-democratic expressions even though this implies the violation of basic civil rights like freedom of speech and expression. In parallel fashion, the IFE argues that the journalists who have requested access to the ballots threaten the very foundations of democratic politics and should have their basic civil rights stripped from them.
Such a leap of logic actually amounts to an implicit confession that something might be seriously wrong with the way in which the votes were counted during the past presidential elections. If everything were in order, an independent review of the ballots would only strengthen Calderón's legitimacy and ratify the IFE's defense of the cleanliness of the elections. The supposed threat to "public order" could only raise its head if the study revealed that the elections were indeed a mess and López Obrador might in fact have been the victor.
But even then it would be wrong to blame the disturbances which would most likely result from this discovery on the act of permitting access to the ballots. The real culprits would be the IFE and the Electoral Tribunal for not having cleared up the electoral results in time. They would be the actors responsible for eroding the confidence in Mexico's democracy, not the journalists who have requested access. The very point of a FOI law is to facilitate independent review of government actions.30 Shooting the messenger never resolves the problem.
Beyond these statements concerning "definiteness" and "public order", the IFE also tries to make the argument that, formally speaking, the electoral ballots are not "government documents" at all and therefore not subject to the access to information law. The IFE cites article 3, subsection III of the FOI law which defines "documents" as:
The files, reports, studies, acts, resolutions, oficios, correspondence, agreements, orders, directions, circulars, contracts, convenios, guides, notes, memos, statistics or any other registry which documents the exercise of the faculties or the activity of the obligated subjects and its public servants, independently of its source or date of elaboration. The documents can be in any form, written, printed, sound, visual, electronic, computerized or holographic.
According to the IFE, the electoral ballots do not document the activity of the electoral authority or its public servants since they are not filled out by IFE employees but by independent citizens. The ballots therefore "are not documents legally speaking, they are only the material expression of the electoral preference of the voting citizenry. The ballots register the popular expression of the vote and this expression cannot be categorized as an activity or faculty of the IFE and its public servants" (p. 11).
Here the IFE conveniently ignores the fact that the ballots are designed, printed, guarded and distributed by the IFE itself and are the basic input for the institution to carry out its central function of deciding who has won the elections. As discussed above, the IFE itself argues in other parts of its response that the ballots are central the "State's role in organizing elections". The fact that citizens and not public servants are the ones who actually fill out the ballots does imply that the documents themselves are private material. Indeed, the transparency law's definition of "information" as "that contained in the documents which the obligated subjects generate, obtain, acquire, transform or conserve in any manner" (article 3, subsection V) is explicitly designed to avoid such a limitation of the reach of the law. The law is an access to "information" law not an access to "documents" law.
In any case, if the electoral ballots are not "documents", what then might they be? The IFE argues that they are simply the "material expression of electoral preferences". But the "materialization of an expression" is, precisely, a "document". Merriam-Webster's Dictionary offers the following definition of the word "document":
1 a archaic: PROOF, EVIDENCE b: an original or official paper relied on as the basis, proof, or support of something c: something (as a photograph or a recording) that serves as evidence or proof.
2 a: a writing conveying information b: a material substance (as a coin or stone) having on it a representation of thoughts by means of some conventional mark or symbol c: DOCUMENTARY.
3: a computer file containing information input by a computer user and usually created with an application (as a word processor).
The Dictionary of the Spanish Royal Academy of Language offers almost exactly the same definition of the word "documento" in Spanish. In its efforts to bend the law, the IFE has gone to the extreme of implying that the electoral ballots are some sort a metaphysical material which we simply cannot define.
This reinterpretation of the nature of the ballots is in direct contradiction to the electoral code as well. Article 254 which speaks of the destruction of the ballots explicitly states that the material to be destroyed is "documentation". In an important decision the Electoral Tribunal has also recently categorized electoral ballots as "documents", distinguishing them from other electoral "material" such as the voting booths, boxes, crayons, etc.31
Finally, the institute also claims that Article 254 implies that the Institute is not permitted to do anything else with the ballots except destroy them once the elections are over. According to the "principle of legality", government agencies are only allowed to do what they are explicitly authorized to do by law. They may not violate a law simply because they think it is inappropriate or unconstitutional. Only if the judiciary explicitly intervenes are they permitted to do so.
Here the IFE cites an important case resolved by the Electoral Tribunal in 1998.32 In this case the Tribunal decided that the IFE is not permitted by law to conduct independent investigations of the content of the electoral packages.
The Federal Electoral Institute is prohibited from establishing different procedures from those contemplated by the Federal Code of Electoral Institutions and Procedures. Ordering the practice of studies or investigations, to be carried out by external parties or by the organs of the Institute, as well as by its own personnel, involving actions different from those set out by law, of the documentation contained in the electoral packages and the information inscribed in the electoral documents of which they are made up, constitutes a practice which is incorrect (viciado) in its origins, because of the lack of faculties in this direction.
But this case was decided in 1998, four years before the passage of Mexico's access to information law. At that time it was indeed the case that there was no law which explicitly ordered the IFE to provide access to the electoral boxes after the elections were over. But now there is such a law. The access to information law passed in 2002 explicitly states that all documents in the hands of government agencies are public unless they fall under one of the categories of reserved or confidential information.
This past April 25th, 2007 the Electoral Tribunal ratified the electoral institute's decision to deny access to the ballots (SUP-JDC-10/2007 y SUP-JDC-88/2007 accrued). To its credit, the Tribunal flatly rejected the institute's arguments with regard to the nature of the ballots and the possible threat to national security. According to the justices, the ballots are without question "documents" and they do not fall into any of the reserve areas, including a possible risk to national security, listed in article 13 of the law.
The Tribunal grounds its denial of access on two different clauses in the transparency law. First, article 14, subsection I states that information can be withheld if there is an "explicit indication" in another law which states that the requested documents are "confidential, reserved, commercial reserved or government confidential". The justices argue that since access to the ballots is strictly protected by law and then destroyed once the Tribunal has decided on the election, it is as if the ballots were declared "reserved" or "confidential" by the electoral law.
The problem with this argument is that the law states that there must be an explicit indication ("disposición expresa") of the confidentiality or reservation of the information. There is no room to construct analogies. Indeed, as argued above, the principle of maximum publicity demands that all information be declared public unless it explicitly falls into one of the reserve areas. In their decision the justices do their best to construct a case for the reservation of the ballots, but they are simply unable to find an explicit mention that prohibits access once the elections have been decided.
Perhaps aware of the weakness of this first argument, the Tribunal offers a second one to strengthen its case. The justices point out that article 42 of the transparency law only allows certain forms of access to information:
The obligation of access to information will be considered completed when the requestor is allowed to consult the documents in the place where they are located; or by issuing simple or certified copies or through any other medium. Access will only be given in the form in which the document permits it, but it will be given totally or partially, upon request.
The crucial phrase here is that "access will only be given in the form in which the document permits it" ("El acceso se dará solamente en la forma en que lo permita el documento de que se trate"). Indeed, the Tribunal conveniently does not cite the previous sentence which lists the different forms of access. According to the justices there is no way to access the ballots, since the electoral packages which contain them are sealed and the law only permits the electoral institute to open them on judicial order during the phase of judgment and evaluation of the election.
Theoretically, this would not prohibit access after the election has been decided when the ballots are no longer of any use to the judicial authorities. Today, the ballots are still sitting inside the electoral packages simply waiting to be destroyed. Nevertheless, the Tribunal argues that even though the ballots still exist and are no longer useful, citizens still cannot access them because the law does not enable the institute to open the packages.
The ballots are therefore simply "unavailable" ("indisponible") to the public. In this way the Tribunal avoids having to find an explicit clause which declares the ballots reserved or confidential and also escapes from the sophistry of pretending that they are not documents or might affect national security. The ballots are public documents but not available. Access would be permitted, but it is unfortunately just not possible.
But this argument fails to take into account the fact that article 42 explicitly refers to forms of access, not to forms of reservation. Title I, Chapter III (articles 13-19) is the section of the law which refers to "reserved and confidential information" and contains all of the stipulations which can apply in order to reject a freedom of information request. Article 42 belongs to Title II, Chapter III (articles 40-48) on "the Procedure of Access to an Agency or Entity". This article therefore assumes that access has been granted and regulates the way in which that access can be provided. It is an inappropriate leap of logic to imagine that this clause can give a government agency reason to deny access entirely.33
The Tribunal's arguments are much more sophisticated than the electoral institute's. But this should not distract us from the fact that they are equally aggressive to the freedom of information law. With this decision, the Electoral Tribunal takes on full legislative powers to inscribe an entirely new category of documents into the law, that of "unavailable documents". By reinterpreting Article 42 as a cause for withholding information, the Tribunal has opened up a massive new black hole into which government agencies throughout the county can throw uncomfortable documents they would like to protect from the public eye.
IV. IMPLICATIONS
The case of Mexico's electoral ballots is a landmark case in the international struggle to expand and exercise the right to know. It has already been highlighted by international NGOs such as the National Security Archive in Washington, freedominfo.org, and Global Exchange in San Francisco. Now that the Mexican electoral authorities have been allowed to get around the law using such ad hoc argumentation, other agencies will surely follow suit. Mexico's international leadership in the area of transparency could be at risk.
In its response to the ballot request the IFE makes a particularly revealing statement:
There is no necessity for sacramental phrases or absolute and overarching dispositions, which literally establish each and every one of the conditions for the exercise of a public subjective faculty (of access to information), since these conditions of exercise can be deduced validly from the systematic and functional reading of the norms applicable to a concrete case, given that the authorities do not proceed in an arbitrary manner, deciding what ever they think is convenient to do (p. 47).
Such a statement amounts to the unilateral erasure of the principle of publicity which gives force to FOI laws. Here the IFE is saying that we should not conceive of the access to information law as being grounded in fundamental, "sacramental" or "overarching" principles such as maximum publicity or the need to hold government officials accountable. Instead, we should proceed on a case by case basis to see whether access is merited in each particular situation.
A case by case analysis is of course always necessary, but such a procedure is not in conflict with the overall principles and exemptions of a FOI law. It is a false dichotomy to state that either we follow overall principles in the application of a law or we look at each individual case. The point of any law is precisely to set out broad principles and clear rules that should be followed when deciding specific cases. Mexico's access to information law clearly states that all information in the hands of government is public, except for that information which is explicitly reserved or classified by the law itself. As we have seen above, the ballots do not fall into either of these categories and therefore must be declared public documents. The law allows no room for special cases.
Now that the Electoral Tribunal has ratified the IFE's refusal to grant access to the ballots the door is left wide open for other agencies to also make up their own "special cases" in the interest of closing down access. For instance, one of the areas which has been particularly resistant to the new transparency law in Mexico is the area of government trusts (fideicomisos).34 These trusts hold government funds, such as customs duties, or a mixture of public and private funds, as in the case of the banking trust. Parallel to the case of the electoral ballots, the Secretary of Finance and Public Credit has used the fact that these trusts are not formally "government agencies" to avoid the transparency law, even though they clearly manage public monies. Mexico's Institute for Access to Public Information (IFAI) has waged an important battle to gain access to information on the trusts. The ratification of the IFE's decision by the Electoral Tribunal sets an extremely negative precedent in this regard.
Now that the IFE's refusal has been left standing this could lead to a massive reclassification of information in the hands of government. Officials could claim that any document which is eventually destroyed or requires special handling is simply "unavailable". In the oral arguments of the case justice Salvador Nava set out the road for this eventual spread of opacity. Nava argues that the electoral ballots are like the bills which the Central Bank holds in its safe rooms: citizens can know how many there are but they cannot go into the Bank to count the bills themselves.
Such an argument is a non-sequitor. First of all, central banks do not hold their reserves in bills, but in government bonds or other securities. Access to copies of these securities would have to be permitted. Otherwise, suspicions would arise with regard to the validity of the central bank reports since investors would have no way of independently verifying them. We would then be in a parallel, or perhaps even more dangerous, situation to the widespread suspicion about the results of the 2006 Mexican elections. Indeed, even if the central bank did hold its reserves in peso bills, access to certified copies would also have to be permitted.
The Tribunal's message is that citizens should be satisfied with government reports, that they should not be given the raw information and documents necessary to verify these reports. As a result, when a citizen or a journalist now requests the receipts of the cement purchases of the Secretary of Public Works the agency will be able to satisfy the request with a document which supposedly summarizes its purchases. There will be no need for it to hand over copies of the actual pieces of paper which document these purchases.
Democracy implies much more than free and fair elections. It also requires a fundamental sea change in the way that government relates to society. In order to fully institutionalize democratic politics, the governing class needs to reconceptualize its role and establish a truly open process of respectful dialogue with social actors.35 Unfortunately, the response of Mexico's electoral authorities to the requests to access the electoral ballots used in the past elections reveals that a profound faith in Machiavellian "reason of the state" and a rejection of citizen demands for transparency remain solidly entrenched among Mexico's top public officials. If this attitude does not change soon Mexico's democracy may be at risk.
