Introduction

In many newer democracies around the world, political parties or other political actors construct third-party enforcers, known as electoral management bodies (EMB), to manage elections, which allows them to take advantage of a longer time horizon and invest in their label without fear that other parties will commit fraud (^{IDEA, 2014}). Parties may want to manipulate the electoral institutions for their own benefit, but they must forge rules that are seen as fair both by voters and other electoral actors so that electoral outcomes are accepted, even by losers. However, as ^{Birch and Van Ham recognize (2017: 487)}, electoral management bodies are immersed in the political game they are supposed to regulate, and so, it is difficult to achieve “de facto independence from political influence…” This implies, then, a tension that has only begun to be recognized: just as parties create formally autonomous or semi-autonomous institutions to limit their actions for their own benefit -they also have strong incentives to cheat on these rules if possible and write laws so they can manipulate or weaken the EMB. This article examines a single electoral system -Mexico’s- to better understand party motivations of non-compliance. We find that leaders of political parties engage in two types of questionable behavior: first, they often collude among themselves to write laws that benefit themselves while harming other types of political actors, such as voters, potential party leaders and office seekers; and second, they ignore the negotiated rules and cheat against each other when compliance would help their long-term goals.

Party leaders often have two sets of interests when dealing with EMBs; one which they share with other parties; and one they do not.¹ Most party leaders and their counterparts want to shore up their ability to win seats through a variety of means such as procuring large public budgets, high barriers of entry for new parties, and control over their ambitious politicians via closed candidate selection.² So, even while party leaders seek to commit electoral fraud against their party rivals, they often cooperate with them to block other types of actors from participating more fully in the political arena. These dual interests can lead to the capture of the regulatory institution by the actors it is supposed to obligate to comply.

Because of the nature of repeated play implied in elections and in electoral rule making, the present paper emphasizes the “time frame problem”. Party leaders and candidates have strong incentives to bring home electoral victories by any means necessary, especially if they know the other party will cheat. If these leaders hold short time horizons in that post may resort to non-compliant behavior because electoral victories under their tenure promote their personal careers. However, if all party leaders continue the same strategy, then cheating and harassment of the electoral authority will prevail, leading to non-optimal outcomes, such as increasing voter rejection of parties. The fallout from duplicitous behavior may not affect the careers of the party leaders because the costs of non-compliance are paid in the future.³ Therefore, they hold strong interests to write rules they can manipulate in the present, and they are willing to forward the costs of their behavior because of the high probability they will not be leaders when negative consequences of their negotiations come to bear.

The literature on regulatory capture helps illuminate a central problem: industry players within a regulated sector are often able to capture the regulator and, in doing so, can raise the barriers to entry for new actors, allowing them to offer substandard “products” at higher costs. As a result, the customers -or voters in this case- often lose out. This work will show that the electoral regulatory agencies that make up the electoral management bodies have been -to a greater or lesser extent- captured by the parties that placed them, leaving them in a weak position against both parties and the executive branch.

One might ask why all nations do not exhibit captured electoral institutions and non-compliant parties. First, not every nation has a strong party system, especially not in newer democracies (^{Mainwaring and Scully, 1995}) and so they lack powerful parties capable of harassing electoral authorities. Second, other political institutions may shore up a weaker electoral authority, as ^{Birch and Van Ham (2017)} suggest. The question remains, however, whether new democracies are able to bring to bear other institutions to protect elections. Many new or challenged democracies in Latin America, such as Nicaragua, Ecuador, Venezuela, Guatemala, and Honduras, share Mexico’s problematic informal behavior.⁴ It may well be that EMBs in developing democracies are weaker in practice than they appear due to general institutional debility.

Mexico is rightly famous for its pacific transition to democracy in 2000, and the important role that political parties and electoral reforms played in the defeat of the long-lived authoritarian PRI (Party of the Institutional Revolution) regime (^{Becerra et al., 1997}; ^{Eisenstadt, 2004}; ^{Greene, 2007}; ^{Langston, 2017}; ^{Merino, 2004}; ^{Lujambio, 2000}). The PRI, the center-right National Action Party (PAN), and centerleft Party of the Democratic Revolution (PRD) were the cornerstones of the nation’s party system and dominated both votes and seats between 1991 and 2015. The PAN controlled the presidency from 2000 to 2006, and another PAN president governed from 2006-2012. The PRI roared back into the executive in 2012 under the leadership of Enrique Peña Nieto.⁵ However, in the concurrent elections of 2018, the three traditional parties were routed by an outsider party (National Regeneration Movement or Morena), led by a charismatic politician. As of now, two of the three traditional parties may not survive the next election cycles, which could spell the end of the nation’s current party system, underlining the serious outcomes of party tactics against their EMBs and other political actors.

This paper examines the process of electoral institution rulemaking and implementation during elections and at other moments during the transition years and into the first years of democracy (1996 through 2014) using official documentation, secondary literature, and newspaper accounts. I also interview several relevant actors to better capture party leader strategies over time: former leaders of the nation’s two electoral authorities (known as the National Electoral Institute (INE or IFE) and the Federal Electoral Tribunal (Tribunal Electoral del Poder Judicial de la Federación, known as the Trife or TEPJF);⁶ former party leaders; and academic experts on the topic.

Electoral institutions and non-compliant actors

Generally speaking, institutions are rules that allow actors to gain greater benefits from cooperation by defining and limiting behavior (^{Knight and Sened, 1998}; ^{Miller, 2005}; ^{North, 1990}). Institutions are outcomes of social decisions chosen by boundedly rational actors in negotiations with an aim of winning benefits for their group, while also allowing others to benefit. They solve the problem of collective action by creating credible sanctioning mechanisms and are meant to regulate behavior into the foreseeable future, although they are often modified. Even if an authoritarian group participates in negotiations while planning to ignore them in practice, the institutionalization of specific rules, the employment of large numbers of personnel, and the procurement of budgets allow actors to coordinate around rules which can compel all actors to comply over time.

We define efficient institutions as those that make actors better off through monitoring and sanctioning, often by creating a third-party enforcer that adjudicates, monitors, and applies punishment for cheating. Inefficient institutions, on the other hand, are those that make some actors worse off, either because it is difficult to coordinate or because stronger groups refuse to allow changes to rules. For example, an inefficient set of political institutions may fail to create strong property rights, allowing the executive to confiscate with relative impunity, which reduces future investment and economic growth (^{North, 1990}). Inefficient rules can also lead to the eventual destruction of the system in which the actors are immersed, whether it be a specific market sector or a stable party system.

Electoral institutions in a democracy are a type of political institution that grows out of group bargaining, whether it be in a single shot game or through incremental rule creation and change. Actors care about the present and the future; and to protect their benefits into the future, they use their political power in the present to assure their position, even though the future is inherently uncertain (^{Acemoglu, Johnson, and Robinson, 2005: 293)}. At the most basic level, the actors involved in electoral rulemaking usually try to protect their seat or vote count or enlarge it (^{Colomer, 2005}).⁷

Party leaders normally modify electoral rules in reaction to external threats, such as franchise expansion, new parties, or voter preference change (^{Benoit, 2007}; ^{Rokkan, 1970}). Electoral rules are made by parties, for parties, with the expectation that these same organizations will then comply with the rules they have created, because over the long-term, they benefit from them. However, it is clear that parties or other political actors in many nations cheat or manipulate the very rules they helped create. By far the greatest attention is paid to autocratic or electoral authoritarian regimes and their efforts to keep opposition parties from competing on a level playing field (^{Birch, 2008}; ^{Eisenstadt, 2004}; ^{Fortin-Rittberger, Harfst, and Dingler, 2017}; Levitsky and Wey, 2009). However, far less is written about why parties in less consolidated democracies refuse to comply with rules they negotiated.⁸

The central actors in this case are those with control over the electoral system, especially party leaders, congressional leaders, and the chief executive. Their preferences are to win elections, take control of the government apparatus, and win more resources. The first two goals are winner-take-all in nature, so that parties compete over them: what one party wins, another party loses. However, party leaders often decide to build a cooperative electoral authority where all are better off because they cooperate in anticipation that the other will do so. This cooperative strategy is only possible if the actors believe the electoral authority is strong enough to oblige the other parties to do the same; otherwise, they must cheat to avoid the worst pay-off. Without a strong third-party enforcer, parties find themselves in a single shot prisoners’ dilemma game in terms of competition at the ballot box. Even in this situation, parties still share interests, such as raising barriers to entry and controlling candidate selection, which allows them to cooperate in other areas.

Estévez, Magar, and Rosas (2007) argue that relationship between Mexico’s electoral authority and the parties it regulates can be understood as a principal-agent model. In their view, the EMB is not independent or autonomous from the parties because the latter retain their ability to hire and fire councilors and control the budget of the EMB. The parties are principals and select their agents to serve on the electoral authority, and then exercise external pressure during their tenures: the councilors either accede to the parties’ demands or refuse to do so. If the councilors (or magistrates) refuse to follow their party sponsor’s bidding, the principals have various instruments to sanction them. In anticipation of this, councilors and magistrates duly represent the interests of the party that sponsored them.

However, the metaphor of a principal-agent relation between parties and electoral authorities is not entirely felicitous. In the economic literature, the agent is not responsible for monitoring and sanctioning her principal, who placed her in the post and pays her to make decisions that are consonant with her demands. Logically, it is difficult, if not impossible, for the agent to restrain her boss. A better metaphor to illuminate the relations between parties and EMB is regulatory capture, in which the government creates a regulatory agency with the capacity to monitory and sanction companies in a certain sector. Regulatory capture is defined as when the agency’s interests are more closely tied to those agents than they are to the public good (^{Stigler, 1971}).

Regulatory authorities learn after repeated interactions with the companies that their external clients have a strong preference not to comply with certain rules, but rather, pursue ends that help their bottom line. In this case, as ^{Stigler warned (1971)}, the agency no longer protects the public from industry abuse, but rather, helps erect barriers to protect them from competitive entry of other companies. By reducing the threat of external competition, sectoral companies can collude amongst themselves and raise prices, while not improving services. In the case of Mexican parties, the “industry” is made up of political parties, and their leaders are better informed than the voting public about how they pursue their goals, at the same time they able to nominate and remove recalcitrant members of the regulatory agency qua electoral authority (^{Carrigan and Coglianese, 2011}).

Mexico can be seen as a case of capture of the regulatory body in which the principal concern of the parties is to deny new entrants into the party system, to deny voters more party options, and to restrain party activists and office seekers in their efforts to capture more power from party leaders.

Three different types of models of authority have been discussed in terms of global electoral authorities; independent, governmental, and mixed (^{IDEA, 2014: 9-10}). Mexico’s model is clearly an example of the independent model, in that the EMB is not part of the executive, it holds responsibility for implementing the laws, and it can manage its own yearly budget (^{IDEA, 2014: 9-10}). Yet, formal independence does not make the EMB autonomous in practice, as the Mexican case demonstrates. The regulatory capture allowed the parties to cheat on the rules they negotiated and passed, which, over the long-term, led to a collective outcome that was worse for all.

On the other side of the electoral equation, one finds the leaders of the EMB, whose responsibilities are to create the conditions so that voters’ preferences are accurately translated into electoral outcomes. However, electoral authorities often hold other preferences that are closer to those of the regulated: they want to maintain their well-paid posts while conserving the outward perception of autonomy and fairness of their institution. Yet many work diligently to follow the law as they interpret it. At times, regulators must antagonize the parties because of the latter’s questionable behavior. If parties or the executive wish to weaken the EMB, they can do so through a serious of maneuvers than run from simple pressure to outright malfeasance, such as placing close allies in the EMB’s leadership councils, threatening to reduce budgets, removing councilors or magistrates without cause, pushing others to renounce their posts because of personal scandals, or promising the current electoral administrators and judges future access to political posts.

^{Lara (2017a: 158)} argues that when parties and their representatives cannot participate in electoral management bodies, they may criticize and undermine the work of the authorities, causing all to be worse off. However, it is also the case that when representatives do participate in EMBs, they can and will undermine the electoral authority when it furthers their electoral goals. Thus, participation in the EMB does not preclude non-compliant strategies; in fact, it can make double-dealing even easier.

^{Alarcón (2016: 20-21)} expresses what most specialists of Mexican electoral system argue: that the parties, in the course of trying to win elections, press the limits of the law, and in doing so, help reveal relatively important weaknesses or problems -thus, one sees that after criticisms of the 2009 election, the executive and later the three parties in congress opened up (somewhat) the party system to new actors via consecutive reelection and independent candidacies. The parties join together in congress to improve the laws so that these drawbacks are mitigated. A related argument is that the three largest Mexican parties used electoral reforms to shore up, pay back, or strengthen one of the three after each contested election. In this understanding, the reforms of the 1990s sought the incorporation of PAN and later the PRD; the 2007 reform was enacted to placate the PRD; and in 2014, the parties solidified the three-party political system. This broad explanation, however, does not uncover the incentives for continually weakening outsider political actors; nor does it capture why and how the parties undermined their third-party enforcer. What the following section will show, however, is that while party leaders correct large-scale problems for the large parties via negotiations, they also took advantage of each reform to impose higher costs on participation for smaller parties, ambitious politicians, and voters. Party control over the negotiation of new electoral rules in Congress and the executive’s desire to support his party helps explain these bi-polar outcomes, as does the short-sited nature of leaders’ interests.

Finally, one could argue that party leaders are not double-dealing actors who only think of short-term gains. Another line of inquiry in the theories of transactions holds that boundedly rational actors often make mistakes when negotiating, and in doing so, make rules that later create unintended -and often negative- consequences (^{Williamson, 1984}). It might be that boundedly rational actors change electoral laws to suit their immediate needs and do not properly calculate the future consequences. The complicated nature of party negotiation, the unknowable effects of different sets of rules, as well as the changes in electoral outcomes and technologies, can produce rules that have little relation with those that the parties set out to pass in congress. If this argument holds weight, then one should see corrections in later reforms to undo the self-inflicted damage.⁹ As this paper will show, intentional duplicity as well as bounded rationality under the stress of multi-party negotiations led to a series of problematic electoral rules.

Mexico’s voted transition and beyond

Mexico’s transition to democracy from 1988 through 2000 was a clear example of a “voted transition” (^{Merino, 2004}). In addition to political and economic crises, as well as an armed rebellion, the main drivers behind Mexico’s transition were the three largest nation’s parties, whose leaders negotiated a series of electoral reforms beginning in 1989 (which created the autonomous National Electoral Institute, or IFE) after the problematic victory of the PRI’s candidate in 1988, which culminated in the transformative 1996 reform. The 1996 reform freed the electoral authority from the interference of the still-hegemonic PRI executive, instituted voting stations and counting procedures that were run by randomly selected citizens and gave opposition parties access to ample campaign resources and media time (^{Merino, 2004}; ^{Becerra et al., 1997}).

Yet, the three parties that led the nation through a relatively non-violent transition, and won the vast majority of seats and votes nationwide (allowing them to control the legislative and the presidency for almost two decades) are now fighting for the future of their organizations, in large part, because voters believe they are corrupt and ineffective. The PRI lost almost 81 per cent of its seats between 2009 and 2018; the PRD saw its seat count reduced by 48 per cent and the PAN by 45 percent.¹⁰ This work does not assign all the blame for the decline of the traditional parties on their non-cooperative behavior in elections and their ability to throw up barriers to citizen participation. Corruption, low economic growth, and large-scale organized criminal activities are obviously important factors in the 2018 electoral crash. The ability of the parties to control representation blunted feedback from voters to party leaders and allowed these problems to fester (^{Langston, 2017}). Negative citizen evaluations of the parties matter only if it is possible to “vote them out”. These barriers (along with many other factors, such as clientelism), made it difficult to punish parties when they did not deliver. Over time, this helped lead to a massive rejection of the nation’s three major parties in 2018.

The creation of Morena under the clear leadership of former PRD leader and two-time losing presidential candidate Andrés Manuel López Obrador allowed the voters to punish the traditional parties in his third attempt to win the presidency. As is clear, the PAN and the PRD suffered serious losses as well as the PRI, meaning the 2018 results were not just a rejection of the incumbent PRI; it was an acknowledgement that the parties were complicit in corruption, bad legislative management, and the growing takeover by criminal organizations of large swaths of the nation’s territory.

Before entering into the account of electoral reforms and malpractice, I present several broad assessments of what is present in the historical record. First, it is important to ask why one sees constant electoral reforming in Mexico. ^{Molinar (1991)} argued that in certain kinds of political contexts, the hegemonic PRI imposed electoral reforms that were more inclined to open the party system or render it more restrictive. Yet, since the onset of rising electoral competition, one sees that party leaders have written, negotiated, and passed in Congress all sorts of self-limiting rules, while, at the same time -and in the same reform- imposing duplicitous procedures on the electoral authority and other political actors, such as activists, ambitious partisans, and voters. This variation demonstrates that it cannot be that a specific political context leads to one type of outcome (self-limiting rules or strengthening the electoral authority) or another (rules that can be manipulated; or cheating during campaigns).

Second, parties in congress, or at least a majority coalition, negotiated these reforms together: that is, after 1986, there was no hegemonic imposition of an entire set of rules. If the PRI wanted support in a major policy program, it would negotiate an electoral reform with the second largest party, the PAN.¹¹ However, in almost all of Mexico’s post 1990 reforms, the PRD was involved, at least during the negotiations. This is an important point because it is not the case that incumbent parties make decisions over whether to cheat or not while the opposition must react to their decision, as some have argued (Chernykh and Svolik, 2015). Third, there has not been a single nation-wide election since perhaps 1982 without at least some complaints made by one party against another, and certainly these complaints have continued since democratization in 2000.¹² At this point, we do not know if this is because the parties cheat in every single election; because they think they can fool the electoral authority that the other parties are not complying with the electoral rules; or because they think they win some sort of benefit from extorting the winner (Hernández, 2019).

Fourth, just as parties claim their rivals cheated in each election, after the 2006 elections, one sees that at least one party (or candidate) demanded an electoral reform after each federal contest. In 2006, after a grueling presidential campaign carried out in television and radio, the losing candidate demanded new rules to control the media and undertake semi-automatic recounts. After complaints in the 2009 mid-term legislative elections, the parties allowed some opening of the party system, which led to an important transformation in 2014. After the 2012 presidential elections, the PAN demanded in the 2014 reforms (as part of the Pact for Mexico for which the PRI needed their support), that the IFE be strengthened to stop gubernatorial interference, among many other issues.¹³ However, it is also clear that party leaders collude to take advantage of these reform moments to act on their common benefits.

The evidence presented below is mixed and includes interviews, legal documents, and secondary literature. I chose the subjects based on lists of former party leaders, government actors, academics, and members of the EMBs. The interviews were semi-structured. In most cases, I asked directly why certain decisions over rules and implementation had been taken.

The secondary literature on the electoral laws in Spanish is quite abundant; although work on the reasons behind the decision making is not. Very few academic studies have studied the effects of the reforms over time. The electoral laws themselves are available online. I am limited by space from listing all the rules which the parties created and later ignored or manipulated; that must be left for another time. Still, in Table 1 many of the reforms are documented and allow us to capture both the variation and the questionable laws that are alluded to in each reform. Finally, instead of listing all the examples of party non-compliance with the formal institutions, I have included the value of the fines that were applied by the EMB on the parties.

The reforms of electoral law in Mexico

Several legal bases for Mexico’s transition to democracy can be traced back to the 1986 electoral reform -the last in which the hegemonic executive imposed its will on other parties- and the 1988 presidential elections.¹⁴ In the electoral reform of 1986, the executive changed the make-up of the electoral authority so it no longer required the coopted, satellite parties then active in the Mexican party system to form part of the majority on the Consejo de la Comisión Federal Electoral. When two PRI party leaders decided to leave the hegemonic PRI, the satellite parties offered the outcasts the use of their electoral registration so that one of these leaders -Cuauhtémoc Cárdenas- could run for president. Under the Frente Democrático, Cárdenas came close to bringing down the PRI’s official candidate, Carlos Salinas; but, because of a computer glitch perpetrated by the Interior Ministry, Salinas won the election with slightly more than 50 per cent of the votes.

After this close call, the hegemonic PRI regime began almost a decade of negotiations over electoral reforms, first with the long-lived opposition PAN, and later including the new unified left Party of the Democratic Revolution, or PRD. The PRI and the opposition parties negotiated electoral reforms in 1989-1990, 1993, 1994, and finally, in 1996, before the PRI lost its majority in the Lower House of Congress in 1997 and the presidency in 2000. Thus, negotiating over electoral rules and elections played a crucial role in the end of authoritarian government in Mexico as each reform allowed for more transparency and fairness in election management, voting lists, voter identification, media openness, and financing, among many other issues.

The crisis of EMB autonomy

Party leaders had both the means and common interests to manipulate or ignore electoral laws, and did so in several ways: first, leaders of the three major parties selected more closely matched agents in the IFE in 2003 and Electoral Tribunal in 2006. Second, after the disastrous 2006 elections, they made it clear they were willing to “fire” the councilors, pressure them to quit, or extend their terms during their period in power, which the PRI government did to the Tribunal magistrates.²³ Third, they wrote electoral laws and regulations that could be easily manipulated in practice. These actions were taken to bypass the campaign spending limits they imposed on themselves. Most parties continued to overspend on elections, while accepting money from other public functionaries and even more illicit sources. Their strategy was to overspend and then pay the fine, which would never again be a true burden as it was due to the elections of 2000.

One should note that the party leaders were willing to accept IFE’s work in a myriad of administrative and logistical tasks that do not affect their ability to spend or erect barriers to entry to the party system. As one can see in Table 1, entire areas of electoral authority activities that were set up during the 1990s continue to be acceptable to the parties, such as the voters’ registry, voters’ identification card, civic education, and managing citizens’ participation in the voting stations on election day. The text in italics refers to rules that restrict entry to the party system or in some way imply duplicitous party behavior.

From Table 1, one can see how parties in congress together with different executives, negotiated important changes to the electoral rules, which they later ignored or manipulated. These changes lend credence to the argument that party leaders cooperate to exclude outside political actors, while they compete and cheat to win votes. One can also see that in the “reversal” section, that in certain instances, new party leaders had to revoke or rework recent reform measures because of their consequences.

Because the Council members placed in 2003 had no citizen councilors who were linked to the leftist PRD, it was regarded as an EMB without the necessary balance among the three major parties, and several of those appointed had little knowledge of electoral matters, or were open supporters of one party or another.²⁴ As a result, it was difficult to argue that the Council could work impartially to interpret and enforce the electoral laws, even if it did in practice. This simple fact would come back to haunt the IFE in 2006, with an extremely close outcome, the PRD could reasonably question the impartiality and fairness of every move made on the part of the Council because their party had not been able to place at least one councilor.

As one former electoral councilor stated, “the parties protected their interests by filling the Council with councilors who acted as “transmission belts” (corredores de transmisión) -that is, councilors who were expected to protect the interests of their specific party in the Council, rather than the institution or election integrity.²⁵ However, as the same councilor pointed out, many councilors did not behave in way -not all protected “their” party’s interests. Still, a delicate balance between the Council and the parties was broken in 2003, and as a result of this and the backlash against the Council in 2006-2007, the parties were able to partially capture the IFE. Through hiring and firing practices, changing the length of tenure, threatening budget cuts, making constant changes to electoral law, the parties sent a clear message: do not sanction us with the full force of the law even if we made the law; and even if we overspend or accept money from illicit sources. With such enormous quantities of both legal and illicit resources flowing through the parties’ coffers, vote buying became more substantial over time (^{Hagene 2015}; ^{Greene and Simpser, 2020}; ^{Lawson, 2009}).

The reforms of 2007-2008 were at least in part an answer to the difficult campaign and post-election period of the 2006 presidential elections. The PRD’s Andrés Manuel López Obrador (AMLO) officially lost the election by a margin of 0.62 per cent of the national vote, but believed he had been robbed of certain victory because the PAN’s candidate, Felipe Calderón, had placed spots asserting that López Obrador was a “danger for Mexico”, because business groups had placed advertisements that criticized the left-wing candidate, and because the IFE refused a full recount. The PRD also complained that the media conglomerates had consistently charged them higher rates for advertising space than they did the PAN or the PRI.

To keep the PRD and its firebrand leader within the bounds of the political system, the parties in congress negotiated a profound restructuring of how parties could accede to mass media during campaigns, particularly radio and television. The reform dictated that parties and candidates could no longer pay for campaign advertisements in federal elections. Instead, the electoral authority would manage the placement of all advertisements, and the television and radio companies were obligated to place them on the air as “public service announcements”, that is, for free. This new plan was also supposed to lower the costs of campaigns, although it is not clear this actually occurred, as Figure 1 below demonstrates.

Source:^{Ugalde and Casar (2018)}.

FIGURE 1 Public financing for parties in Mexico, 1997-2018, in tens of thousands of MX pesos 

It is not clear that the party leaders understood the consequences this new rule would have on their ability to run modern, reactive campaigns using spots in television that do not annoy the average voter, which lends credibility to an explanation based on limited rationality and unforeseen consequences.²⁶ Based on this point, one must recognize while party leaders did work intentionally to close off competition in the party system in many instances, they also had to negotiate as humans with cognitive limitations.

The selection of councilors (as well as the magistrates of the Tribunal) was changed to a staggered calendar to afford better transitions between two CGs. At the same, however, leaders of the PRD demanded the removal of the president of IFE’s General Council because of his perceived bias during the electoral process. As a result, several councilors, including the president of the council, either resigned or lost their posts. This was the clearest indication to that date that the party leaders were willing to break the autonomy of the electoral authority when they saw fit.

The parties also decided to construct a new Accounting Unit (Unidad de Fiscalización) whose chief was decided by the president of the Council and voted in with a majority of the Council. The new Accounting unit took auditing powers away from the direct purview of the Council supposedly so its tasks could be carried out in a more professional and less politicized manner. It was meant to keep the thorny issues of financing from mixing with the other matters that the Council and the parties had to solve. But this autonomy from the Council caused problems in 2012 and the Unit was later returned to the Council in the 2014 reform. As one former council mentioned, once the Auditing Unit was brought back into the Council, INE was again able to place significant fines.²⁷ Again, we see that some decisions were not successful and were later reversed.

From Figure 1, above, it becomes clear how well financed the parties are, both during the campaign season and in those years without campaigns. The national parties received MX$1 1844 000 000 in 2018, which is roughly $90 million dollars divided unequally among the registered parties and coalitions. Though the parties receive such hefty sums of public financing, they still refuse to respect the spending limits, and are willing to paying fines placed by the electoral authority once the election is over and these fines never again approached $1 billion pesos of 2000.

While most analysts see the 2007 as a response to the 2006 elections, the party leaders took advantage of the negotiations to punish smaller parties. On the new ballots, instead of a single option with all the coalition partners included, each party would have its own box and emblem for the voters to select, along with the name of the coalition. Before this reform, it was impossible to know how many votes each party in a coalition won because they were placed together on the ballot. Now, the big three would know how much of their coalition’s overall vote share came from the smaller parties, and how much came from their own supporters, making it easier for the smaller electoral organizations to lose their registration.²⁸

Second, new parties could attempt to win registration only every six years, instead of every three, making it more difficult to gain access to funding and media time. This move clearly had a target: those citizens or political leaders who wished to compete under a different party label against the big three, which reduced entry into the party system. The parties passed another new rule which allowed the parties to claim that any candidate selection process was open and democratic, a rule that made it more difficult to challenge the parties’ selection practices.

To sum up the effects of the 2007-2008 reform: first, many of its most important proposals were a reaction to the perceived failures of the elections of 2006; second, the party leadership took advantage of the reform to place many other, lesser known articles that in fact, reduced the ability of ambitious office seekers, new parties, and activists to participate in the party system. The leaders of the major three parties were willing to negotiate changes to the use of media in campaigns and prohibitions on free speech to appease one of their own. But they also used their congressional majorities to further their shared interests in blocking new entrants and reducing the number of players.

The Federal Electoral Tribunal is far less studied than its administrative counterpart,²⁹ the IFE-INE; yet, its judicial decisions are now the last word in all issues relating to elections. At first, the Tribunal was a court of appeals for some of the decisions made by the IFE-INE regarding elections. However, in a few short years, the magistrates of the Tribunal extended, via judicial action, the scope of their authority to all areas of electoral activity, in effect, overtaking the IFE-INE in its area of action and becoming the judge of last resort (^{Martin, 2007}). With this, the parties realized that the IFE-INE was no longer the most important authority and began appealing the most important decisions to the Tribunal almost automatically. In the early years, the Tribunal worked closely with IFE-INE, but over time, their relations became more antagonistic. The Tribunal began to send almost everything back to the IFE for review, so the parties would semi-automatically repeal the IFE’s findings. Finally, the Tribunal began to revoke the rulings of the IFE-INE, not just return them to be improved, which gave the Tribunal more power to oppose the INE and to strategically assist certain parties over others, especially if a party were holding the executive.

This, however, would eventually cause the Tribunal grave problems as the parties in 2006 chose their allies to become magistrates. The open jostling to place party allies in the Tribunal had two major effects -constant wrangling between the two agencies of the EMB and a greater deterioration of the independence of many of the decisions of the Trife. One of the most questionable decisions of the Tribunal was its authorization of Jaime Rodríguez Calderón’s place on the 2018 presidential ballot at an independent candidate (supported by the PRI to draw off votes from other candidates), despite proof that he had gathered signatures illegally. On the same day, the Tribunal handed down a decision to prohibit Armando Ríos Piter from competing as an independent because he had not complied with these same rules. Pressure exerted by the parties and the executive eventually weaked the agency to such an extent that the president in 2019 was able to oblige the president of the Tribunal to relinquish her leadership post because she did not support the president in a decision. Newspaper attacks against supposed corrupt acts taken by magistrates in the Sala Superior became common; and enormous pressure to support the executive in questionable decisions.

It is interesting to note that the parties and the executive had more success at controlling many decisions of the Tribunal than they did in the INE signaling that the party leaders and the executive were not captured to the same degree. This fact made the work of the INE councilors more difficult, as decisions handed down by the IFE’s Council could be challenged by the Electoral Tribunal, which became beholden to executive interests. Actual impeachment of the leaders of the two organizations turned out to be unnecessary -the pressure that congress (or the executive) brings to bear is normally enough to obligate an unmanageable councilor to step down from her post.

Electoral reforms of 2012 and 2014

Once again, the changes to the electoral law in 2014 began with electoral problems in a previous race. In the mid-term federal elections of 2009, media elites and NGOs complained bitterly about the behavior of parties in power and during elections (Hernández, 2015: 126). The complaints became even stronger when several respected political commentators called for a null vote campaign to demonstrate to Mexico’s party elite that it had to open up to new members and to end its collusive behavior.³⁰ To meet the criticisms of elite political commentators and several nongovernmental groups, the presidency and party leaders at the end of the Calderón administration (2006-2012) negotiated several initiatives that were signed into law in April, 2012 that permitted independent candidacies, referendums, and consecutive reelection.³¹ A further and even more profound restructuring of the electoral authority was carried out in the 2014 electoral reforms and formed part of the Pacto por México under then-president Enrique Peña Nieto of the PRI.

Source:^{Cristalinas (2014, chp 11)}. *In coalitions. Constant 2000 MXP. Note: This figure does not include fines from Friends of Fox (498 million MXP) or Pemexgate (1 billionMXP) from 2000.

FIGURE 2 Fines on parties set by IFE-INE, in thousands of MX constant 2000 pesos 

The 2012 presidential elections brought to light one of the greatest problems of the Mexican electoral system -spending far more than the legal limit. This also created pressure to once again modify how parties are audited. The parties in Congress passed three new laws; a new party law, a new law that regulated electoral crimes; and the law that undergirded the activities of the new electoral authority -now called INE.

As can be seen from this figure, the PAN is by the far the least likely -if not to cheat- then to get caught over-spending, while the PRI and the PRD were the least trustworthy in terms of following spending rules. However, some interview subjects state that the PRD is less able to track its own spending. In the 2012 elections, the PRI took money for its presidential campaign that had been incorrectly reported as spending for the federal legislators (senators and deputies) so Peña Nieto would not be charged with overspending. Furthermore, the candidate of the PRI and his campaign did not report roughly MX$85 million pesos. So, in fact, the IFE did catch many offenses, but not all, or could not prove them all (^{Urrutia, 2013: 8}).

One of the worst examples of illicit spending in 2012 was the PRI’s massive scheme to distribute debit cards to its operators in certain districts in 2012 to be exchanged for votes, called the Caso Monex (^{Aristegui Noticias, 2012}). It was discovered that the parties did not report what they disbursed on the day of the election for their representatives, and such spending would have breached the spending limits ^{(Cristalinos, 2014: 222}). If all parties were obligated to spend under a certain amount, then all would gain the benefits of perceived fairness, at the same time they would not have to spend so much money on campaigning. When asked about this, former party leaders who were interviewed responded that short term interests defeated long term benefits.³² However, it is important to note that the PRI was never found guilty of wrongdoing in this case, under the argument that it was not clear that the money was actually meant to buy votes.

Because of these recurrent problems with campaign spending, the parties once again sat down in 2014 to negotiate a way to both strengthen the INE by creating a new system of campaign accounting and councilor selection (^{Solís Acero, 2018}). A new accounting platform allows campaign spending to be monitored in real time (^{Solís Acero, 2018: 62}). Finally, election results can be thrown out if a campaign is found to have overspent in a close election by more than 5 per cent, among other activities. The problem is that actually revoking the election is ultimately a political decision on the part of the Tribunal.

In this same reform, the parties -led by the PAN- announced that they would centralize much of the power of the state electoral institutes to the national electoral authority. The goal behind this centralizing move was to weaken the PRI governors’ hold over their state electoral authorities. A complete centralization was watered down, but the name, Instituto Federal Electoral, was changed to the Instituto Nacional Electoral (INE) and the Institute was assigned yet more tasks, such as selecting the councilors of the new State Electoral Organisms, now called the OPLES, and using the INE to track spending for all elections, instead of only federal races.³³

The 2014 reforms also opened the party system to new actors. These modifications were extensive and included independent candidacies (which had not been permitted since 1946), consecutive reelection for a variety of posts (prohibited at all levels for all offices since 1933), and new quotas for women candidates. However, at least two of these reforms have been seriously weakened. For example, the new law requires that the candidates be re-nominated by the same party that selected them for their first term in office (unless they leave their party within a specific period before the next election). Independent candidates (those who do not hold party registration) are now allowed on the ballot and given public financing. However, in practice, the parties -especially in the state assemblies- wrote laws that they manipulated in both the national and subnational arenas, such that independent candidates are rarely on the ballot and are not able to compete successfully (Hernández, 2015). Finally, in the 2014 reforms, the parties actually raised the barrier to legislative representation from 2 to 3 per cent of the national vote, making it more difficult for the smaller parties to maintain their registry.³⁴

From the review of the electoral reforms and the implementation of the rules from 1996 onward, we know that party leaders and the executive have incentives to negotiate with other major parties to capture the regulator, to make laws they can later manipulate, and to ignore many of the rules they write to force themselves to comply. Over the past several years, Mexico’s parties followed non-cooperative spending strategies in almost all elections because they knew they would not be fined as excessively as they had been for their 2000 activities. The changes to the 2003 IFE negotiations, the changes in the Tribunal, and the ability of the parties and the executive to place their favorites as regulators all led to this outcome. On the other hand, for topics that do not imply spending, the parties comply with the electoral authorities. Finally, as suggested above, the parties kept the party system closed to new entrants through their control over electoral laws, which the authorities were then forced to apply. Still, it is noteworthy that the parties also reversed some of their decisions that had unintended consequences (under new party leaders). Thus, these are not mutually exclusive causes: parties both intentionally limit, restrain, and weaken, while they also make mistakes and try to remedy them.

Party leaders and the executive have successfully weakened several of the bases of EMB autonomy. They were able to force out the sitting president of the Consejo of the IFE in 2007; they lengthened the tenure of the magistrates on the Tribunal after they had taken office. In 2018, the executive reportedly pressured the president of the Tribunal to relinquish her post (although she remains a magistrate). Together, these manipulations lead one to see how constant pressure and rule changes allowed the parties and executive to at least partially capture their enforcer.

Yet, the lack of representation and fair play had consequences for the parties and their candidates. The terrible cataclysm of the 2018 elections has perhaps altered the party system for good. This system-wide rejection of traditional parties was not only due to the relentless corruption and impunity of the Peña administration or the inability or unwillingness of the opposition parties to stop it. Rather, voters rejected most of the candidates of the three major parties and turned to López Obrador and Morena because of the mountain of political failures that piled up since 2000. Terrifying homicide rates; a lack of sufficient economic growth; and rising prices (together with stagnant wages) also damaged the reputations of the parties as did newspaper reports of their wrongdoing. The shorter-term gains of institutional change were overwhelmed by long-term losses when Morena used corruption debate as a central issue against the three major parties in 2018.

While at least part of the reason for Morena’s wide-ranging victories in municipalities, states, and of course the federal legislature and the presidency was the traditional parties’ lack of compliance with their own rules, the new administration of AMLO seems even less willing to strengthen the electoral authorities than its predecessors, in large part because it wants to continue to win races with huge margins to remain in power. For example, in 2019, the president’s party in congress threatened to reduce the term in office of the president of the CG of the INE from nine to three years, which would be applied retroactively to the sitting president (^{Animal Político, 2019}). If this type of institutional weakening continues, the EMB will be less likely to act against the president’s interests, perhaps leading to worse outcomes.

Conclusions

Both the INE and the Tribunal are well-financed, professionalized bodies that help the nation carry out dozens to hundreds of elections each year. Ballots are printed, voting stations are set up, advertisements are placed, judgements are handed down, and resources are distributed and accounted for, among many other important duties of the two agencies. However, during more than 20 years, Mexico’s major parties and the executives chose a mix of cooperative and non-compliant strategies, even when it became clear that this mixed strategy would eventually harm their own interests. This paper has delved into the issue of short-term gains, and the beliefs of the groups in power that they would be able to deny entry or at least minimize the risks of admitting other actors into the game of electoral politics. The dual nature of the relation among Mexican parties -competing with each other to win elections and control resources while at the same time colluding to rebuff the participation of new actors- created an environment in which it was far more beneficial to capture the regulatory agency and write exclusionary laws to block or weaken representation.

The parties in congress have constantly made constitution and legal changes to the electoral system. When the EMBs carry out the new laws, however, the parties often complain that the authorities are not performing fairly or completely, and that they allow the other parties to overspend or carry out other duplicitous behavior. But it is difficult for the electoral authorities to limit non-compliant behaviors in part because of how parties select the councilors and in part because the councilors know the party leaders can punish them through new laws, budgetary restrictions, and attacks in the media. Another factor that is constant across time is that parties castigated certain groups: the activists and ambitious office seekers within their own parties; regular voters; and women, whom they were very reluctant to nominate to winning districts. The worst result is that even when outsider actors are benefited by new rules, in practice the parties found ways to reduce their participation and influence.

Party leaders miscalculated the willingness of the voters to permit the constant cheating at the same time they misread the growing power of the leader of today’s largest party, Morena. It remains to be seen which among the three parties survives this latest test. However, the party in power continues to wreak havoc on the electoral rules and the authorities that are paid to implement them. Of course, many questions remain, mostly dealing with variation among parties, variation over time, and the different strategies parties would take in the same election or electoral reform to further their interests.