1. Introduction
On July 2, 2006, Mexican voters elected National Action Party (PAN) candidate Felipe Calderón as the next president of Mexico. Calderón's victory was extremely narrow; he won under 36% of the total vote and less that 0.6% more than his leftist rival, Andrés Manuel López Obrador. This potentially problematic situation was aggravated by López Obrador's decision to challenge Calderón's victory, both in the courts and in the streets. López Obrador's protest campaign culminated on September 16, when tens of thousands of his followers gathered in downtown Mexico City to acclaim him “legitimate president” of Mexico. Meanwhile, in the legislature, leaders of López Obrador's Party of the Democratic Revolution (PRD) oscillated between hints that they would collaborate with Calderón's administration and signs that they would adopt a posture of untrammeled hostility.
Post-electoral controversies raised the specter of a Left that had abandoned parliamentary tactics and returned to mass mobilization as its principal political strategy. At worst, they presaged the sort of political upheaval that could threaten Mexico's young democratic institutions. There was certainly no missing the symbolism involved in López Obrador's choice of the date on which he would take the oath of office—November 20, the anniversary of the Mexican Revolution—nor could students of Mexican history fail to recall that a contested election had sparked that decade-long conflagration.
How did Mexico find itself in the middle of such a crisis? Why had the country's vaunted electoral regime, generally regarded as a model for other democracies, failed to produce an outcome that all parties considered legitimate? Were Mexican political institutions so shaky that the actions of a single man could cause their collapse? And, given the answers to these questions, what does the future hold for Mexico's political system?
Over the last decade, research on Mexican politics has focused on (1) institutional reform, especially in the electoral sphere, and (2) mass behavior, especially voting. Both areas of research are, of course, essential to understanding Mexico's transition from a one-party dominant regime to a multiparty democracy. However, scholarly attention to them has tended to minimize the importance of political leadership and informal arrangements among elites (as did O'Donnell and Schmitter in 1986).
This article argues that the way these elites interact plays a pivotal role in the current political situation. It first summarizes Mexico's transition to democracy over the last 15 years. It then addresses the simmering tensions between the PRD and the PAN during the administration of Vicente Fox that boiled over in the 2006 elections. The third section suggests that Mexico's current climate of polarization is a function of elite attitudes and interactions, rather than those of the mass public. The fourth section shows how these same interactions exercise a far greater influence on Mexican politics than do the institutions most often implicated in poor governance. The route out of Mexico's political impasse thus runs through pacting and compromise among members of Mexico's current political class, rather than further institutional tinkering.
2. Incomplete Transition
For close to seven decades, a single party (known today as the Institutional Revolutionary Party, or PRI) won all elections for significant posts. Over time, however, modernization weakened the corporatist and clientelist apparatuses through which the “official” party and the state had ensured social control. The collapse of Mexico's economy in the 1980s further undermined autocratic institutions and provoked mass disaffection with the old regime.
In the face of mounting social unrest, representatives of Mexico's political establishment negotiated a series of reforms with the leaders of the main opposition parties during the 1990s. These elite pacts, most notably the 1996 “Reform of the State,” ultimately leveled the electoral playing field. In 1997, the PRI lost control of the Chamber of Deputies, and in 2000 PAN candidate Vicente Fox captured the presidency.
Data from standard measures of democracy nicely capture both the scope and the limitations of Mexico's political transition during the 1990s. In 1991, Mexico scored a zero on the combined Polity IV index; by 2001, it scored an eight.1
The combined Polity IV score ranges from −10 (utter autocracy) to 10 (full democracy). A score of zero indicates that autocratic features of the regime evenly balance.
Freedom House scores range from 2–14; higher scores indicate less freedom.
A closer view of Mexico's political transition reveals the unevenness of democratization across different institutions and spheres of governance. For instance, Mexico's electoral regime and party system were quite well developed (see Todd Eisenstadt's essay in this symposium). As in the old regime, the military and the security services remained small and firmly under civilian control. Finally, despite the domination of broadcast television by two relentlessly commercial networks, Mexico's mass media had become quite open by the time of Fox's election in 2000 (Lawson 2003). By contrast, progress toward reforming the police, the judiciary, the prosecutorial apparatus, and other parts of the bureaucracy remained painfully slow. The PRI remained the country's largest party; businessmen with longstanding ties to conservative factions of the regime continued to monopolize most sectors of the economy; and corrupt bosses affiliated with the PRI controlled most labor unions.
As president, Fox presided over modest democratic deepening. Civic groups and opposition parties successfully challenged the PRI's remaining strongholds at the state and local level; independent newspapers sprouted throughout the country; state-level electoral authorities became more independent and professional; prominent PRI figures and organizations began to defect to the opposition; and the passage of a federal Transparency Law exposed government operations to public scrutiny. Although corruption remained a serious problem, especially in the criminal justice system, the administration itself managed to avoid major scandals.
For most Mexicans, the PRI's defeat in 2000 represented the culmination of a long process of democratic transition and the beginning of an equally arduous process of democratic deepening. For the Left, however, alternation in power between the old ruling party and the conservative PAN constituted only partial or cosmetic change. Three episodes during Fox's tenure seemed to confirm their fears.
In 2003, the PAN and the PRI joined forces to name the new leaders of the Federal Electoral Institute (IFE) over the objections of the PRD (see Eisenstadt's article). In contrast to the previous set of “Citizen Councilors,” who included a number of distinguished academics and activists, the new cohort included a number of political unknowns and party hacks. Their selection signaled the breakdown of the partisan consensus that had characterized the political accords of 1996–1997.
Two years later, PAN and PRI legislators voted to impeach López Obrador, then mayor of Mexico City, on the grounds that he had violated a court injunction in a zoning dispute. Had the legal proceedings continued, they would have prevented López Obrador from seeking the presidency. The Fox administration backed down in the face of widespread public opposition, international pressure, and massive demonstrations in Mexico City organized by López Obrador. Most Mexicans saw the whole affair as an attempt to trump up charges against a popular rival.
A third insult came in the midst of the 2006 presidential race, with the passage of the new broadcasting law (the “Ley Televisa”) that was notoriously favorable to Mexico's two main television networks. During the second half of the race, television coverage of Calderón became more favorable, while reporting on López Obrador turned rather sour. In the leftist narrative, all of these events signaled a conspiracy between the government, the PAN, the old ruling party, leading businessmen, and a now-perverted electoral authority to deprive their candidate of victory.
Panistas (PAN partisans), of course, saw matters in an entirely different light. Their party stood for the same Christian Democratic principles that it had represented steadfastly since the late 1930s; by contrast, the PRD represented both the radicalism of the Marxist left and the corruption of the old PRI (from which many of the PRD's founders had come). It was the PRD that had rejected Fox's offer to form something like a government of national unity in 2000, and it was PRD obstructionism that had prevented partisan consensus in the selection of a new set of IFE Councilors.
Despite López Obrador's moderate position on many policy issues, his administration as mayor of Mexico City struck opponents as eerily reminiscent of PRI rule. For instance, López Obrador incorporated whole hog into the PRD apparatus almost two dozen PRI organizations, several with decidedly unsavory reputations. Episodes of corruption among his top aides, some captured on videotape, raised serious questions about financial probity, as did López Obrador's refusal to endorse a local transparency law modeled after the federal statute.
For his opponents, post-electoral controversies only confirmed their instincts: López Obrador was simply unwilling to accept the results of an election that the IFE, the Electoral Court, and most international observers considered free and fair (see Eisenstadt's article). His ad-libbed responses to critics in speeches after the elections—e.g., “to hell with your institutions”—betrayed a casual attitude toward the rule of law that would have imperiled democracy had he won. This perspective contrasted starkly with Calderón's pledges to respect the autonomy of regulatory institutions, insulate the office of the public prosecutor from direct control by the executive, reform the judicial system, and further devolve authority to state and local governments—precisely the steps Mexico needed to overcome its autocratic legacy.
Since 2003, the strategies adopted by Mexican political elites in their partisan disputes have proven more tendentious and incendiary than analysts predicted. For instance, few political observers in 2001 would have anticipated PRI and PAN attempts to prevent López Obrador from contesting the 2006 elections through an act of legal legerdemain. Even fewer would have guessed how far López Obrador was willing to escalate his tactics after July 2, 2006.
3. Elites or Masses?
Do trends at the elite level reflect increasing polarization among ordinary Mexicans? Over the last five decades, support for the PRI in the mass public has declined at a rate of about 3% per election cycle. The weakness of the PRI's presidential candidate in 2006, Roberto Madrazo, only accelerated this process by hastening defections from the old ruling party (see Langston's contribution to this symposium). Because the PRI was ideologically and socially amorphous, its unraveling should theoretically have divided Mexico along lines of class and ideology. Left-Right differences should also become more salient as the issue of democratization faded from the agenda, forcing people to choose between very different political alternatives rather than simply selecting the one that was most likely to defeat the regime.
Nevertheless, the way in which voters have attached themselves to the PRD and the PAN does not seem to follow such a clear logic. Most Mexican voters do not base their electoral choices on the policy positions adopted by parties and candidates (see Moreno's and Bruhn and Greene's articles in this collection). Still less do Mexicans vote along class lines (see Moreno). Although indicators of social status—such as living standards, education, skin color, and occupation—influence voting behavior at the margin, for ordinary Mexicans, region is a far more important predictor.3
Results are based on simulations from a multinomial logit model of vote choice, in which the dependent variable took on one of four values (Calderón, López Obrador, Madrazo, or none/undecided). Independent variables included: age, gender, living standards (as measured by an index of material possessions, education, church attendance, region, politicalengagement, skin color, and urban or rural residence). Data are taken from the Mexico 2006 Panel Study, Wave 2. For full results, see Lawson 2006, available at: http://web.mit.edu/polisci/research/mexico06/Pres.htm.
The episodes so central to polarization at the elite level have played out very differently in the electorate. For instance, polling data indicate that there is little support for continued protests by López Obrador; even many of those who voted for him express ambivalence about his tactics.4
Mexico 2006 Panel Study, Wave 3 and accompanying cross-section; Consulta Mitovsky, National Household Survey, August 2006.
Parametría, “El desafuero de López Obrador,” National Household Survey, August 2004; Consulta Mitovsky, Household Survey in the Federal District, September 2004; Consulta Mitovsky, National Household Survey, January 2005; Consulta Mitovsky, National Telephone Survey and National Household Survey, April 2005; Consulta Mitovsky, Household Survey in the Federal District, February 2005.
Acknowledging the truth of this argument, however, tells us little about why elite conflict has become so pronounced. The principal answer to this question lies in the patterns of party-building in Mexico. During the period of one-party rule, the PRI's eclectic nature gave rise to a fragmented opposition. Because opposition politics promised few tangible rewards, it tended to draw more extreme or ideologically purist members of society, on both the Right and the Left (Greene, forthcoming). Today, PAN and PRD activists come from strikingly different backgrounds. PAN candidates to Congress in 2006 were generally introduced to politics through their ties to the private sector and the Church; PRD candidates came up through labor unions and popular social movements (see the Mexico 2006 Candidate and Party Leader Survey, described in Bruhn and Greene's contribution to this collection). Although many leaders in both parties have attended public school, PAN politicians are far more likely to have attended private or parochial institutions. As a result of these patterns of political recruitment, party elites share relatively few cultural reference points.
Despite steps toward internal democracy in both parties, “old guard” elements still exercise substantial influence. The PAN remains a “club” party, in the sense that membership is not automatically open to anyone. Rather, those who wish to join must first be accepted as junior members (miembros adherentes); after a minimum trial period of six months and participation in various party activities, they may then apply to become full, dues-paying members (miembros activos). Both types of members could vote in the 2006 presidential primary, but only full members can vote for party leadership positions or candidates for other offices. Not surprisingly, the party's current leadership remains far more conservative than party voters, not to mention ordinary citizens. In the case of the PRD, presidents have exercised rather wide discretion in whom they appoint to the National Executive Committee. Many of the current members were placed there by López Obrador. This fact may help to explain why, despite the steady influx of pragmatic PRI defectors into the PRD, its leadership supported López Obrador's post-electoral protest movement. These party elites are, in turn, the principal source of political polarization in Mexico.
4. Elites or Institutions?
For those political scientists who emphasize the role of institutions, the roots of Mexico's current political predicament lie in a cluster of familiar constitutional rules. First, Mexico's electoral system contains a large component of proportional representation, which in turn encourages multipartism. A multiparty system is not inherently problematic, but it becomes so when paired with a second institution: presidentialism. The combination of these two institutions virtually guarantees divided government. The adverse effects of these arrangements are compounded by the lack of run-off elections for president, which permit the election of non-Condorcet winners, and by the length of the presidential term (six years). Finally, the prohibition on consecutive reelection renders politicians less accountable.
Dysfunctional institutions, however, cannot account for the most salient features of Mexico's current political topography. Most obviously, they cannot explain why one of the best-designed electoral systems in the world failed to produce a result that party leaders on the losing side would accept. If institutions are the main issue, governance problems should be the product of gridlock, rather than political polarization. Today, the reverse is true: as a result of likely collaboration between the PAN and elements of the PRI, legislative gridlock is now relatively unlikely; on the other hand, alleged electoral irregularities have provoked a crisis.
Choice and leadership have more to do with today's situation than do formal institutions. In the case of post-electoral protests, for instance, other men in the same situation would have made different decisions than did López Obrador. In 1988, PRD candidate Cuauhtémoc Cárdenas proved less vigorous in protesting the official results of the election than López Obrador is today, even though the Left had much stronger grounds to do so than they did in 2006. Likewise, had Calderón lost the election, there is no doubt that he would have accepted the result or challenged it through strictly constitutional channels. The problem, then, lies less in how Mexico's president was elected than in how elites reacted to his election.
The simple fact that institutions have played a role in permitting the overrepresentation of extremists at the top of Mexico's main parties does not consign Mexico to crisis. Even fairly doctrinaire politicians can compromise, as the 1996–1997 inter-party negotiations showed. Nothing in the current context compels the PRD's leadership in Congress to adopt a relentlessly obstructionist stance, and the electoral benefits of doing so are at best unclear. Nor do present circumstances prevent Mexico's president-elect from reaching out to the Left.
There is not necessarily anything wrong with further institutional reform in Mexico, of course. But such reform is important as a symptom of agreement among the main political parties, not as its cause. In the end, it is the way particular leaders interact that will propel events toward compromise, or toward crisis.
