Mexico has undergone major political changes during the past four decades. From being a stable hegemonic party system in the 1960s it became a multiparty system in the 1990s with three main parties that together took 90% of the vote, though none managed to cross the 50% threshold. New institutions were created to promote free and fairer political competitions and greater protection of civil liberties. Alternations of executive power in important states and counties became more frequent. Whereas four decades ago no one doubted that the Party of Institutional Revolution, the PRI, would win presidential, congressional, and virtually all other elections, in 1997 not only did the PRI lose control of the Chamber of Deputies, but it also lost Mexico City's rst mayoralty election. The July 2000 presidential election was the climax of these changes. The PRI's candidate, Francisco Labastida, was defeated by the National Action Party's Vicente Fox, 43% to 37%, with the Party of Democratic Revolution's candidate, Cuauhtemoc Cardenas, receiving most remaining votes. Moreover, Fox's Alliance for Change, of which the National Action Party (PAN) is the major component, won a plurality in the Chamber of Deputies and the PAN added governships in the states of Morelos and Guanajuato to the half dozen that it already controlled.Political scientists have sought to explain Mexico's political changes from a variety of theoretical standpoints. Many of their explanations have focused on the structure of Mexican elites and how it has been changing. To capture the elite structure, a handful of models have been employed without any agreement about which model is most accurate. It is probably correct to say that political scientists agree only that Mexican elites have long constituted an exceptional case, one that has been dif cult to t into any existing model. I intend to review and critique this modeling effort, paying particular attention to how the interpenetration of political, bureaucratic, and business elites has been depicted by various scholars. I begin with a brief exposition of the several models that have been used, after which I assess their applications as the elite structure changed from the hegemonic formation of the 1950s and 1960s to today's signi cantly more differentiated con guration.
Competing Models of Mexican ElitesWithin elite studies there is no consensus about how to model elites. At least ve models have been proposed and each has been used in studying Mexican