Iwish in this essay to peer through the Latin American looking glass, or abertura, to see what light may be shed on the ongoing struggle to democratize the South Korean political system. In Latin America the richest literature on the problems and prospects of democratization emerged along with a more successful decompression of authoritarian regimes (with exceptions, like Chile) than has yet occurred in Korea or Taiwan. So, it is a literature that may hold comparative insights for East Asia. Moreover, the Latin American field has been a model of world theory merging with home-grown theory and practice, each influencing the other. This is a relief to someone from a field whose leading lights describe the economic prowess of the people they study as miraculous (that is, unbelievable) and who find episodic democratization equally odd, and worthy of yet more self-revelatory genuflection. I will first assess the rise and apparent demise of the Latin American dictatorships, then question the diagnosis and cure suggested by several prominent Latin Americanists, and finally proffer an alternative explanation that I think illuminates both the Korean and Latin American cases.
It is worth recalling that the most powerful explanation of the serial emergence of authoritarian systems in Latin America, Guillermo O’Donnell’s analysis of bureaucratic-authoritarianism, posited an interacting pattern of world system, world time, industrialization strategies, and domestic political coalitions and regime types. The bureaucratic-authoritarian system arose, so O’Donnell argues, in sequence with the exhaustion of import-substitution industrialization (isi), vertical ‘deepening’ of industrialization (in intermediate and capital goods), and the onset of exportled development; thus ended the populist and nationalist phase identified with the Perón and Vargas regimes in Argentina and Brazil. Industrialization strategies prompted by what Albert Hirschman called ‘late-late’ development had the political consequence of authoritarian rule, rather than the democratization that previous analysts in the modernization school had predicted. footnote1
The bureaucratic-authoritarian regime combined the following characteristics: a social base in ‘a highly oligopolized and transnationalized bourgeoisie’; decisive weight given to a transnational stratum of experts in metropolitan economic orthodoxy and specialists in coercion; absence through liquidation of the institutions of democracy and various channels for the representation of heterogeneous interests; exclusion of the popular sector, especially the working class; economic exclusion of small and medium business; transnationalization of the economy; denationalization of society.
South Korean authoritarianism (or what I have called its bureaucratic-authoritarian industrializing regime) shared many similarities with the recent Latin American experience. In my view the period of military rule in 1961–63 coincided with a state-led transition from isi to export-led development, and the elaboration of a formal authoritarian system (the ‘Yushin’ regime, 1972–79, and 1980–87 in Chun Doo Hwan’s guise) coincided with a profound industrial deepening. If O’Donnell’s model does not work in Latin America save Argentina, as some critics say, it seems to do so in Korea.
The international or security dimension made a difference in the Korean case, as Park Chung Hee inaugurated heavy industrialization (steel, petrochemicals, automobiles, shipbuilding) and developed his authoritarian system in the context of the sharp changes occasioned by Nixon’s opening to China and his ‘New Economic Policy’ toward Japan, both unveiled in August 1971. Nixon’s combination of incipient us neo-mercantilism, corresponding imperial inattention to industrial deepening, indulgence for authoritarianism (both the Philippines and Korea ‘went authoritarian’ within weeks of each other in 1972), ‘Nixon doctrine’ reliance on regional gendarmes, and shrewd pursuit of detente with Russia and China provided both the goad and the space for Park’s actions. I would not want to guess whether the external or internal