Citation

Goldsmith BE, He B. J. Peace Res. 2008; 45(5): 587-611.

Copyright

(Copyright © 2008, SAGE Publishing)

DOI

10.1177/0022343308094322

PMID

unavailable

Abstract

Does the type of regime really make no difference to the likelihood of violent conflict over basic issues of stateness such as separatism and decolonization? Can democratic peace theory be successfully applied when dealing with the national identity or stateness question? This article extends the application of the democratic peace to the process of decolonization. It examines conflict between imperial states and their colonies during the process of decolonization and investigates the question of whether democracy affects the likelihood of conflict. The central finding is that, contrary to the implications of some prominent theories of state formation and democracy, democratic imperial states are significantly less likely to go to war with their colonial possessions in the process of achieving independence. Further, the authors find only a monadic, not dyadic, democratic peace effect. The regime type of the colony does not have a significant effect on the likelihood of war. It is the nature of the regime of imperial states, rather than that of colonies, that is a significant factor. In addition, the predominant source of this effect appears to be the institutional constraints placed on executive action within democracies, rather than the influence of mass politics or the effects of political competition. Regarding power-related factors, power parity between sovereign and colony makes conflict more likely (a colonial power-transition effect), but imperial decline actually makes war with colonies less likely. Sensitivity analysis reveals that a number of other hypothesized effects cannot find robust support. Simulations are used to assess the magnitude of the effect of regime type pre- and post-independence. Overall, the article contributes to theory development by investigating different institutional aspects of democracy and by distinguishing monadic and dyadic effects.