Regimes and the limits of realism: regimes as autonomous variables

Two distinct traditions have developed from structural realist perspectives. The first, the billiard ball version, focuses purely on interaction among states. The second, the tectonic plates version, focuses on the relationship between the distribution of power and various international environments. It is the latter tradition that suggests why regimes may be important for a realist orientation. However, it also opens the possibility for viewing regimes as autonomous, not just as intervening, variables. There may be lags between changes in basic causal variables and regime change. There may be feedback from regimes to basic causal variables. Both lags and feedback suggest an importance for regimes that would be rejected by conventional structural arguments.