dependence—would effect the likelihood of large-scale emulation or innovation in response to
external vulnerability. To illustrate the plausibility of these hypotheses I used historical examples
from the experiences of five rising or declining great powers over the past three hundred years:
France, Prussia (later Germany), Great Britain, China, Japan, and the United States.
The arguments presented above are not a challenge to neorealist balance-of-power theory
and related works on emulation, innovation, and diffusion that incorporate variables from
balance-of-threat and offense-defense theories. Rather, they supplement systemic realist theories
by specifying how international pressures might combine with intervening variables at the unit-
level to produce foreign and defense policies. The next stage of this research will be devise
measures of the different components of domestic agential power and to subject the hypotheses
advanced above to empirical testing. A central argument of this paper has been that domestic
agential power varies across different states and over time. One avenue for research might
involve testing the neoclassical realist hypotheses using cases drawn from different historical
periods. Another avenue would be to test these hypotheses using cases involving second-tier and
third-tier states in past international systems, as well as cases involving Third World states today.
39
Convention