States are rational, and regime type matters. That is, countries will tend to do what is in their best interest. However, what is in the best interest of a country depends on what sort of country it is. Thus, Russia’s best interests are probably very different from those of India, China, Europe, or the United States, Russia desires to maximize oil prices means it needs to redistribute wealth away from the producers of wealth to the extent possible. Growth-oriented countries, by contrast, need to create wealth to the extent possible.
Still, none of this happens in a vacuum. Countries learn what is the best way to achieve their goals. In particular, is modeling (which is analogically adopted from the research of Albert Bandura) or “conquest-and-reward” (which is analogically adopted from the research of B.F. Skinner) the best method for removing bad behavior from the repetior of states?
First, some background on the approaches:
Conquest-and-reward is an operationalization of the Law of Efffect, which was first described by E.L. Thorndoke in 1898 and refined over the next decades to the following
Responses to stimuli that produce pleasing effects are more likely to occur again
Responses can be “positive” (some new good thing is added) or “negative” (some old bad thing is taken away). Punishment doesn’t work predictably, as it interacts with volition instead of motivation. Instead, to prevent a bad behavior from reoccuring, you would subject the organism to an upleasant state of affairs, and lift the bad experience when the organism complies. “Conquest-and-reward” (C&R) is a basic feature of all learning systems, whether mechanical, animal, or human. C&R, by itself, does not explain all variation in learning behavior. However, it is a great start.
Modeling is a more complex form of learning that relies on an orgasm using the behavior and consequences of other observable organisms as a form of “offloaded experience.” In modeling, an organism observes a behavior, acquires the gist of it, and reproduces the behavior (with variation) in similar situations. So, in the classic Bobo Doll study, for example, children watched a model abuse a doll in specific ways, and proceeded to abuse the doll in novel ways.
With states, my suspicion is that C&R is more important than modeling. Or rather, they operate in different ways. C&R, through condition, impacts what a country actually wants. America’s learned response from Vietnam was to avoid counterinsurgencies, for instance. Modeling, in contrast, provides methods for a country to get there. The successful integration of much of Old Core Europe into the European Union, for instance, modeled behavior that would allow the central European states to get what they want (rich and free of Russian domination). Thus, if a country wishes a bad outcome, such as the conquest of a neighbor, modeling provides a mechanism only of demonstrating how that could be done successfully, while C&R can reverse the bad behavior.