The theory fundamentally differs from defensive realism by depicting great powers as power-maximizing revisionists privileging buck-passing and self-promotion over balancing strategies in their consistent aim to dominate the international system.
Offensive realism is a prominent and important theory of international relations belonging to the realist school of thought, which includes various sub-trends characterised by the different perspectives of representative scholars such as Robert Gilpin, Eric J. Labs, Dylan Motin, Sebastian Rosato, Randall Schweller and Fareed Zakaria.
Ultimately, Mearsheimer's offensive neorealism draws a much more pessimistic picture of international politics characterised by dangerous inter-state security competition likely leading to conflict and war.
[30][31] Since global hegemony is nearly impossible to attain due to the constraints of power projection across oceans and retaliation forces, the best end game status states can hope to reach is that of a regional hegemon dominating its own geographical area.
In a defensive neorealist mindset, security increments by power accumulation end up experiencing diminishing marginal returns where costs eventually outweigh benefits.
[35][36][37] Mearsheimer challenges these claims by making the argument that it is rather difficult to estimate when states have reached a satisfactory amount of power short of hegemony and costly to rely extensively on balancing as an efficient power-checking method due to collective action issues.
Combined, these two variables allow him to establish that great powers tend to favor—to the contrary of defensive neorealism predictions—buck-passing over balancing in all instances of multipolarity except for those that include a potential hegemon.
[43] This lack of balancing is best explained by the regional hegemon's newly acquired status quo stance, which follows from the geographical constraints on its power projection capability.
While the inputs and critics below provide a good sample of the theory's contributions and the kind of arguments that have been addressed against it, the listing should in no case be considered as exhaustive.
[50][51] Additionally, the inclusion of new variables such as geography alongside the distribution of power enhances offensive neorealism's potential to make specific assumptions about states' pursuit of aggressive actions and resort to balancing and buck-passing.
According to him, the inclusion of the non-structural geography variable to explain great power behaviour shifts the theory's focal point of analysis from system-wide dynamics to regional ones.
In this sense, Layne questions the ability of the water constraint to transform a power-maximizing state into a status quo power and contradicts Mearsheimer by arguing that a regional hegemon remains subjected to the quest for security, thereby striving to attain global hegemony.
[59][60] Moreover, Snyder argues that no consideration is given to transnational threats such as terrorism, and that Mearsheimer's emphasis on security makes him ignore states' non-security interests such as ideology, national unification and human rights as an essential aspect of international politics alongside power competition.
[61] Additionally, Toft points out that Mearsheimer's concentration on military capabilities and issuing state capacity for territorial conquest "implies a risk that his analyses miss a host of other ways of gaining and exercising influence".
[62] Similarly, political scientists whose primary focus is bargaining models of international conflict note that offensive neorealism ignores the fact that war is costly.
Instead of being a regional hegemon with the strategic aim of dominating the Western hemisphere while preventing the rise of peer competitors in Europe and Northeast Asia, these scholars believe that empirical data points to the fact that the United States has sought and achieved global hegemony, which in turn biases Mearsheimer's predictions regarding future U.S. strategic behavior, mainly in terms of its military involvement overseas.