I helped to teach a course on international institutions last term. The course included an overview of the debates from the 1980s and early 1990s over multilateral agreements and the possibility of mutually beneficial cooperation between states. According to most IR textbooks, these debates established an enduring point of contention between Realist and Liberal Institutionalist perspectives – the latter focusing on absolute gains (welfare improvements) and the former on relative gains (share of power resources). I am not sure this is a very accurate summary of current debates in the discipline, I don’t think that there has been much research on relative and absolute gains in quite some time (this and this were the only articles I could find from the last decade). If my grasp on the recent history of the discipline is correct, that this is likely because the initial research programme on relative gains initiated by Mastanduno and Grieco fizzled out pretty fast when the empirical evidence proved to be mixed and inconclusive. More hardcore rational choice theorists demonstrated that the claims made about relative gains being a barrier to cooperation did not really stand up to scrutiny. Liberal Institutionalism won the argument to a large extent, but then mutated into a pretty generic rational choice institutionalism concerned with the same sort of problems of credible commitment and collective action as much mainstream economics and political science – losing touch with discipline-defining debates. So relative vs. absolute gains provides a neat way for textbook writers to distinguish between two ‘paradigms’, but it’s not a division that provides much of a guide to current research in journals like International Organization.
But the division does appear in textbooks, so it’s part of the education of many thousands of students every year. Unfortunately, the argument that if states are concerned with relative gains (share of power resources) they will forgo the benefits of cooperation is bunk and can be shown to be so very easily. In writing this up I discovered that this argument may have been made by Duncan Snidal 25 years ago, no doubt more rigorously. But the fact that its not hard for people like me who lack Prof. Snidal’s expertise in game theory to show flaws in the relative gains argument demonstrates how flimsy the argument is, even granting all the assumptions made by its Hobbist proponents.
Let’s adopt the same soft rational choice approach of many 80s and 90s IR theorists. I’m not convinced by this approach, with its contextless hypothetical examples and stereotyped, oversimplified scenarios, but it’s the ground on which proponents of the relative gains argument chose to pitch.
Suppose states A, B, C and D are states, each with a GDP of 10 trillion dollars. Let’s use GDP as a measure of material capabilities, as many neo-realists do (and with good reason). Each state therefore controls 25% of the material capabilities in the system. If (offensive) neo-realism is correct, each desires to increase their share. They are expected to forgo economic gains if it would also involve larger gains for another state, as this would lead to a reduction in their relative power. Logically, if they are rational as defined within rational choice theory, they would even accept a lower absolute income in order to have a greater share of power resources. It seems, therefore, that asymmetric trade deals that benefit some of our actors more than others would never be struck. This is not the case, however.
Let’s say A offers B a take-it-or-leave-it trade deal that will benefit A by $2trn and B by $1trn. This example follows the discussion on page 105 of Theory of International Politics where Waltz explicitly discusses an imaginary deal skewed 2:1 in favour of one state. If B is concerned with relative gains, should they take this seemingly unfair deal? Contra neo-realism, yes: it increases B’s share of power resources in the system by 0.6% and puts it ahead of C and D in terms of their power ranking. So concern with relative gains will lead towards cooperation even though the terms are unequal. Let’s say that B can offer a similar deal to C and C to D. Assuming that each deal really is a take-it-or-leave it offer in which no better alternative is available, each offer will be accepted because each time it will increase the share of power of both states involved – providing relative gains.
Finally, after this chain of deals, if D can offer A a similar asymmetric deal, will A take it? Unless A is very concerned with a very specific sort of relative gain (i.e. gains or losses relative to the most powerful in the system), then it will. The deals that A and D were not involved in reduced their share of material power, and making a deal – even if it benefits D more – will increase A’s share of power resources.
At the end of the chain of deals, each actor will have $13trn GDP and so the same 25% share of power resources that they started with. Even if they cared only about relative gains, each state has ended up better off and participating in mutually advantageous cooperation.
So under this scenario, relative gains are not a barrier to cooperation. It should be easy to see that under more realistic conditions, the same will apply. A state is unlikely to want to be left out of a multilateral trade deal that provides benefits to others, even if the deal is skewed. In his book Ruling the World Lloyd Gruber argues that something similar to this process created the clamour to join the WTO in the 1990s – even though many states in the global South preferred its forerunner the GATT. It’s only when the number of states is very small and the distribution of benefits very skewed that we should expect relative gains to matter and to prevent cooperation.
Alternatively, if we abandon the idea that all states are threatening to each other and acknowledge that states are primarily focused on certain specific rivals then relative gains considerations might be relevant. There’s real world evidence for this: witness concern over French arms exports to Russia in the aftermath of its annexation of the Crimea. As pointed out by my students, one of the most compelling examples of relative gains presenting a barrier to cooperation comes from international environmental negotiations, where the US seems to have been concerned to avoid any agreement that would disadvantage it relative to China.
But in the scenario that neo-realism posits, where all states are rivals, somewhat counter-intuitively relative gains considerations will no longer be relevant as a loss of relative power compared to the other contracting state may be compensated by the relative gain compared to others.
Starting with broadly Hobbist assumptions, there are other reasons why states might avoid entering into positive-sum multilateral agreements with each other. Waltz notes that states may be concerned with the vulnerability that dependence on a potential enemy or rival might bring. States might guard their sovereignty jealously and be wary of any loss of de jure or de facto independence. But whilst the issue of relative gains might prevent cooperation in some special cases, it isn’t a general barrier to cooperation under conditions of anarchy – even if the system is characterised by competition and aggression. Glancing at the historical record, this should have been fairly obvious: transnational capitalism emerged coterminously with the turbulent European state-system.