'Absolute Gains or Relative Gains' Kind of Misses the Point

Sorry, why are we doing this? (Image: UNSW eLearning)

Students in my department spend a lot of time comparing/contrasting the two biggest “-isms” (so-called realism and liberalism) and the debate about whether states pursue “absolute gains” or “relative gains”. The “liberal” camp argues that states pursue “absolute gains” that benefit them overall, or words to that effect. The “realist” camp counter-argues that states do and should pursue only “relative gains” that privilege them more so than other potential rivals. There is often a discussion here about a prisoner’s dilemma, or a repeated prisoner’s dilemma that comes in tow. No matter, I remember this debate from when I was an undergraduate. I was reading the same stuff they are, and that was over 20 years ago.

I was too naive to notice this when I was an undergraduate, but I see now that this debate misses the point entirely. While it looks like a “debate” that was apparently important enough to fell a lot of trees for paper to print the terms of it, this debate is rather silly and out of focus. My inspirations for this scorcher of a 🔥 #take come from R. Harrison Wagner and Robert Powell’s different takes on this. I’ll (try to) elaborate.

For one, Wagner (2007, p. 41-2) will repeat some of the points made by Powell (to follow), though fn. 60 emphasizes one major reason why I think this debate misses the point. If you come from a bargaining perspective, all gains are relative because all goods are divisible. No good that is important enough to potentially be militarized could be anything but divisible, and the division of it produces gains that can only be relative. That is: goods are zero-sum. In the classic case that I like to teach, two states are trying to proverbially divide a dollar and maximize their individual payouts in the process. The proverbial dollar could be the openness of State B to State A’s goods. It could be the division of territory. No matter what, there is no division possible of the good in question in which both sides are better off. One side getting a better deal means another side is getting a worse deal. Zero-sum goods can only produce “relative gains” in that way.

Thus, I think so-called liberal argumentation in the abstract about the downstream benefits of cooperation misses an important point, or at least loses sight of it in the confines of the debate they elect to have. A payout downstream from an agreement is a dividend, not an absolute gain in the terms in which it’s often discussed. Consider the 1993 agreement between China and India regarding their contentious border. The agreement reaffirmed the status quo, one largely imposed on India by China following the failed “push” in 1962. However, the confirmation of the status quo comes in relation to all of what India covets. Accepting the status quo in terms of what was conceivably on the table is a relative loss for India. China got more (i.e. it retains what it has). India got less than it wants (which is a loss). Just because there’s been some kind of goodwill that’s followed from that doesn’t change the fact the division of a good is ultimately relative. “Absolute gains or relative gains” obscures more than it reveals.

Powell’s 1991 article and 1994 article are must-reads on this topic as I think he does a great job making clear how much is unclear or poorly stated in what’s superficially a really obvious debate. The 1991 article offers a simple framework for thinking about the problem while noting the terms of the debate are wholly divorced from the methods being used. The debate occurs using similar methods as economists might use studying the behavior of firms under conditions of perfect competition or monopoly. However, neither strategic context changes the incentive that economists impute for a firm. It’s always about maximizing profit. Thus, a debate about apparent behavior that is observed is not an empirical debate about what states “do”. It’s actually a debate about what assumptions we elect to make about what they do. We’re altering their preferences and not their behavior, per se, and debating the relative merits of these competing assumptions.

Ignore for the moment that Powell (1991, 1310) contends that any use of the prisoner’s dilemma in which an actor can be eliminated is no longer a prisoner’s dilemma. That misapplication of a simple model and misinterpretation of its parameters are important details, but we’re always going to talk past one another. Why are we having this debate, then? It’s no longer a debate about empirics, which can be resolved. It’s a debate about what we assume states do, and we’re not going to resolve that anymore than we’d resolve a debate about the perspectives we have toward the world.¹

Powell (1994) expands on this in reviewing another salvo in this debate. The article itself is multiple and touches on several issues dividing so-called realists and liberals, but he also emphasizes that we’ve been mistaking effects for causes. Seeking absolute gains or relative gains, however defined, is a function of the security environment and concerns states may have about things like the security dilemma or the offense-defense balance. Thus, we can assume that states are concerned with relative gains, but then the degree to which it’s concerned with relative gains becomes an outcome to be explained. When cause and effect are conflated, the question we’re begging is poorly understood.

I might expand on this more. There’s a lot more than can be said about the difference between what states “do” and what they “should do”. No matter, I at least want this here for now for when it comes up in class. “Absolute gains or relative gains” is a debate that misses the point entirely. I don’t think what we elect to discuss in the confines of this apparent debate is a productive use of energy at all.