Modeling A Conflict-Free World Based On Trade And Military Alliances
A recent article from the MIT Technology Review examines a study conducted by Matthew Jackson and Stephen Nei of Stanford University.. The study, which combines network and game theory, has shed light on the importance of trade links between nations with regards to producing a stable network of countries, and therefore, peace.
Jackson and Nei initially looked into whether a hypothetical network of countries could ever be stable with only military alliances, where no country is vulnerable to attack, and no country has an incentive to change alliances. Their calculations revealed that network stability against war was impossible with only alliances; the creation of new alliances creates instability in existing alliances. They then introduced the possibility of links being formed through alliances or trade. This change dramatically altered the stability, creating a much more densely connected and stable network, somewhat resembling our current world.
It is interesting that with the simple addition of economic considerations(trade) into the network, the payoffs are changed with regards to the strategies of either maintaining a peaceful status quo, or attacking another country. The introduction of trade lowers the payoff for attacking due to the potential disruption of trade. In class we worked with models which included only one form of payoff (typically a dollar per edge), however in Jackson and Nei’s model, there are two forms of payoff; the payoff from a potential war and the payoff from trade, or lack thereof.
The article also touches on the potential predictive power of such models using the analogy of the spread of forest fires, the idea being that the outcome of the fire is more dependent on the structure of the network rather than the initial spark. This brings to mind the idea of embeddedness from Chapter 3. More common neighbors between any pair of nodes implies a denser network, which affects any type of propagating network effect. In the case of the additional economic ties, they seem to form a sort of network glue, which prevents the network from becoming too unstable, and also prevents an initial shock to the network (e.g. terrorist attack) from wildly propagating in an unmitigated fashion. I would be curious to see what sort of predictions can be made regarding conflict using a real-time model based on current global conditions such as economic ties, market conditions, natural disasters etc….. I have no doubt that this would be a monumental task, but given the power of big data, it doesn’t seem too far fetched for the near future.