The peace bubble
Associate Professor of Public Policy, Harvard Kennedy School; author of Securing the Peace: Durable Settlement of Civil Wars (2009), and co-author of God’s Century: Resurgent Religion and Global Politics (2011)
Steven Pinker and Joshua Goldstein have produced impressive and important books on the decline of violence and war across time and space. Whereas Pinker discusses violence in more general terms and Goldstein limits his analysis to war, both scholars make the argument that violence has declined over the past 100 years, but in particular since WWII.
Pinker attributes this decline to the success of the modern state and the imposition of order across multiple levels in society. Not only has war across societies declined as a result, but so too has criminality and violence among individuals. Goldstein tracks similar trends in war; but for him peacekeeping and the United Nations are critical in helping to usher in this period of peace.
The books offer us an optimistic view of contemporary history: both essentially reduce to the claim that the world is getting more peaceful, and as a result, better. Sadly, I remain unconvinced by their characterization (peace),and its implications (a better world), but even if they are each right, I think what they have described is a bubble: a peace bubble.
First, their analyses hinge largely on data and trends in these data. In looking at their data however, a critical question emerges: are they sampling on the extreme? In statistical terms this would amount to sampling bias. One of their responses might be “what about the Thirty Years’ War?” which was responsible for killing one in five Europeans. But this begs a different question: if it happened before, what is to assure us the trend they have identified is uni-directional? Again, what if we are witnessing a kind of “peace bubble”?
In addition to this sampling bias and the possibility of extremes, could it be the case that the decline in violence is not an artifact of the main sources of data? We have excellent data of the wars fought among European powers, for instance, but even these are subject to considerable debate both on empirical (how many Europeans were there and how many died in battle?) and definitional grounds (do victims of war-induced famine count, were they counted by some scholars?). In more contemporary terms, we still don’t know, for example, how many Chinese died during the Chinese Revolution or Cultural Revolution, or even the more recent 1994 Rwandan genocide. Body counts themselves are notoriously difficult to assess, yet each author relies (Pinker to a lesser extent) on these data to support his respective arguments.
This last point highlights the scholars’ understanding of violence in only its most physical manifestation; its implications for death or life. Consider Pinker’s discussion of bullying. Pinker makes the case that bullying has been one of the forms of violence targeted for elimination, and I agree this is a good thing. But the subject of bullying opens the door to a very penetrating question: what if it is possible to be more cruel yet less violent?
Most people tend to think of bullying as physical intimidation, which is easier to identify than the much more destructive (and painful) psychological intimidation which often follows successful efforts to halt physical bullying. When researchers looking at schoolyard bullying, for example, broadened the definition of “aggression” to include psychological cruelty, they found that girls were just as “aggressive” as boys. Moreover, internet bullying is psychologically and emotionally devastating and on the rise, but only rarely results in any physical injury. So again, harm is increasing even as physical violence is declining. Rape is yet another example: in most cases it does not result in a death, and because it is systematically under-reported, it is possible that rape could increasing globally even as violence and war both decline (Pinker admits he has weak data on the global front). Such a narrow indicator of violence may cause us to overlook critical areas where harm is still done, just not physically.
And Pinker is not alone here. Goldstein too relies on physical harm—death—as his critical indicator. What about the psychological impact of war (e.g. post traumatic distress disorder)? Due to advances in the organization of war from medical technology (e.g. surgery, antibiotics) and mobility (e.g. the helicopter) and to the relatively much smaller scale of wars nowadays, many soldiers who would have died in earlier wars have survived physically (and not been counted), but been shattered emotionally and psychologically. Many cannot work and cannot love. Consider brain injuries. According to one estimate, mortality from brain injuries was 75 percent greater in Vietnam than in the most recent Iraq war. Although some soldiers recover, many do not, suffering both the physical trauma and psychological stress of war far beyond the battlefield. In neither case would the harm sustained be counted in Pinker’s and Goldstein’s analyses.
Making the argument that “all that is true, but it still matters that deaths are fewer” is akin to responding to a critic of the US war in Iraq by saying “aren’t you better off with Saddam Hussein dead?” The point of each argument is that the world is getting better; whereas if we decouple physical violence—and death—from harm, such a line of argumentation is called into question.
More generally, most people in the West believe violent death is a universal empirical indicator of harm (and so it is). But what if the harm that is being done to people today cannot be captured by physical violence? This is a major part of the fight between Palestinian Arabs and Israeli Jews: the Arabs say a Jewish settlement is, in and of itself, an act of violence, which then justifies a violent response (say, a suicide bomb). But in this interaction the Israelis have had the easier argument, because they can (and do) claim that their settlements are non-violent. We in the West have arrogated to ourselves the very definition of what counts as violence and what does not, and this leaves us vulnerable to lethal blind spots when we attempt to bargain with or coerce people who do not share our axiomatic connection of harm to death.
Consider how advances in technology make it possible to cause grave injury without killing, by say, deliberately creating refugees. This raises the question of whether the number of peoples left homeless by war has risen in proportion as the number of violent deaths has declined. If so, we would have another example of decreasing violence masking increased harm. I say this to remind us all that forced mass expulsion is considered a grave breach of the Geneva Conventions of 1948 and 1949, and to remind us of Slobodan Milosevic’s strategy in Kosovo: rape and kill a few in order to cause the bulk of the population to flee. Again, relatively few corpses would result, but the harm would be grave.
Finally, I raise an alternative narrative and the emergence of the moral hazard resulting from the industrial revolution’s relationship with warfare as proving to be a mixed record. On one hand, if killing conduced to coercion, then the tools of war that made killing more effective should have conduced to shorter, less bloody wars. Yet this proved a fallacy of composition, because too many countries were industrialized at the same time, so wars became nearly suicidal. Then the apotheosis of weapons came: the nuclear weapon. Now it became impossible to use unlimited means to pursue absolute ends, and under that nuclear umbrella, the term limited war entered the lexicon of international politics, and not as an ideal.
I would argue that not only did this initial condition subsequently make it possible for bad leaders to murder their own people with less fear of invasion and conquest from without, but that today, it is the permissive condition which makes it possible for them to harm their people without killing them. In other words, I raise the uncomfortable issue of moral hazard: were there not circumstances where war was a good thing? On the other hand, it is fair to point out that the same permissive conditions and technologies which make it possible to harm without killing, may make it possible to prevent harm without killing. The key, it seems to me, is to redefine harm and to separate it from killing.
All this is to say that although Pinker and Goldstein have done us a great service in writing these impressive works of scholarship, the optimism is likely either premature or misplaced. Not only am I not willing to accept that physical violence is the only or most appropriate metric for measuring a decline of violence, but I remain unconvinced that the barbarism of humanity is behind us. Even if we concede they are right about the trend save at the margins, we are given no sound argument for why the trend is uni-directional: what has happened before could happen again. What Pinker and Goldstein see as the victory of peace, I see as a bubble that can all too easily pop.
Visit here to view a webcast of a John F. Kennedy Jr. Forum at Harvard Kennedy School on Jan. 30 titled “Is War on the Way Out?” with Pinker, Goldstein, Toft and Kennedy School Professors Stephen Walt and Joseph Nye.