25 August 2014

Storm chasers

We were talking about violence – let me explain. No, there is too much. Let me sum up.

The world is not going to hell in a hand basket, at least not immediately and despite considerable evidence to the contrary. Most measures of violence, including conflict and homicide, are exhibiting positive trends on negative trajectories. But this might be because we’re looking at the wrong horizon; we could well be missing the storm gathering in front of us as we’re too busy looking back, congratulating ourselves on what we just survived. It doesn’t necessarily follow that violent acts will continue to diminish, and even if they do, there will still likely be blips and backsliding and periods of man’s inhumanity to man that will need to be addressed by the international community (one hopes. The stunning failure of R2P has really left me disheartened, as did this suggestion. HAS HISTORY TAUGHT US NOTHING ABOUT PARTERNING WITH MONSTERS?! DO NOT ACCEPT THE PRINCE AS YOUR IR DOCTRINE, OBAMA). We can (again, hopefully) anticipate these same blips and plan out our reactions accordingly by analysing current trends in violence. So what are these trends, and how do we use them to forecast what awaits in the clouds before us?

Let us start with the excellently thorough Global Burden of Armed Violence from the Geneva Declaration on Armed Violence and Development. This sucker is the gold standard in discussing violence, along with the Small Arms Survey, and it deserves a much closer read than I gave it. The discussion that follows is, in the interests of full disclosure, mostly taken from the 2011 report, but some stuff from 2008 might have snuck in here.

Where the findings from the HSR were mostly rosy – things have gotten so much better over the past 10,000 years! – the GBoAV strikes a more measured tone, informing us that “at least 526,000 people have died directly or indirectly from armed violence – both conflict and criminal violence – every year in recent years (for what it’s worth, this was down by some 200,000 from 2008). One in every ten of all reported violent deaths around the world occurs in so-called conflict settings or during terrorist activities.”

Two quick notes about their data – first, they only use direct conflict deaths, though they do acknowledge that indirect conflict deaths (like those from conflict-related malnutrition and hunger, cholera, measles, and other preventable causes of morbidity and mortality) are “certainly the largest portion of the burden of conflict deaths.” A conservative estimate of the indirect/direct ratio is 4:1. And the case study for this? Our own fair Congo (though the next time around, it will probably be Syria), where at the turn of the century the International Rescue Committee “launched a major effort to better understand the human costs of armed conflict in the DRC.” Based on some six surveys that stretched from 1998 to 2007, IRC estimates that 5.4 million people died as a result of conflict. That is a staggering number of people in a very short timeframe; possibly without precedent in a nation that is no stranger to death on a grand scale (here is where I again suggest the excellent King Leopold’s Ghost, which explains that during its stint as the personal fiefdom of King Leopold and immediate aftermath – a period that stretches roughly from 1880 to 1920, the population of the territory was reduced by at least half. Half of what, you rightly wonder? “Only in the 1920s were the first attempts made at a territory-wide census. In 1924 the population was reckoned at ten million, a figured confirmed by later counts. This would mean that during the Leopold period and its immediate aftermath the population of the territory dropped by approximately ten million people,” most of whom died from ‘indirect’ causes including disease, exhaustion, and malnourishment). The primary approach IRC utilised to determine its death toll was a ‘verbal autopsy’ – a randomized household survey (it cost a heroic amount of money and gets our Health Advisor riled up whenever OCHA demands to know why we don’t have more specific M&M data). A number of folks – including our friends at the HSR Project – have challenged this figure and the use of survey-based approaches to calculating mortality rates, claiming that IRC overestimated ‘excess deaths’ by almost 60%. Without getting too much into the nitty gritty (too late now!), it boils down to an argument over baseline crude mortality rates (CMR). For what it’s worth, the Centre for Research on the Epidemiology of Disasters calculated a CMR in line with what the IRC was using, though they did note that “the overall CMR in all provinces in the DRC has decreased or remained stable over the past decade”, which only makes sense, as the lion’s share of the high-intensity conflicts have abated.

It’s a bit hypocritical of me to point the above out, considering I was so blithely dismissive of structural violence last week, but it is what it is.

Second, the report is necessary limited to recoded conflict deaths. They have a fascinating case study of Yemen in this regard – basically explaining that their own statistics for Yemen are profoundly undercounting violent deaths. I frequently bemoan how difficult it is to get people to care about Congo, but yikes, is Yemen even a profoundly forgotten conflict (you should probably go read about it. I’ll wait).

So, steadfast in the knowledge that the numbers we’re working with actually rather dramatically downplay the costs of conflict in human lives, let us move forward. Though the global average of violent deaths is about 7.9 per 100,000, at least 58 countries “exhibit violent death rates above 10” and account for almost two-thirds of all violent deaths. “Of the top 14 states most affected by armed violence (with violent death rates exceeding 30 per 100,000 population), only five have more than 1,000 conflict deaths in an average year (Colombia, DRC, Iraq, Sri Lanka, and Sudan).” Put another way, the majority of violent deaths (90 per cent!) do not occur as a direct result of conflict or in a conflict setting. Even so, one-third of the most violent states were either actively experiencing an armed conflict or had recently emerged from one and both kinds of violence (conflict and non-conflict, as if that was all that mattered) disproportionately impact the developing world, especially Africa (both) and Central and South America (non-conflict, in this case homicidal violence).

Yet another note about methodology (I’m super into semantics for some unknown reason): as you observant readers might have noticed, the report makes a distinction between conflict deaths and ‘intentional homicides’ that occur in a non-conflict environment. So…what is it when you have a homicide in an IDP camp, for example? In reading the GBoAV, I was reminded of an incident in Kabul when a series of shootings near our based sparked security warnings out the wazoo, but ultimately turned out to be more of a Hatfield/McCoy thing than insurgent action. “It is often difficult in fragile and post-conflict contexts to determine whether a death can be attributed exclusively to organized or interpersonal violence, or to political or economic motivations. Killings that are believed to be motivated by political or economic objectives may be the result of both or neither. In countries ranging from Afghanistan and Yemen to Mexico and Nigeria, the merging of organized criminal violence with armed conflicts of varying intensity renders a simple binary distinction between ‘conflict’ and ‘non-conflict’ meaningless.” How illuminating.

What are a few of the somewhat more concrete trends that the report teases out? Most violent deaths (as many as 60 per cent of all homicides) are the result of firearms. Most victims of violent deaths are men, though that varies dramatically by region. “In ‘high-violence’ countries, women generally account for about 10 per cent of the victims, while they represent up to 30 per cent in ‘low-violence’ countries. This suggests that intimate partner violence does not necessarily rise and fall with other forms of armed violence, and may not decline as other forms of armed violence are reduced.” Isn’t that peachy? Kidnap-for-ransom is a growing phenomenon, with approximately 1425 cases reported in 2007 in Latin America Asia, Africa, and the Middle East (this is a bit the problem with this field. I look at 2007 and say to myself, sweet fancy pants, this data is ancient! How can I possibly blog about it? Which is both somewhat valid and completely ridiculous, as the report was only published in 2011 – so, not ancient – and is intended to be cumulative. It’s rather like that time in grad school that I was dinged by a prof. for not updating a paper to include an analysis of events that had occurred the morning said paper was due).

One development that enjoyed quite a bit of print space in both the GBoAV and the HSR before it is the upswing in non-state violence. In fact, this is one of the few trends that is not decreasing in either comparative or absolute terms and most conflicts today involved at least one non-state actor (depending on how you define it, this category can include such a profoundly diverse group of characters as the army formerly known as Blackwater, Los Zetas, ISIS, and, for good measure, the LRA. No wonder it’s so sexy). Oddly, it seems that non-state conflicts are also becoming some of the most intense. Per the HSR – “battle deaths from non-state armed conflicts increased more than threefold from 2007 to 2011.” I think this one captures so much attention in the academic realm because of its implications for the state system; there’s a lot of speculation that we’re witnessing the dissolution of the state’s monopoly of the legitimate use of violence. Harkening back to the last post, were this trend both real and sustained, it would undermine Pinker’s first critical explanation for our march toward utopia. But it’s not just Pinker – this goes back to Weber as a necessary condition for the modern state-centric system. This reading is perhaps a touch hyperbolic – I don’t think that our use of the state as the foundation of the international system will end any time soon, but I do think it explains why so many states stagnate.

On the flip side of the coin, we have a goodly number of states non-legitimately using their monopoly on violence (in the coming years, I think it will be absolutely fascinating to read about how the militarisation of American police plays into this narrative). “In some regions, the state (or state agents) commit or are implicated in acts of armed violence. At least 30 states register more than 50 reported extrajudicial killings per year (at the time, that list did not include the US). Forced disappearances occur ‘frequently’ in more than a dozen countries and ‘occasionally’ in 20 others (from what I could tell, the US was also not included in these totals, despite our heinous and profoundly illegal tendency to render people and their families).

The HSR reminded us that, though state-sanctioned violence might well be waning, “governments often are the greatest threats to human security when the turn against their own citizens.” This alone forces me to remain sceptical of their entire ‘better angels’ argument; as long as the international community is willing to allow a despot to starve literally hundreds of thousands of his own people to death with little more than an tut-tut and artificial lines in the sand about how all of those people are killed, there are no angels. It’s not even putting a Band-Aid on a lost limb – it’s walking past someone who was just hit by a car and murmuring that someone should really call 911, and you would, but you just can’t use your minutes for just anything.

In a finding that is not altogether surprising, but does sometimes require reinforcement, GBoAV notes that “the aftermath of war does not necessarily bring a dramatic reduction in armed violence. In certain circumstances, post-conflict societies have experienced rates of armed violence that exceed those of the conflicts that preceded them.” This ties back into the link between violence and development. In what is a depressingly well-established cycle, conflict undermines development which can lead to rises in criminality and inter-personal violence which in turn tend to coalesce into formal conflict. Indeed, post-conflict states run a 20-25 per cent risk of relapsing into war. “So long as such countries must contend with high youth bulges (exceeding 60 per cent of the total population), soaring rates of unemployment, and protracted displacement, the risks of renewed armed conflict remain high.” (And we’re going to look at some of these aggravating factors soon, mostly ‘cause I’m a little bored and find conflict drivers fascinating). The report goes so far as to caution us that, when one observes a drop in violent deaths in a very recently post-conflict society, it might well be that the conflict in question demolished the surveillance mechanisms which would allow for an accurate tally of homicides and other violent deaths.

Let’s dig into this recidivism a bit more (I baked some brownies this weekend, so I’m well-provisioned to approach bleak topics). We’ve already talked about the conflict cycle somewhat, if I remember right, but it’s still really interesting. For one thing, it’s not a given for many people that the conflict cycle is, in fact, circular at all: many analysts see it as a linear movement from conflict to underdevelopment. “In almost all cases, armed violence generates negative consequences for people’s quality of life and the achievement of the MDGs” across pretty much all the indicators, including HIV/AIDS prevalence, while drops in violence are reflected in improved MDG performance. Moreover, the more intense the violence (generally measured in numbers of deaths), the larger its development gap. “Repeated cycles of violence over the past decades are linked to high poverty rates; in countries experiencing ‘major’ violence at any point during the period 1981 to 2005, poverty rates are, on average, 20 per cent higher than in countries that were minimally or not affected by violence.”

But does it go both ways? “At the micro level, there is mounting evidence that individuals, households, and communities affected by certain forms of armed violence – especially war – tend to underperform in social and economic terms. Similarly, a number of macro-level assessments demonstrate how states plagued with underdevelopment are particularly susceptible to disproportionately high rates of violence.” That would be a yes, then. It’s also worth noting the absolutely staggering economic costs of conflict, which fly a bit in the face of that old chestnut that war is good business. “The annual economic cost of armed violence in non-conflict settings, in terms of lost productivity due to violent deaths…could reach as high as USD 163 billion – 0.14 per cent of the annual global GDP.”

The conflict cycle isn’t exactly rocket science and its whys and wherefores are quite logical: states with epidemic levels of violence spend lavishly on armed/police forces rather than social or economic programming; the aforementioned loss in worker productivity due to death and displacement; actual physical damage to infrastructure; loss of FDI... “In proportional terms, countries that register lower levels of human development exhibit more violence….taken together, approximately one-fifth (19 per cent) of the world’s population resides in lower- and medium-income countries experiencing high and very high levels of lethal violence.”

Elaborating on this theme is the knowledge that “when examined in the aggregate, it is obvious that the global burden of armed violence is weighted unfavourable against the poor. The large majority of the estimated 526,000 people directly killed each year as the result of armed violence reside in low and medium-income settings.” This is especially true with regard to non-conflict armed violence, as high homicide rates are often tagged to extreme poverty and hunger, lower primary education enrolment, and high infant and child mortality. Taken together, the data suggests that violence is not simply correlated to absolute poverty, but to inequality. “Indeed, the higher the concentration of income among the rich, the higher the total levels of homicidal violence (US, I’m looking at you).”

That’s quite a bit of information to digest, and it really only scratches the surface of the scholarship out there. So how is a humble if forward-thinking humanitarian supposed to plan for the next global hotspot? I think the simple answer is that you really can’t. As with so much of social science, the topic of trends in violence is something of a kaleidoscope: ten scholars look at the same set of situations or statistics, and somehow managed to generate 18 theories on the major factor underpinning it and double that number of forecasts or recommendations. We’re all hammers in search of a specific kind of nail.

I didn’t even bother to wade into the discussions surrounding the ‘democracy recession’, conflict minerals, political inequality, internationalisation of intrastate conflicts, etc. etc. All of these are worthy topics deserving of deep reflection and analysis, and I’m not going to do that at all. Instead, we’ll next explore some of my preoccupations – climate change and demographics (I like food and am a chick). In the meantime, have a brownie on me and try not to think too much about conflict chocolate.