By Rob Hickey, Guest Blogger
Source: Getty Images
Finding relationships between phenomena is human. This attribute of the human mind has had strong survival value in nature and we therefore carry it strongly with us today. In one of its many modern incarnations, it has allowed us to develop hypotheses regarding what might make good (and bad) public policy.
Despite the benefits of this evolutionary gift, an active mind without the application of logical rigor can sometimes lead mentally healthy and logical human beings astray. Over the past 100 years, well-intentioned U.S. policy-makers have developed and implemented policies that have expended immense amounts of public money while simultaneously failing to achieve their aims. Others have resulted in unsettling unintended effects.
The ability to survive in hostile and chaotic environments has bestowed us with the side-effect of occasionally seeing patterns that are not there, and to often simplistically assume that an effect results from a single cause. We attempt to overcome the incorrect and bizarre paths down which this leads us by methodologically analyzing data (i.e. the glory of science) which may link phenomena (or disprove these linkages) and which isolate the object of analysis from the surrounding chaos by controlling or explaining away extraneous information.
Identifying the Linkage
In their paper, Quantifying the Influence of Climate on Human Conflict, published in the journal Science on August 1st, Solomon Hsiang of Princeton University, and Marshall Burke and Edward Miguel from the University of California, Berkley, aim to do just that. They present a body of evidence which they claim implies a causal relationship between rising temperature and increased interpersonal and intergroup conflict.
Their paper concludes that “for each 1 standard deviation (1σ) change in climate toward warmer temperatures or more extreme rainfall, median estimates indicate that the frequency of interpersonal violence rises 4% and the frequency of intergroup conflict rises 14%.” They find that, since inhabited areas are expected to warm by 2-4σ by 2050, anthropogenic climate change could have a critical impact on human conflict. They define this impact as “important” if the authors of the 60 studies that they analyzed state that the “effect is substantive” or if the change in conflict risk is greater than 10% for each standard deviation change in climate variation.
The Signal and the Noise
The paper has its detractors, such as Dr. Halvard Buhaug from the Peace Research Institute Oslo, Norway who has criticized the study and pointed out that not one case of real-world conflict was identified that would not have occurred without climate change. This criticism is important and does illustrate the complexity and care which must be taken in attributing one particular cause to a particular effect, especially in a system as complex and seemingly un-dissectible as human society. In the interests of full disclosure, it is relevant to note that Dr. Buhaug’s previous academic work refutes linkages between climate variability and armed conflict.
But, this is also is something of a straw man line of reasoning, as it implies that the authors of the study are claiming that climate change alone can cause conflict, rather than exacerbate existing tensions. The authors never make this claim, but they do anticipate this potential interpretation of their work by writing:
“We do not conclude that climate is the sole – or even primary – driving force in conflict, but we do find that when large climate variations occur, they can have substantial effects on the incidence of conflict across a variety of contexts.”
One of the main points of contention around the web seems to be that the study does not properly control for other societal variables which affect violence.
The authors attempt to address this difficulty by saying that a:
“problem occurs when researchers control for variables that are themselves affected by climate variation, causing either (i) the signal in the climate variable of interest to be inappropriately absorbed by the “control” variable, or (ii) the estimate to be biased because populations differ in unobserved ways that become artificially correlated with climate when the “control” variable is included.”
So, the study intentionally does not include control variables, as they may themselves be proxies for climate change and may erase some (or all) of the relationship between temperature and violence if included. Instead of trying to account for societal differences, the studies that were examined are only those which compare one society to itself during different intervals of time. While some differences may exist, this approach endeavors to minimize most of them.
Is this solution good enough to separate the temperature signal from the socioeconomic noise?
Is Violence Increasing?
The findings of the study make intuitive sense given the notion that the world is going to hell in a hand basket. It seems clear that the earth is warming and violence around the world is on the rise.
When it comes to violence being on the rise – not really. In addition to the fact that we are terrible at assessing risk and macro-trends given our propensity to rely on anecdotal evidence, we also have the ability to accurately understand and interpret data.
This includes Harvard Psychologist Stephen Pinker, who in his book The Better Angels of Our Nature: Why Violence Has Declined, makes the case that violence is actually decreasing around the world.
To make use of one particular case from the U.S., it appears that gun violence and homicides dropped sharply between 1993 and 2011, according to a Pew Research Center analysis of government data. During this same period, 17 out of 19 years showed higher than average temperatures (from 1901-2012) in the contiguous 48 U.S. states:
What is going on here?
Frankly, this one case from the U.S. does not necessarily run contrary to the finding of causal linkages between warming and violence. Beyond it being a sample size of one, a number of important contravening factors could be at play here that may result in a decline.
The Pew Research Data presents some of these, including the collapse of the crack cocaine market in the early 1990’s and increasing economic growth, which created employment for those who might otherwise have turned to lives of violent crime.
It appears then, that examining a single geographic location over time does not always control for other societal drivers or dampeners of violence enough to draw such conclusions, especially over relatively short time scales. Even within countries, various sub-national geographical divisions may exhibit wildly different conditions that may allow for the development of violence.
This may, for example, be particularly true between different jurisdictions in highly federalized countries such as Germany or the United States where legislation and socio-economic health may differ substantially. In this context, the location of the violence within the country becomes highly relevant. The more geographcally consistent the time-series data is, the better it will be. Unfortunately, for some of the data used, violence statistics are aggregated at the national level and in some cases, the national studies have periods as short as 20 years.
But maybe a large enough sample size (60 studies were used) and time duration (in one case this span was 12,000 years – from 10050 BCE–1950) might be able to overcome differences in economic health, legislation and other factors which change the dynamics a particular areas over time – that the noise is self-cancelling. This is an assumption that this study’s authors make.
Pinker’s book on the decline in violence provides some clues on how, in a world that is scientifically proven to be warming from human activity, and where rising temperature appears to result in rising violence, that overall bloodshed might decrease. These include the rise of the nation-state (which holds a monopoly on force and punishment), increasing commerce, an increased respect for women’s rights, increasing exposure to other cultures and viewpoints, and the increasing use of reason being applied to potentially explosive situations. These ‘calming’ influences are all given as potential pathways that may allow for the expression of our nobler virtues.
What Might The Link Look Like?
Arguments against a linear relationship between temperature and assault have been made, based on a finding that at high temperatures, assault actually decreases. Without taking into consideration the time of day when assault occurs, it appears that in Minneapolis (where the data on which this study is based is from), assaults decline at temperatures above 74˚F:
If true, this brings the connection between anthropogenic climate change and increasing assaults at higher temperatures, into question. Other studies however, such as Bushman et. al. (2005) found that the time of day is also highly relevant for the incidence of assault and shows linearity in the data when taking this into consideration. This appears to refute the claim that extremely high temperatures actually discourage assaults. When controlling for the time of day, the temperature/aggravated assaults curve appears quite different from the above:
According to the Bushman paper, one postulation as to why aggravated assaults dropped during the hottest part of the day, when there is a proven relationship between temperature and violence, is because “there are strong inhibitions against committing aggressive acts at work, church, or school” despite the heat. The potential role that institutions might play (the effects of sunlight or circadian rhythms could also be considered here) in mitigating violence also appear to resonate with Pinker’s work.
Over the Precipice of Civility
Getting back to the original study, the Hsiang et. al. (2013) paper on violence and climate suggests possible mechanisms under which climate change might increase the incidence of violence. Interestingly, these overlap significantly with Pinker’s pacifying forces in that rising temperature and climate change may erode them.
These include the relative allure of conflict over economic activity, when productivity declines, or crumbling state power resulting from falls in tax revenue and law enforcement capabilities. Warming may also exacerbate social inequalities and violent attempts by the poor to redistribute wealth (a potentially ironic comic-tragedy of free-market ideologies which ignore the environment). It is also likely to increase food-related price shocks and the violence that they cause.
Increased violence may also be attributable to the migratory pressures of climate change on refugee populations and a lack of the bare necessities to sustain them. Finally, as anyone who has sat in traffic in a hot car might attest to, there appears be a mechanism whereby hot weather increases aggressive tendencies and reduces rational thinking. Whether it is a change in heat that triggers this, or higher constant heat is somewhat irrelevant, as temperature is tending upwards and will exacerbate both.
So, while violence may decrease overall in keeping with recent trends, warming may decrease the rate at which it has been dropping and could remain somewhat ‘hidden’ from policy debates. One of the unnerving aspects of this is that warming can potentially reverse the stabilizing effects of the institutions of the nation-state and the economic opportunity that a stable climate provides. Countering the degradation of the forces which promote this downward trend in conflict may be an insurmountable task for individual nations to address.