The Societal Violence Scale: Explained

Physical integrity rights have always occupied a central position in the human rights discourse, and for good reason: the right to life and liberty is the first priority of virtually every human rights instrument and is the precondition for the enjoyment of other human rights. Likewise, physical integrity violations by the state have traditionally pre-occupied human rights scholars and activists, also for good reason: the modern state, by definition, seeks to monopolize the use of force within its territory and normally has a wide range of repressive tools at its disposal, with the primary responsibility to protect its citizens. Too often, however, these tools are used to threaten rather than protect the physical security of citizens, who have little recourse for relief.

Nonetheless, it is increasingly impossible, and perhaps irresponsible, to ignore the threats to physical integrity posed by non-state actors. Newspapers are filled with atrocities perpetrated by non-state actors. Violence against women and children is epidemic in most countries. Forced labor thrives almost everywhere. Armed groups like the ISIS have shocked the international community with the scope and brutality of their abuses. Indeed, on the basis of annual US State Department Human Rights Reports, it is hard to avoid the conclusion that physical integrity threats by non-state actors substantially eclipse abuses by state in most countries, most of the time. For example, it is not uncommon for a medium-sized state to report tens of thousand cases of sexual assault (e.g. Canada) and/or domestic violence (e.g. Vietnam). Some reports dispense with numbers altogether, instead providing estimates of the percent of victims reporting abuse. UN and NGO sources cited by the State Department estimated in 2013 that 70% of marriages were forced. If a country also suffers from honor killings, dowry deaths, FGM, bride-kidnappings, not to mention sex trafficking (e.g. most countries), the number of female victims alone can be staggering. Few states routinely terrorize tens of thousands, much less hundreds of thousands of their citizens but those that do are counted among the worst human rights offenders. Sadly, this level of societal violence is commonplace.

In the worst cases, society is beset by violence on all sides: high levels of societal violence coexist with, and often feed, high levels of state repression. In addition, some agents of the state may collude with criminal elements, exercising violence for their mutual, personal benefit. The results for human security are particularly pernicious: there is no safety in retreating to the private sphere for relief from violence and no recourse to government protection. However, even in the absence of significant state violence, societal violence can flourish. The problem is acute in stable democracies as well as unstable and undemocratic regimes.

If the ultimate goal is to protect the physical integrity of the person, it is hard to justify ignoring threats posed by non-state actors. Indeed, under the “tripartite” principle that states have a responsibility “to respect, protect, and fulfill,” the state is obliged not only to refrain from human rights abuses itself, but also has the duty to prevent abuses by third parties, including private actors, and to take positive measures for the provision of rights to everyone under the state’s jurisdiction.¹

Assessing the degree to which state versus non-state actors threaten physical integrity rights and understanding the relative threats posed to and by different non-state actors minimally requires more systematic and disaggregated measures of non-state actor abuses² to complement the several projects that measure state-sponsored violence, including the Political Terror Scale (PTS), which the authors help produce. The Societal Violence Scale (SVS) focuses attention on physical integrity rights violations by non-state actors. Indeed, the PTS and SVS may be used together as a kind of modified “misery index” for human insecurity, providing a more comprehensive picture of physical integrity rights abuses globally.

The present project seeks to develop measures of societal violence based on annual US State Department’s Human Rights reports.³ The Societal Violence Scale ranks countries on a 5-point scale (from the lowest level of societal violence to the highest) based on three criteria. First, we look at scope: the proportion of society that is victimized. Thus, widespread violence against women (who account for 50 percent of the population) figures more heavily in the final score than widespread abuses against human rights defenders, who represent a very small number. We also look at the severity of abuses. For example, evidence that human rights defenders are killed weighs more heavily than beatings of human rights defenders. Likewise, while women are routinely subjected to sexual violence and domestic violence, the addition of other types of violence against women like gang rape, sex trafficking, and/or FGM/C adds to our assessment of severity.

Unlike the PTS, the SVS also measures societal violence disaggregated by victim and by perpetrator. Disaggregation is necessary to determine more precisely if, when, and where threats to human security originate. In other words, it is important to know who is being victimized and who is perpetrating the abuses in order to craft effective responses.

Violence perpetrated by non-state actors may be organized into three categories, based on increasing levels of organization:

• Violence perpetrated by individuals or ad hoc groups
• Violence perpetrated by “corporate” actors: that is, organized groups for private profit
• Violence perpetrated by organized, armed groups for group benefits

Victims of societal violence are myriad, including:

• Women
• Children
• Victims of national, ethnic, and racial violence
• Victims of religious and sectarian violence
• Refugees, asylum seekers and internally displaced persons
• Sexual minorities
• Forced labor
• Labor activists
• Human rights activists
• Humanitarian workers
• Journalists
• State actors
• Victims of civil conflict
• Others (added as needed)

Coders are instructed to provide as complete a picture as possible of each relevant victim category with the understanding that this may entail double (or triple) counting. For example, if the report notes that indigenous women are particularly vulnerable to forced labor, the information is recorded as evidence of abuse of women, ethnic minorities, and forced labor. To represent it only as abuse of women would understate the particular vulnerability as members of indigenous minorities and underestimate the incidence of forced labor. Likewise, if it were only recorded as abuse of ethnic minorities, we would not fully appreciate the vulnerabilities of women and again underestimate the incidence of forced labor. Similarly, recording it in forced labor would obscure the degree of abuse of women and ethnic minorities.

Reproducing the data in all three relevant categories of victims–women, ethnic minorities, and forced labor – presents a more comprehensive picture of the threats suffered by each but prevents easy aggregation of the scores across victim category. There are good reasons to eschew easy aggregation in any case. In particular, it is not clear that each victim category ought to be given equal weight. A severe problem with violence against women is a severe problem facing roughly fifty percent of the population. While a severe problem with violence against labor activists, while no less morally reprehensible, simply affects a much smaller number of victims. Thus, any aggregate measure of societal violence cannot be arrived at by simply adding up the scores from each category but must be carefully constructed and transparent about the weighting mechanism.

Although complicated, the effort to disaggregate societal violence by victim and perpetrator promises several advantages. On a very practical level, different threats are likely to require different responses. For example, policy responses to domestic violence perpetrated by thousands, tens of thousands, or even hundreds of thousands of individuals against an equal number of individuals in their homes would obviously differ from efforts to diffuse sectarian conflict among an equal number of people organized to advance or defend group rights or benefits. Researchers will also benefit from disaggregation across both perpetrators and victims. Researchers who have specific interests in understanding violence by their perpetrators or by their victims can readily identify the appropriate cases. In addition, disaggregation may also allow researchers to explore patterns of abuse, by perpetrator and/or victims, in relation to other relevant social, political, economic or geographic characteristics.

Thus, the SVS uses two distinct but closely related coding schemes to measure violence against specific victims and overall.

Coding by victim/perpetrator category:

Level	Interpretation
	This is the default position, indicating no or relatively fewer reports of abuses against members of this group.
	There is evidence of a notable problem with violence against members of this group.
	The report reveals a serious, widespread problem with violence against this group.
	Reports of violence against the group are pervasive in scope as well as severe in nature and may assume a variety of forms. It affects a significant proportion of the group.
	Violence against members of the group is ubiquitous in scope and egregious in nature. It affects a large proportion of the population of the group and assumes a variety of forms.

Violence across the country is also coded on a 5-point scale:

Level	Interpretation
	Societal violence is limited in scope and severity, with relatively few victims and few perpetrators.
	Societal violence is a problem, affecting a significant number of victims, albeit across few victim categories and of a less severe nature.
	Societal violence is widespread and serious in nature. It affects a significant number of people across several victim categories.
	Societal violence is pervasive in scope, severe in nature, assumes a variety of forms and affects a large proportion of the population typically across several victim categories and perpetrators.
	Societal violence is ubiquitous in scope, egregious in nature and assumes a variety of forms. If affects a large proportion of the population, commonly crossing numerous victim groups and perpetrators.

Feedback is welcome!