One trouble that many of us who work in and study security sector reform have is that we find actually measuring security a challenging concept. Due to a variety of factors – lack of good data, difficulty in operationalizing complicated social issues, and the specific aspects to focus on are just a few examples – what the proverbial “person in the street” sees as an everyday, tangible reality may not be readily quantifiable to the practitioner. To that end, in an effort to unravel and test some of the available sources that may (or may not) lead to better insights into police and, more broadly, state governance performance, a colleague and I began an initial examination of the validity of homicide rates as an indicator of state security. I believe that given the lack of reliable and security sector-relevant transnational data, especially for many developing states, there is something here upon which we might fruitfully develop a more effective vehicle for understanding the state of the state.
However imperfect a measure, homicides are at least accepted by practitioners and scholars as an indicator of state capacity. The true measure comes with homicide rates applied to context. Africa, with its range of states, allows for a deeper exploration within each national context.
Measuring Security in Nigeria and Côte d’Ivoire – A Comparative Perspective
How do you measure security? Clearly, security is experiential, subject to perceptions and experience. However, this hardly lends itself to large-N comparisons among states. Our paper sought to model how people experience the state. More precisely, how citizens interact with the organs of the state that a majority of people see and often deal with daily: state security actors.
Based on the notion that measuring security and safety is both a straightforward concept (e.g., number of crimes, budgets, staffing) but concurrently notoriously difficult to measure at the interpersonal level, there is clearly a space between numerical indices of security and personal anecdotes, but we wanted to test whether one variable – homicide – matched against two deeper cases studies, would give some of those interpersonal insights.
Not to complicate things further, we wanted to look at a couple of states that were being buffeted by resource competition, particularly the oil-producing states of West Africa.
Our two cases demonstrated that homicide rates can give some insights into policing capability, which extends into a reasonable valid proxy for broader state capacity. In both Nigeria and Côte d’Ivoire, above-average homicide rates correlated with perceived policing dysfunction, though the lack of well-founded data (telling in itself) sheds doubt on the efficacy of homicide data for more than the broadest of indicators.
More specific conclusions, by topic, include:
Weak Security Apparatus. Analyzing specifically for Nigeria, Robert-Okah and Wali’s recent observations hold true not just for Nigeria, but Côte d’Ivoire as well. They hold that a weak security system “arises from inadequate equipment for the security arm…in addition to poor attitudinal disposition of security personnel…some personnel get influenced by ethnic, religious or communal sentiment.” This results in citizens sabotaging government efforts by supporting and fueling insecurity, allowing criminals “to escape the long arm of the law.”  The lack of material goes beyond weapons, encompassing a dearth of vehicles, fuel, information management and communications equipment, among others.
Oil. There seems to be little connection between oil and state capacity. We intended to isolate oil-producing countries based on their production levels, or their expected mid-term production capacity. Côte d’Ivoire, as a low-to medium producer of oil (though with great potential for growth) and Nigeria as Africa’s largest producer, are chronically unable to connect their respective oil wealth to enhance internal security. The evidence does not support in either case oil as a nation-wide variable that enhances the potential for violence, but in neither case does oil wealth lead to enhanced state capacity to deal with everyday violence or the more regional-specific internal identity-related violence.
Homicide as Indicator of Institutional Capacity. Neither case is universally what one might consider a failed state (recognizing that this term has seen a reduction in practice), but there is a clear and apparent weak institutional capability. In both cases the national homicide rates – well above the global average – indicated a gap in the institutional capacity of the police, which in turn is a barometer of nationwide institutional capacity. However, the problems with the data (most pronounced in Côte d’Ivoire) make drawing more specific inferences doubtful. In the case of Nigeria, “The foundations…are very shaky and have resulted in the deterioration of state governance and democratic accountability, thereby paralyzing existing set of constraints including the formal and legitimate rules…the state of insecurity in Nigeria is a function of government failure.This “paradox of plenty” – where in a very rich country has very poor people – leads to the insecurity of lives and properties.  Murphy’s analysis on the interplay between oil production and state incapacity outside of those institutions that matter most to oil production plays out to reasonable degree in both cases:
The current international system that makes international recognition – not internal legitimacy or functionality – the key to state authority works to the benefit of dysfunctional oil producers in the developing world. Enclaves that are valuable to oil consumers – and to the domestic elites who facilitate and benefit from international legitimization – function well enough. They include oil and gas fields, export terminals, oil-related shipping, and offshore infrastructure around which defensive perimeters can be drawn. 
Significance for Regional Instability. The ultimate question surrounds a state’s capacity to withstand internal pressures when combined with resource competition. In both Nigeria and Côte d’Ivoire, they experienced internal instability, though this instability was not a direct consequence of oil production, but rather was a symptom of greater social, political and economic dysfunctions. Police reforms may assist in mitigating some of the greater concerns surrounding crime and its prevention, but until macro-level issues, including economic growth and governance, these reforms will likely serve only as a temporary salve. Until these issues are ultimately addressed, perhaps with the commodity-derived incomes, regional instability emanating from and affected by Nigeria and Côte d’Ivoire will likely continue.
So, where might a path for this research lead? Among the next steps, I believe, are (most obviously) further refinement and testing through other cases, beginning initially with the littoral states of the Gulf of Guinea, and then extending outside the region, with selected states in Latin America and the Middle East, controlling for the presence and absence of oil and other readily-extractable resources. Moreover, this research helps broaden the appeal to improve modeling on estimating homicide rates, which in turn will help further assessments of governance and state capacity. Improving citizen safety and security is highly contextually dependent, but I believe that using homicide data as a cross-national indicator can help all concerned parties to address both where and how to focus attention and resources.
Joseph L. Derdzinski is a visiting associate professor of political science at the Colorado College, and a senior international development and governance adviser. Much of his research and consulting centers on the study of the democratic consolidation processes of post-authoritarian states. Among his more recent activities are research on measuring governance in West Africa, serving on election observation missions in Egypt and Afghanistan, including 2014’s contentious election audit, and policy research to help the then newly-formed Ukrainian government develop an accountable national guard.