ACPED Data

The VERSUS project was the first of its kind to examine conflict patterns as a function of elite politics.

It asked: why are the patterns of modern conflict so poorly explained by the proposed causes of conflict? In 2017, when Versus began, conflicts were often explained by the ‘failure’ and ‘fragility’ of governments, poverty, sectarianism, political exclusion and even environmental factors. But these factors could not explain why conflicts differed so much in practice, and why countries with similar political, economic and social structures had very different conflict risks and trajectories.

VERSUS presents a new paradigm for understanding the relationship between governance and conflict. It argues that political relationships between subnational elites and regimes incentivise violence in states under stress.

The VERSUS team worked from four cornerstones:

(1) Governance in states under stress is best understood through the relationships between regimes and subnational elites

            We investigated and proved this through extensive research into patronage, the architecture of governance, formal and informal power, networks of power and influence, appointments, and subnational authority volatility.

            This component of the work was the most extensive: previously, there were no systematic or thorough records of how power was created, held, changed or distributed within unstable states. Indeed, the presumption was that power was informal, highly centralized in individuals, and deeply corrupt. We found – through extensive data collection and investigations, that the opposite was true:

A.    Power is formal- having a formal government appointment is often the sole way that people understand ‘power’ and ‘authority’. With an appointment comes access to funds, offices, staff and authority. It also represents a position on a ladder- or hierarchy- of power within a party or institution. These people become elites.  We documented these positions at the cabinet level in our ACPED dataset, and we documented individual countries subnational authority structures in multiple articles.

B.    The hierarchy is important and has an architecture- that structure is highly conditional on the country, but at the top, there are networks of power associated with individuals. This was documented in individual country papers. What did we learn?

a.     Largely inclusive of ethno-political and ethno-regional groups within states

b.     Power is far more formally assigned than informally manufactured

c.     Authority at senior formal levels (like cabinets) is inclusive but unbalanced

d.     Why unbalanced? That is how elite influence on the executive is demonstrated

e.     However, large groups, powerful agendas and agents are not necessarily more represented: smaller groups, those that are easier to manipulate, are more likely to be ‘over-represented’ in cabinets

f.      Appointments have proliferated at all scales

C.    When we examined ‘architectures’ and methodologies for uncovering these structures, what did we learn from this investigation?

a.      Countries have a geography of power and hierarchy of power

b.     Regimes are highly inclusive of different ethno-political groups and regions

c.     There are still significant problems, some stemming from the type of inclusion

d.     Subnational elites and subnational politics have rarely received enough attention

e.     But these relationships and their character underlie real and frequent variations in power, conflict and development 

D.    Informal power is not widely understood as translating into authority, but more likely to be associated with communal ‘wisdom’ or ‘advice’.

(2) Elite relationships are built on the exchange of positions, resources and violence management

It is less important to many elites to deliver on governance objectives than to uphold the structure of power. In this way, the welfare of populations is secondary to the pursuit of power and ‘climbing the ladder’.  

A.    If having an appointment is the sole way to access funds, authority and progress, then keeping such appointments is vital.  People will invest in these positions, and use violence to retain them (or gain access to them).  

B.    The networks of power around the senior elites are highly competitive and contentious. The leader is surrounded by often combative and contesting individuals whose power must be balanced and mediated in order to not become volatile. This ‘elite management’ is also difficult. In our individual state investigations, we created the ‘elite architecture’. Below is an example of the Ethiopian system, in which individuals (with political agendas of their own) are appointed for variable periods of time.

C.    We worked extensively on the idea of ‘alignment’ and worked on a special theme on this subject.

(3) Elite relationship dynamics shape modern conflict patterns

How does this affect the types of conflict we see? The power structure shapes the incentives for violence, and the elites – and how they react to those incentives- shapes the type of violence that will occur. This article links the shifts in elites and appointments to a move toward democracy;  this one links the subnational politics of African democracies to specific forms of militia behaviour on behalf of government elites.

But the larger story is that the shift in violence rates and form is parallel to the change in government structures, size and appointment competition.

A.    Very few countries have rebel/insurgent problems, and most have militias. These are operating on behalf on political elites (very commonly, known political figures) or local/regional "big men". For example, Boko Haram was sustained and supported by the governor of Borno state (Ali Modu Sheriff) in the early period, and while BH went on to engage in much more extensive violence, it is the violence in Nigeria's Middle belt and now, in the NorthWest, that is much more in line with common conflict patterns.  Smaller militias are prolific, often have specific political objectives, their relationship to political figures allows for impunity, they surge during periods like elections when their client sponsors need targeted violence, and then engage in other forms of quasi-political actions and crime in other periods. 

B.     There has been a very large shift in the type and number of groups committing violence across Africa has led to a consistently high rate of conflict that is difficult to eradicate. These groups are not involved in political negotiations, but function more as violent labour for political elites (see the attached article). With more political elites in the system due to widespread inclusion and government appointments, there is a drastic increase in zero-sum political competition. This competition is directly contributing to the need 

C.    In keeping with the Ethiopia examples, consider the outcome here (Arat Kilo is the center of power):

(4) Political elite relationships mean that elites benefit from conflict, but it creates much higher levels of violence exposure and vulnerability for populations

A few attributes of modern African conflict (and those in Yemen) are critical here:

A.     Armed groups differ with respect to their political goals, and this influences the form they take. Traditionally rebels are engaged in a struggle for state control. An example of this is the RSF in Sudan- they graduated from a militia operating in Darfur (as the Janjaweed), but once their leadership had national ambitions, their actions were geared towards engaging and challenging the state's security forces. Al Shababb is another very traditional 'insurgent' or 'rebel' group: national political control is their ambition (whether they will achieve it or not is a different story- as they very rarely do). 

B.     These groups and the forms of conflict they engage in is not a sign of 'government weakness' : it is a feature, not a bug of the political systems that have evolved. These are extremely common circumstances, and their volatility has a lot to do with the political calendar and the need to compete. An ACLED based conflict exposure index (https://acleddata.com/conflict-exposure/) demonstrates that if we consider how many people are exposed to violence by these groups- state forces expose 11% of global population in 2023, and rebel expose 2%, but militias expose more than 4 times that of rebels and 50% more events (indicating their actions are more prolific but widespread rather than concentrated like rebels)

Jihadists can be a bit of an exception to this rule: in part because their political motivations are not clear (and often anti-state/anti-system). But in truth, while that is evident more for IS related groups than for AQ affiliates, there is too much variation in the expected political objectives even between IS affiliated groups (ie. ISWAP acts and controls differently than ISGS; and many of the groups that have loose affiliations with IS (like those in Cabo Del Gado, Moz) don't have the control or capabilities yet to move on from hit and run insurgent behaviour.  

C.    It is more important to refute this more typical interpretation of conflict:

And instead consider this one:

D.    But what we learn from these cases is the following:

1. Conflict is widespread, varied and often beneficial to political interests

2.     Political competition causes conflict

3.     Central hierarchies have interests, not always preferred local outcomes

4.     Groups and elites respond to the ‘violence market’

5. Conflict isn’t a breakdown in governance: it is a feature of modern governance and political change

In conclusion, VERSUS created multiple real-time measures of power distribution across many select unstable states. It determined how violence erupts from political and governance processes in varied environments and examines how common internal and external shocks create new trajectories of governance, violence and potential for political resilience.  It had five objectives and they were well met, with considerable advancement ongoing.

Versus has advanced a new paradigm on subnational political architectures and environments over static institutionalism;

It has generated several measures of comparative political power distributions in developing states that capture the degree and depth of regime and elite relationships;

It designed and tested scenarios to explain how, when and where violence erupts as a strategic function of architectures and environments;

It examined how ‘resilient’ regimes, elites and vulnerable members of society in response to internal and external shocks, through exposure.

It collaborated with development practitioners and civil society to understand the new patterns of power and conflict emerging.