Given that measurement and specification errors are likely not at fault, these null findings could be evidence of a flawed theory. However, given the body of knowledge assembled by conflict researchers, this seems unlikely. How can factors such as group population, remoteness from the state, and lost political autonomy that have robust and well known effects on the onset of conflict, but have no bearing on the goals these conflicts are fought for? Furthermore, anecdotal evidence provided by quotations from rebel leaders suggests that there issomerelationship between the governability of a territory and the goals that a group will fight for.
One potential explanation for this lack of evidence is that the stochastic component of the relationship overwhelms the deterministic component in explanatory importance. If “war is in the error term” (Gartzke 1999) and difficult to predict evena posteriori, then it logically follows that the goals groups fight for in these conflicts could be equally difficult to predict. This framing shifts the area of inquiry from how rebels choose the goals to fight for in territorial civil wars, to why these conflicts are so difficult to predict.
Thus far I have referred to the state as merely something that inconveniences potential rebels by reacting to their secessionist aspirations. By introducing agency to the state in this causal story, we can begin to answer the question of why we cannot observe a relationship between territorial governability and group goals. More importantly, by investigating how states work to prevent secessionist conflict, we can gain further insights into the relationship between territorial governability and goals in civil war.
CHAPTER 3: THE CURSE OF GEOGRAPHY
The Nigerian military maintains a significant presence in the petroleum rich Niger Delta region. While the threat of secession is significantly lower than it was during the height of the Ogoni self-determination movement in the 1990s or the Nigerian Civil war in the 1960s, the government is apparently still worried given the frequency of military operations against rebels in the region (Walker 2009, BBC 2016, Owolabi 2017). This concern is not misplaced; loss of the oil revenues generated by the Niger Delta would severely hinder the government’s ability to meet its obligations, and militant attacks on oil facilities (Uguru & Faul 2016) have led to up to 30% reductions in production (The Economist 2016). Regular military exercises anger local residents who say that the government should “change its military approach” and “address the developmental challenges facing the region,” instead (Akwagyiram 2017).
At first glance, this belligerence may seem puzzling. Gone are the demands for secession and independence of the Biafrans; in their place, Ijaw groups call for “federalism and self-determination” (Opejobi 2017). Why the heavy hand if separatism is less of a concern today? One possibility is that the government is unwilling to tolerate the income loss a revenue sharing agreement would entail. Another is that the region’s petroleum reserves are so valuable that the government is not willing to risk any chance of losing them to a successful secessionist movement.
In fact, acquiescing to the region’s demands for development could make secession more likely. Modern nation-states are territorial political entities, defined by their borders with other sovereign states, and encompassing the territory within those borders. As such, the administrative competence and institutional capacity of a group is only part of the equation for how difficult governing will be after achieving independence. The other is how burdensome it will be to govern and control the specific territory that a group will control.
The governability of a territory can be influenced by numerous geographic factors such as the location and abundance of natural resources, ease of accessibility, and quality of infrastructure. Similarly, the human geography of a territory can determine how governable it is for a specific group.
Are people clustered in dense urban populations, making the administration of the territory easier? Are there numerous other ethnic groups in a territory that may be unwilling to submit to majority rule by the self-determination group?
Improving infrastructure, paving roads, and expanding electrification would all decrease the difficulty of governing the Niger Delta if it were an independent state. While developing the Niger Delta might satisfy activists in the short-term, the improvements to the region could decrease the difficulty of governing it sufficiently that they may decide independence is a feasible goal. Investing without addressing local grievances, the flow of oil revenues from the Delta to Abuja, could raise raise the risk of secession by reducing the amount of post-independence state-building required while not removing the desire to escape the state’s control.
If geography influences a region’s suitability for independent governance, and thus likelihood of secession, then this gives states a powerful source of information they can use to head off potential secessionist movements. If an excluded ethnic group inhabits territory that is particularly well suited to sovereign governance, and therefore secession, then the state may take pains to discourage the group from considering secession by increasing their coercive capacity within the group’s territory. This attention will manifest as increased levels of state capacity relative to less governable areas.
Yet secession is a strategic process and Chapter 2 has largely ignored the role the state plays in this dyadic phenomenon. Unsurprisingly, states are not content to sit idly by and let groups try to secede with swathes of their territory. The decisions states make in the face of this dilemma can explain why we do not observe more secessionist conflict given the plethora of aggrieved minority groups, and why governments often appear to prefer dealing with low level violence to meeting groups’ demands for regional development.
Understanding when governments will choose appeasement instead of preemption is beyond the scope of this project, but one plausible explanation is that more politically excluded groups are less likely to be appeased because they are less important for governments to maintain winning coalitions. Regime type could also constrain the government’s choice of strategy with coercive preemption less palatable to a larger electorate.
It is important to conceptually disaggregate state capacity for this argument. State capacity entails both ensuring a monopoly on the legitimate use of force (Weber 1965, Tilly 1985, Olson 1993) and maintaining political institutions (Acemoglu, Johnson & Robinson 2001). Improving a region’s
infrastructure could increase the attractiveness of secession by providing more tools for a hypothetical future state. Conversely, increasing the policing and surveillance capabilities in a region deters secession by increasing the costs of militarily challenging the state. Grievances are related but orthogonal, as public goods provision may decrease grievances if they are driven by disparate outcomes but may do little alleviate them when they arise from political exclusion.
The puzzling consistent lack of evidence for a relationship between governability an goals in Chapter 2 becomes less puzzling when we take a step up the causal chain. Secessionist conflicts often begin in places abundant with resources located far from the centers of state power. These factors affect the likelihood of secessionist conflict because dissidents will only rebel when they expect to be able to form a functional state within the borders of their territory following independence. Given the large geographic component of governability, governments use this information to target the groups most likely to secede for preemption efforts. In effect, attempting to find a relationship between governability and goals will end in failure because doing so ignores the effect of government on this relationship.
To test this argument, this chapter proceeds as follows. I argue that states are aware of the role of territorial governability in secession and use their knowledge of different territories within their borders to identify the most likely candidates for secession and proactively work prevent such movements from emerging. I conduct a cross-national test of this argument using geospatial data to measure both governability and local state capacity. Results indicate that governable areas geographically far from the center have elevated levels of state capacity compared to similar areas located closer to the capital. This pattern suggests that governments are deliberately cultivating a presence in these areas despite the cost of doing so. I close by discussing how this strategy can explain many of the patterns we observe in civil conflict throughout the world.