Articles and Posts from ISQ

In this fine study, Bell and Wolford investigate whether the discovery of oil resources increases the risk of civil war onset even before the actual exploration begins. In particular, the authors contend that the effect of petroleum discovery should be mostly visible in weak states where oil exploration can produce substantial changes in the relative power between the government and the rebels. The argument is developed formally and tested with data culled from UCDP and several other sources. The empirical results generally corroborate the main theoretical contention.

Oil discoveries allow the authors to gauge actors’ expectations about future power distributions and, thus, to empirically capture current preferences for conflict or cooperation. Bell and Wolford argue that oil discovery functions as an exogenous shock with great potential to shift bargaining power in dynamic ways. Newly discovered petroleum reserves enable governments to militarize, repress domestic challengers, and alter incentives for credible commitment. Anticipated wealth from oil exploration reduces government’s motivation to negotiate a peaceful deal and makes it more likely to renege on previous concessions. Conversely, oil discovery might also provide an impetus for agreement with the rebels as the central authorities will be better equipped to accomodate demands for a more equitable distribution of domestic resources. Finally, ‘black gold’ discovery fundamentally shapes rebels’ incentives for war and peace: as Bell and Wolford note, the unearthing of oil deposits predisposes insurgents to initiate conflict in order to forestall adverse shifts in the power distribution.

While wading through the theoretical discussion, the protracted South Sudanese insurgency (1956-2005) readily came to mind. [1] One of the reasons for the collapse of the 1972 Addis Ababa agreement - perhaps, the tipping point - was the discovery of petroleum deposits in 1978 around Bentiu, a town located in the Upper Nile region of South Sudan. In blatant violation of the 1972 deal (the agreement had granted the southern regional government authority to legislate on the exploitation of mineral deposits in the South), Khartoum attempted to redraw the boundary with the South and dispatched Northern soldiers to guard the oil fields. The move contributed to the collapse of the Addis Ababa agreement and added fuel to the smouldering SPLM (Sudan People’s Liberation Movement) insurgency. Noteworthily, when civil war broke out again in 1983, the Bentiu oil installations were among the first sites attacked by the SPLM. Hence, the South Sudan case appears to validate the logic developed by Bell and Wolford according to which rebels have rational incentives to initiate military action against a government that is expected to accrue additional resources from the exploitation of newly-discovered petroleum reserves.

Bell and Wolford are to be commended for starting a serious conversation on the role of mineral resource discoveries and civil warfare. Being tasked by the ISQ online editor to perform a replication of this well-crafted study, I undertook the following steps. First, I re-ran the original models on the original data and the results were consistent with those reported in the article. Second, I re-ran the original models after cross-checking authors’ coding for civil wars onsets with the UCDP data. Apart from minor inconsistencies (a handful of civil war onsets were not coded or were improperly coded), [2] I also excluded three cases of civil war onsets - US 2001, US 2004, Romania 1989 - which conform neither with the UCDP conceptualization of internal conflict nor with the conventional definition of civil warfare as armed combat between two or more political organizations subject to a common authority at the outset of hostilities. In the US case we are dealing with terrorist attacks while in the case of Romania we are observing a revolution - two phenomena which involve violence but which are conceptually distinct from civil warfare. UCDP codings are carefully vetted but, in some (thankfully, rare) instances, remain prone to unit homogeneity concerns - a serious issue that I discuss at length elsewhere. Nevertheless, my expectation was that the exclusion of these three cases as well as the recoding of the handful of inconsistencies would not significantly alter the main findings. Indeed, when re-run on the adjusted data, the models produced substantively similar results.

Third, and more importantly, I replicated the basic models in Bell and Wolford using a different dependent variable: civil war presence (continuation), with data taken from the UCDP. The logic of bargaining failure in the context of anticipated changes in relative capabilities should extend to actors’ incentives for credible commitment in the midst of warfare. Powell, for instance, notes that a key function of persistent fighting is to forestall adverse shifts in the power distribution. Oil discovery creates expectations about favorable changes in the relative power balance, and reduces actors’ desirability to strike a deal. Essentially, oil discovery during conflict shifts antagonists’ discount rates, and increases the likelihood of civil war continuation since neither the government nor the rebels have an incentive to strike a bargain with the anticipated change in the balance of military capabilities. Table A presents the main models proposed by Bell and Wolford with the new dependent variable (a dichotomous indicator for whether civil conflict is ongoing). The result under Model 3 suggests that, as conjectured above, the discovery of oil reserves is a robust predictor of civil war continuation in weak states. This is a finding that reinforces the argument developed by the authors: the unearthing of petroleum deposits changes the balance of power, and makes leaders of weak states less inclined to prefer peace to the continuation of conflict. Obviously, more systematic analysis is needed to unpack the relationship between discovery of oil reserves and internal war duration. As a further step, I envision an analysis of conflict duration with the government-rebel group dyad as the unit of analysis and with the inclusion of other covariates that have been found to alter actors’ incentives for credible commitment, such as rebel movement fragmentation. I hope that Bell and Wolford will successfully embark on this project as well.

[1] The authors use the case of Sudan to illustrate the delayed realization of oil discoveries.

[2] Suriname 1987 ; Spain 1980 ; Spain 1985 ; Azerbaijan 1992 ; Niger 1996 ; Togo 1991 ; Uganda 1979 ; Thailand 2003 ; Laos 1989 .

florea rep



ISQ On Twitter

The International Studies Association

Representing 100 countries, ISA has over 6,500 members worldwide and is the most respected and widely known scholarly association in this field. Endeavoring to create communities of scholars dedicated to international studies, ISA is divided into 7 geographic subdivisions of ISA (Regions), 29 thematic groups (Sections) and 4 Caucuses which provide opportunities to exchange ideas and research with local colleagues and within specific subject areas.
Help   |   Thanks   |   Privacy Statement   |   Terms Of Use