“All in all, I wish we had discovered water”. Sheik Ahmed Yamani, Oil Minister, Saudi Arabia (Ross 1999).
“We are in part to blame, but this is the curse of being born with a copper spoon in our mouths”. Kenneth Kaunda, President of Zambia (Ross 1999).

These comments portray well the increasingly popular notion of the ‘resource curse’. At first they strike as entirely counter-intuitive. After all, they are known as mineral ‘endowments’ precisely because of the positive connotations that are attached to them. However, recently a combination of poor economic performance and some high profile conflicts in mineral rich countries has thrown this traditional consensus into confusion. This has lead to a situation where Sheik Yamani and President Kaunda can make such assertions in agreement with a growing body of empirical and theoretical evidence. Such a view directly challenges the notion that politics is the root cause of conflict and due to the differing policy implications associated with these positions it is extremely important to resolve this issue to allow for a better understanding of conflict. This argument will focus exclusively on resource abundant economies taking advantage of the higher profile nature of these conflicts that allow for a fuller analysis of the root causes. In doing so it quickly becomes apparent that minerals do play an important role in conflict in these countries. However, this should come as no surprise. Their very being resource abundant makes such a fact unavoidable. In an economy that is dependent to a large extent on mineral endowments any major spending programme will naturally be drawn heavily from such wealth. Thus, the existence of a link between mineral wealth and conflict has little explanatory worth for the actual onset of the event. In looking at the cases of Angola, Indonesia and Zaire it can be clearly demonstrated that though minerals are of significance they are only so in the context that political choices led to. The discussion begins with an evaluation of the ‘resource curse’ argument looking closely at its empirical grounding and two main explanatory models: rent seeking and the rentier state. The robustness of this analysis is then questioned and the relationship of mineral resources and politics to the root of violent conflict is assessed through the use of detailed case studies.

The strength of the ‘resource curse’ argument rests heavily in an array of recent empirical evidence that must be considered briefly. The following is not a complete list but presents a fair selection of the empirical grounding of the ‘resource curse’ position. Sachs and Warner (1999) prove that a significant positive relationship exists between the rate of natural resource exports and lower growth rates. Collier and Hoeffler (1998) and Hegre (2002) show that higher primary export rates are positively related to the onset of conflict. Studies by Fearon and Laitin (2003) and de Soysa (2002) using oil exports, and Humpfreys (2003) using oil production, all find a positive relation to war onset. Lastly, Fearon (2004) demonstrates that greater presence of contraband (drugs and/or gemstones) is correlated with increased conflict duration. Resting heavily on the basis of this evidence two main explanatory models are used to account for the existence of the ‘resource curse’.[1]

The rent seeking argument derives closely from Collier’s (2000) distinction of ‘greed’ and not ‘grievance’ factors as being the motivating force behind conflict. In this way it sees that the financial rewards that can be gained through the control of mineral endowments acts as a major incentive for conflict. Thus, the aim of such conflict has no political goals but is purely motivated by potential financial remuneration. The abnormally high rents that are associated with minerals, especially with oil, are, therefore, particularly provocative and help explain the association between mineral endowments and conflict as found in the statistical evidence. The discovery of North Sea oil off Scotland’s coast is an illustrative example of such financial incentives. Following the discovery the Scotland National Party achieved a dramatic boost in popularity largely due to their campaign that sought to reap the benefits of Scotland’s oil for Scotland only. Though not a case that led to conflict it is nonetheless useful in showing the impact of perceived financial gains on peoples’ behaviour. Le Billon (2001 in Ross 2004) draws some important further distinctions. He sees that the mineral’s concentration, its distance from the capital and the extent of capital required for its extraction all influence the onset of conflict. As such he considers that point resources distant from the capital will be the most likely to induce secessionary conflict. The logic is that as point resources require capital-intensive extraction processes the benefits are accrued by the state rather than the population in the asset’s vicinity. It is only through secession that the people native to the mineral abundant region can fully realise the benefits of the resource and this provides a strong pretext for conflict. Essentially, rent seeking is economic opportunism and the greater the potential remuneration the greater the risk of conflict.

The theory of ‘Dutch disease’ explains how the discovery of a new mineral deposit may not be the absolute benefit it is commonly perceived as but may in fact be somewhat of a mixed blessing. The discovery can lead to a surge in mineral exports at such a scale that it causes the currency to appreciate and harms the competitiveness of all other exporters. The effects are further compounded by the fact that this new industry will attract capital and labour at the expense of all other industries. In this way the new industry can be seen to lead to a state of de-industrialisation and to greater resource dependence. By a small modification of this analysis to include an emphasis on political as well as economic factors the theory of rentier state is created. Karl’s (1997) ‘petro-state’ demonstrates the theory well. The extensive rents associated with oil production free states of the need to extract revenue from their domestic societies, thereby, making them less accountable to the societies they govern. Such a situation negates the need of certain institutional and bureaucratic capacities that are of central importance to the long-term development of political institutions and the economy. This leads to weak government that is less able to cope with political instability. Furthermore, the narrow base of government revenue makes it more prone to price fluctuations increasing the chance of it being exposed to pressures that it cannot survive. The cumulative effect of these weaknesses establishes the rentier state as inherently less stable.

However, much of the strength of the rent seeking and rentier state models lie in the existence of a statistically proven relationship between mineral endowments and conflict. In looking at the methodology of the tests it quickly becomes apparent that this foundation upon which the ‘resource curse’ explanations are based has chronic weaknesses. A brief glance back to the statistical evidence cited in support of the ‘resource curse’ is helpful. All seven surveys agreed to the existence of a ‘resource curse’ giving a sense of consensus to the matter. But an analysis of the means by which this agreement was reached shows very little in the way of consensus. Of the seven surveys cited four different variables were used, a range of four states/wars covered, five resources types considered, all over five different time periods. Not one test was based on the same criteria as another. Then consider the effect of different datasets and the comparability of these tests is even more fragile. From 1960 to 1999 it was possible for Fearon and Laitin to count 97 civil wars but Collier and Hoefller only 78 (Ross 2004). These inconsistencies in approach help to highlight the weakness of the empirical evidence that alludes to the existence of a ‘resource curse’. A further inspection of the empirical evidence reveals more problems.

The use of primary commodity exports as a resource measure is a particularly unhelpful form of analysis. As all developing nations are dominated by primary commodity exports it does not have much use as an explanatory tool and indicates only that less developed countries are more at risk. Likewise, the commonly used criterion for a conflict of 1000 battle-deaths presents considerable problems. It gives no indication of differing levels of severity and takes no account of civilian casualties, greatly reducing its usefulness and that of the numerous tests that use it. Lastly, beyond questioning the empirical strength of the ‘resource curse’ it is necessary to note the numerous tests that directly contradict it. Di John (2007) contests the slow growth permutations of oil abundance showing that between 1965 and 1980 the average growth rates of oil exporters as a region was second only to East Asia and in 1980 to 1998, though there was a marked reduction in growth, it was still not below the level of other low income economies. Studies by Elbadawi and Sambinus (2002), Fearon and Laitin (2003), and Collier, Hoeffler and Soderbom (2004) all find there to be a weak or no effect in the relationship between primary commodity exports and the onset of war.[2] Lastly, Hegre (2002 in Ross 2004) finds no significant effect between mineral exports as a percentage of total exports and war onset.

As such the empirical evidence underlining the ‘resource curse’ is inadequate and this poses serious question of the applicability of the explanatory rent seeking and rentier state models. In looking at the examples of Angola, Indonesia and Zaire it is possible to show conclusively the shortcomings of resource based explanations and prove that politics is the root cause of violent conflict.

“Angola is in many ways the ultimate primary commodity economy. Its economy was based first on slaves, then on coffee, then on diamonds and, increasingly over the past thirty years, on oil” (Cramer 2007). It should, therefore, be a very good example for the rent seeking argument. Indeed, if you were to take a snapshot at random of Angola during its conflict the rent seeking argument may appear as true, or at the least as strong. Oil and diamonds were absolutely central to the conflict and a state of resource dependence was in effect for the duration. Furthermore, UNITA’s position of strategic strength in the diamond region fits in well with Le Billon’s (2001 in Ross 2004) finding that there is a relationship between conflict and distance from the capital. However, it is important not to be distracted by the prominent role of minerals throughout the conflict and we must look back to the actual beginning of the war to fully establish its true cause. What we can see shows clearly the inadequacies of a resource-based explanation. The beginning of the war in the 1960s had little to do with oil or diamonds and must be understood in the context of politics. Cramer notes how the liberation war emerged in a wide range of political dimensions: “urban tension, rural repression and rebellion…and nationalist politics”(Cramer 2007). It was the actual conflict that led to the resource dependent position and not the other way around. At independence manufacturing accounted for 25% of GDP and had been growing at the fastest rate in Africa between 1960 and 1973 at an average real rate of 11% per annum (Cramer 2007). Thus, the heavy resource dependence throughout most of the conflict is misleading and distracts from the political cause. Even to the end the conflict in Angola should be seen as being politically motivated. Le Billon (2002 in Lujala, Gleditsch and Gilmore 2005) notes how war can reach a stage where one side is no longer in pursuit of actual victory but is instead content to generate revenue from resources under their control. However, the swift end to the war after the killing of Savimbi shows that politics was the driving force behind the conflict to the very end and that the use of resources was simply a means to achieve this end.

The situation in Angola was one in which the resources controlled by the UNITA rebels was in the form of alluvial or secondary diamonds. As such it required very little in the way of capital for its extraction and proved a viable means from which to fund the war effort. However, alluvial diamonds are unique amongst minerals in this respect and it is important to consider the effect of point resources on the rent seeking model. As the argument states the greater the resource rent the greater the incentive it creates for a challenge to control it. Thus, oil is typically attributed as being the most conflict inducing mineral. Yet such an assumption has major flaws, the most significant of which is the collective action problem encountered due to the time consistency error. Point resources require intensive capital processes for their extraction and, therefore, have a long time lag between the initial instigation of conflict and the stage where the actual derivation of wealth occurs. “Resource rents may increase the desire of nascent rebels to capture the state, but not their ability to do so- or even to initiate a civil war. Rebel groups often face credit constraints: just because the value of victory is high it doesn’t mean rebels can raise money for arms and men” (Ross 1999). This constraint helps demonstrate the inadequacy of rent seeking as an explanation of conflict onset. In practice any such motives can only be realised in the context of a larger political issue in which the collective action problem has already been overcome.

Rentier state theory views mineral endowments as having a direct effect on the institutional capacity of the state, acting to weaken its level of governance and leaving it more prone to conflict. In doing so it belittles the role of politics but through the use of Indonesia and Zaire as case study examples it can be shown beyond doubt that politics is the key determinant of a state’s direction.

From 1961 to 1978 the percentage share of Zaire’s top three exports (all minerals and predominately copper) grew from 52% to 91% (Dunning 2005). This undoubtedly represents a major dependence on resources and many ‘resource curse’ advocates use it as an example for the rentier state theory. They may well find significant correlation between their theories and Zaire but in considering the political context of the situation the idea of resources as the cause is proved incorrect. Olsen’s (2000) “stationary bandit” is helpful here. The “stationary bandit” has an incentive for economic growth as it is from this that he will be able to maximise benefits. Though not implying that Mobutu was a bandit it can be seen that he too would be a maximiser in the long-run and endeavour to encourage economic growth. However, Dunning (2005) adds an important twist to this analysis and notes the importance of trade-offs between economic growth and political stability. “According to a number of analysts, Mobutu believed that investments in economic infrastructure, including those as simple as maintaining the network of roads left by the Belgium colonials, would pose a threat to his hold on political power by facilitating collective mobilisation against his regime” (Dunning 2005). Such a view may strike as paranoid but his regime was fragile from the outset and characterised by high levels of societal opposition. It was against this backdrop and with the goal of continued power that Mobutu formed policies that were instrumental to the greater resource dependence Zaire descended into. The focus of production into the hands of his allies and the emphasis on the more easily controlled resource sector led to de-diversification. To simply use the fact that at the official outbreak of conflict the country was resource dependent and present this as evidence of a ‘resource curse’ would be to completely miss the point. The root of conflict in Zaire can be traced back to Mobutu’s early years and it was the political choices he took to avert political crisis that led to the resource dependent position and not the other way around.

A comparison of the fate of Indonesia to Zaire helps to show further the need for a political understanding of violent conflict. Like Mobutu, Suharto’s rule was heavily characterised by high levels of opposition and, again like Mobutu, he had reservations of the political feasibility of economic diversification. However, unlike Mobutu he was able to devise a method of growth that was advantageous to both him and the Indonesian state. Through the promotion of Sino-Indonesian entrepreneurs he was able to encourage the growth of a non-resource sector but in a manner that did not represent a political threat because their ethnic minority status meant they lacked the popular support of normal Indonesians (Dunning 2005). In this way it was possible to diversify the economy and from 1966 to 1997 real GDP grew at an annual rate of 5%, and manufacturing as a percentage of GDP rose from 10% to 25% (Dunning 2005). The difference between the fates of Zaire and Indonesia cannot be understood through the lens of the ‘resource curse’. Both nations began as resource dependent yet Indonesia was able to diversify rapidly. Politics was the dominant driving force in both but Suharto, unlike Mobutu, was fortunate to have a viable means around his political challenge.

That the discussion has sought to highlight the prevalence of politics as the root cause of violent conflict does not mean that mineral endowments have an insignificant role in conflict generally. As seen in Angola diamonds were of central importance in UNITA’s war effort (Cilliers and Dietrich 2000). After the Cold War patronage ended it was their control of the diamond fields that allowed for the continuation of fighting. As such it is not only from a better understanding of conflict onset that important policy lessons can be gleamed. Measures such as the Kimberley Process are extremely important and can make a significant impact on conflict alleviation. Yet this should not detract from the fact that politics is the root of violent conflict in less developed countries and it is from a greater understanding of this that more effective preventative measures may emerge. The empirical evidence of the ‘resource curse’ is insufficiently robust to base explanatory models upon. It should come as no surprise, therefore, that the rent seeking and rentier state models provide an entirely inadequate analysis. The case studies of Angola, Indonesia and Zaire are highly illustrative of the fact that politics is the root determinant of conflict. The situation currently unfolding in Zimbabwe is a useful note on which to conclude. The country is increasingly resource dependent and if civil war were to erupt tomorrow it could fit neatly into the ‘resource curse’ model. However, to see this as an explanation of the conflict would be completely wrong. The current situation in Zimbabwe undoubtedly derives from politics. As recently as on the 21^st November 2007 a new bill was passed giving the state a controlling stake of all diamond and platinum mines, compounding to the country’s already astronomical problems. As can be seen elsewhere and quite possibly now in Zimbabwe, the often very long prelude to conflict during which resource dependence deepens with the collapse of the manufacturing industry gives a misleading picture of the event when it finally erupts. As such resource explanations fail to pay due attention to the political factors that lead to the resource dependent position in the first place and in doing so severely misinterpret the root cause of violent conflict in less developed countries.

References

-J. Cilliers and C. Dietrich (eds.), (2000), Angola’s War Economy: The Role of Diamonds and Oil, (South Africa: Institute for Security Studies).

-P. Collier, (2000) “Doing Well Out of War: An Economic Perspective”, in Berdal, Mats and D. Malone, Greed and Grievance: Economic Agendas in Civil Wars, (London: Boulder).

-C. Cramer, (2007) Civil War is No a Stupid Thing: Accounting for Violence in Developing Countries, (London: Hurst and Company).

-J. Di John, (2007) “Oil Abundance and Violent Political Conflict: A Critical Assessment”, Journal of Development Studies, Vol.43, No.6.

-T. Dunning, (2005) “Resource Dependence, Economic Performance, and Political Stability”, Journal of Conflict Resolution, Vol.49, No.4.

-T. Karl, (1997), The Paradox of Plenty: Oil Booms and Petro-States, (Berkeley: University of California Press).

– P. Lujala, N. Gleditsch and E. Gilmore, (2005) “A Diamond Curse?: Civil War and a Lootable Resource”, Journal of Conflict Resolution, Vol.49, No.4.

-M. Olson, (2000) “Dictatorship, Democracy, and Development”, in M. Olson and S. Kahkonen, A Not-So-Dismal Science: A Broader View of Economics and Societies, (Warminster: Footnote Graphics).

-J. Raath, (2007) “Mugabe grabs platinum and diamonds”, The Times Newspaper 21/11/2007.

-M. Ross, (1999) “The Political Economy of the Resource Curse”, World Politics, Vol. 51, No. 2.

-M. Ross, (2004) “What Do We Know about Natural Resources and Civil War?”, Journal of Peace Research, Vol.41, No.3.

Further Reading on E-International Relations

• Is Nationalism Inherently Violent?
• International Relations Is Not Post Postcolonialism in the Twenty-First Century
• South Korea Is Not In Democratic Backslide (Yet)
• Is Decolonisation Always a Violent Phenomenon?
• Imperialism’s Legacy in the Study of Contemporary Politics: The Case of Hegemonic Stability Theory
• A Spectre of Conflict: Shaping Lebanese Sectarianism and Identity Politics