Natural Resources and Violent Conflict - unu-merit [PDF]

ANNUAL REVIEWS

Further

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Click here for quick links to Annual Reviews content online, including: • Other articles in this volume • Top cited articles • Top downloaded articles • Our comprehensive search

Natural Resources and Violent Conflict Eleonora Nillesen1 and Erwin Bulte2 1

UNU-MERIT, 6211 TC Maastricht, Netherlands; email: [email protected]

2

Development Economics Group, Wageningen University, 6700 KN Wageningen, Netherlands; email: [email protected]

Annu. Rev. Resour. Econ. 2014. 6:69–83

Keywords

The Annual Review of Resource Economics is online at resource.annualreviews.org

civil war, greed, grievances, resource scarcity, resource abundance

This article’s doi: 10.1146/annurev-resource-091912-151910

Abstract

Copyright © 2014 by Annual Reviews. All rights reserved JEL codes: D74, O11, O12, Q34

We discuss the literature on natural resources and violent conflict. The theoretical literature is rich and ambiguous, and the empirical literature is equally multifaceted. Theory predicts that resource booms or discoveries may attenuate or accentuate the risk of conflict, depending on various factors. Regression analyses also produce mixed signals and point to a plethora of mechanisms linking resources to conflict. The empirical literature is gradually evolving from cross-country conflict models to micro-level analyses, explaining variation in local intensity of conflict. This transition has resulted in more credible identification strategies and in an enhanced understanding of the complex relation between resources and conflict.

69

1. INTRODUCTION

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Internal armed conflict: a conflict between a government and a nongovernmental party, with no interference from other countries (Gleditsch et al. 2002) Armed conflict: a contested incompatibility that concerns government and/or territory where the use of armed force between two parties, of which at least one is the government of a state, results in at least 25 battle-related deaths (Gleditsch et al. 2002)

The relation between natural resource wealth and the onset or duration of conflict remains a much-debated topic in economics. One group of scholars associates resource scarcity with conflict. According to this neo-Malthusian perspective, dwindling resources imply intensified incentives for conflict. The opposite perspective holds that an abundance of resources invites conflict. Natural resource stocks may represent a prize worth fighting for, and trading resources thus gained may be used to finance (ongoing) conflict. In addition, mismanagement or theft of resource rents, or environmental damages caused by large-scale resource extraction, may fuel grievances. Yet a third perspective emphasizes that the relation between resources and conflict is complex; varies for different types of resources; and is conditional on many other factors, including the institutional context and political regimes. Which perspective is correct? In this review, we survey the literature on resources and conflict. This literature is potentially very broad, not in the least because of the variety of resources that can be considered. Although few readers will doubt that conflicts over oil and diamonds are part and parcel of the resource-conflict nexus, some would argue that land degradation due to overgrazing or water scarcity due to weather shocks also fits in this category. A rapidly growing literature now considers the relation between conflict and environmental factors such as increasing temperatures and declining (and more erratic) rainfall patterns. For example, droughts in Africa may be associated with reduced opportunity costs of conflict for herders and farmers relying on rain-fed agriculture. However, droughts may also increase livestock mortality, lowering incentives to engage in cattle rustling. A recent issue of the Journal of Peace Research (Gleditsch 2012) is devoted to the relation between weather shocks (climate change induced or otherwise) and arrives at the conclusion that the evidence supporting the hypothesis that droughts or temperature shocks cause conflict is mixed, at best. In our survey, we do not focus on weather shocks and instead consider exhaustible and renewable resources, for which the dynamics of abundance are governed by (local) management decisions. We also restrict ourselves along another dimension. There are many different types of conflict, ranging from interpersonal conflict to interstate fighting. Some analysts probe the relation between resources and broad measures of social conflict, including mass demonstrations and riots. Other analysts look at the complex relation between resource wealth (typically oil) and international interventions, such as the war between Iraq and Kuwait and its aftermath. However, the great majority of research focuses on the link between resources and civil conflict. This focus is unsurprising, given that civil conflict has been the dominant form of warfare for the past two decades and still is today (Lacina & Gleditsch 2005). Although the incidence of civil conflict has declined since the end of the Cold War, in 2009 the proportion of the world’s countries involved in internal armed conflict was still 15.1% (Hegre et al. 2013).1 Civil conflict is therefore another main focus of our review. This article reviews the literature on conventional natural resources and civil war. We include studies that consider the onset, the duration, the incidence, or the intensity of conflict, and we consider both the micro literature, which is based on household or firm data, and the macro or cross-country literature, which is based on aggregate statistics. As becomes clear below, analyses of the relation between resources and conflict typically face several econometric challenges. Prominent among such challenges is the fact that standard resource measures are not obviously proper exogenous regressors in models explaining variation in conflict. One reason is simply that resource stocks and flows are governed by exploration and extraction decisions, which are

1

The drop in the number of armed conflicts is particularly pronounced in Sub-Saharan Africa (see Themnér & Wallensteen 2011).

70

Nillesen




Bulte

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

affected by (the shadow of) conflict in various ways. For example, the threat of conflict may induce firms or incumbent leaders uncertain about their political future to increase current extraction rates (van der Ploeg & Rohner 2012). It may also attenuate incentives for firms to invest in extraction capacity, which would have the opposite effect of lowering extraction rates. Indeed, the shadow of conflict may distort the optimal structure of ownership; perhaps extraction by public firms (nationalization) is more attractive than extraction by technically superior private firms, if the risk of conflict or the threat of reneging on contracts introduces holdup problems for firms (van der Ploeg & Rohner 2012). In addition to concerns about simultaneity bias or reverse causality, both resource extraction (or exploration) and conflict may be affected by a range of other variables (e.g., political regimes, level of economic development) so that failing to control for these factors may introduce omitted variable bias into regression estimates. In light of obvious limitations to conducting randomized experiments in the domain of conflict, researchers typically employ econometric techniques such as instrumental variable methods to tease out causal effects. As in other fields of economics, however, the appropriateness of such instruments is often the subject of intense debate. Disputes about econometric issues contribute to the fragmented nature of the literature on resources and conflict and explain some of the contradictory evidence, as discussed below. This article is organized as follows. Section 2 sketches the theoretical framework that has guided much of the empirical work in this domain. We introduce the simple static contest function that is at the heart of many conflict models, albeit often implicitly, and then discuss more elaborate theories extending the basic model. Section 3 presents the cross-country evidence, which includes relatively simple ordinary least squares (OLS) models linking conflict to resource dependence and more elaborate and credible identification strategies based on exogenous variation in quantities or prices. Section 4 presents evidence at the micro or household (or firm) level on the relation between resources and conflict. This subliterature focuses on the incentives for individuals to engage in conflict or to pursue alternative livelihood strategies. Finally, in Section 5, in light of all the evidence, we critically discuss where this literature is going.

2. RESOURCES AND CONFLICT: THEORETICAL CONSIDERATIONS One may view conflict as a consequence of bargaining failure. After all, because conflict is costly (absorbing production factors with positive opportunity costs and possibly inviting destruction and damages), it would presumably be better to peacefully divide the contested asset. However, bargaining may break down for many reasons, including commitment problems, indivisibilities, agency problems, and asymmetric information (e.g., Fearon 1995). If so, engaging in violent conflict emerges as a second-best possibility. A large body of evidence suggests that the propensity to engage in such conflict, the intensity of the conflict, and the outcomes of the conflict are not invariant with respect to stocks of natural resources. The basic intuition underlying the idea that an abundance of natural resources matters for conflict is best illustrated with a model containing a simple contest function (see also Hirshleifer 1995). Assume that there are two groups in society, one of which may be the government; each group consists of Ei, i ¼ 1, 2 members. Each group has to decide how many of its members to allocate to a production task, Wi, and how many of its members to allocate to contesting a resource base or flow of resource rents, Fi: Ei ¼ Wi þ Fi .

ð1Þ

Assume for simplicity that the returns to productive effort are given by a simple linear production function pW i ¼ aWi (where a is a productivity parameter) and that the returns to contest effort are www.annualreviews.org




Natural Resources and Violent Conflict

71

given by pFi ¼ pðFi , Fj ÞR. In this expression, p(×) represents a contest success function (CSF), j
i, and R represents the (exogenous) value of the resource that is contested. A standard CSF reads as follows: p1 ¼

F1m

F1m , þ F2m

ð2Þ

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

where pi represents the share of the resource that is grabbed by group i (or, alternatively, the probability that group i wins the contest and grabs everything) so that p2 ¼ (1  p1). The parameter m captures the extent to which an advantage on the battlefield translates into greater material payoffs. For the extreme case of m ¼ 0, each group obtains half of the resource value, and for m ¼ 1, the p(×) function approaches a winner-takes-all step function. This parameter may represent a measure of resource pointiness (as in Wick & Bulte 2006)2 or a measure of institutional quality that reflects the security of property rights (as in Butler & Gates 2012). With total group payoffs F defined as pi ¼ pW i þ p i , the optimal allocation of group members across the two activities, in an interior equilibrium, is given by F1 ¼ F2 ¼

mR . 4a

ð3Þ

In expected value terms, the resource is equally divided between the competing groups. The intensity of conflict, measured as the amount of effort allocated to contesting the resource, is increasing in the value of the resource: ∂Fi =∂R ¼ m=4a > 0: Hence, a positive shock to the resource measure—more abundant natural resources or higher resource prices—causes income-maximizing groups in society to rationally allocate more effort to conflict. Equation 3 predicts that resource income and production income have opposite effects on the intensity of conflict. If the productivity of labor allocated to production goes up, the opportunity costs of conflict effort increase so that optimally less effort is allocated to the contest: ∂Fi =∂a ¼ mR=4a2 < 0: This very simple model can be extended in many directions. An obvious shortcoming of the basic conflict model is that it does not allow for side payments or bribes or otherwise include domestic policies and interventions that could alleviate grievances of various social groups. For example, a sizable resource curse literature emphasizes the fact that resource rents are often (mis)used as tools of political patronage. It is well documented that many so-called rentier states, in which public coffers depend on the inflow of resource rents, implement large-scale redistributive programs or policies to galvanize support from the opposition. For example, offering inflated yet unproductive public sector employment might be one approach to avoid conflict and to stabilize the incumbent ruler’s regime. This issue is explored in Bjorvatn & Naghavi (2011), who study the case in which the state wishes to stabilize its rule by transferring wealth to the opposition but cannot make binding commitments regarding future transfer programs. They find an inverted-Ushaped relation between resource rents and the risk of conflict. Although an increase in resource rents increases incentives for rent seeking and conflict by the opposition, it also provides credibility to extensive political patronage programs. For sufficiently large resource rent flows, the latter effect dominates (but see also Aslaksen & Torvik 2006). The assumption that the inflow of resource rents can create extensive patronage networks suggests that the quality of public institutions may be best viewed as an endogenous variable.

2

Pointiness reflects the degree to which the resource is concentrated (clustered) in space—easy to conquer and defend, or diffuse and spread across the landscape.

72

Nillesen




Bulte

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Recognizing this point, Hodler (2006) develops a model in which property rights to produced output is endogenous and varies with the intensity of fighting (an effect external to warring factions). Specifically, when warring groups allocate more effort to fighting, property rights security deteriorates, and groups retain a smaller fraction of the output they produce (pW i ¼ aWi ) for themselves; the rest of the output enters the contestable pool, where it may be grabbed by everybody. Weaker institutions obviously erode incentives for production, so conflict over resource rents may lower aggregate income. Because theory predicts fighting effort will increase in the degree of fractionalization, resources will invite conflict (and slow growth) in countries with more different ethnic groups. Hodler’s data support this prediction. Caselli & Coleman (2013) also pursue the perspective that ethnic fractionalization may be a salient factor in the resource-conflict nexus. Their innovation is the observation that group membership may not be permanent or even be recognizable. If members of a losing group may pass as members of the winning group, then “losers” will mimic the “winners,” and the spoils from the conflict will have to be shared with an expanding group of individuals. Because this sharing dilutes the net gains from conflict, the incentives to engage in conflict are eroded. Because especially ethnic cleavages between rival social groups may allow for easy identification of group members— compared with, say, cleavages based on religion—the ethnic composition of societies may be an important mediating factor that explains the frequency and intensity of conflicts over resource rents. However, uncertainties remain about how and when ethnicity matters. For example, Montalvo & Reynal-Querol (2005) point to ethnic polarization—rather than to fragmentation— as a determinant of conflict. Adding to the interplay between ethnicity and resources, the geographical location of resource stocks (and how these deposit-rich areas map onto the homelands of different ethnic groups) may also be relevant for conflict. Intuitively, if an ethnic group dominates in a region that is particularly rich in resources (relative to the rest of the country), and if the group’s chances of successful secession are greater than its chances of winning an all-out conflict over control of the state, then two relevant threat points for war emerge, and bargaining over peaceful sharing of resource rents breaks down. Conflict is inevitable. Morelli & Rohner (2010), who study this case and others, propose a unified framework to study the interplay between ethnicity, multiple types of conflict (i.e., secessionist and centrist), and the spatial distribution (clustering) of natural resources.3 In addition to being relevant as a factor shaping domestic violence and secession tendencies, the geographical location of resource stocks may also matter for international conflict. Caselli et al. (2013) demonstrate that the location of resource deposits vis-à-vis international borders affects invasion decisions. Of course, extending the basic model in other directions is possible. In an interesting contribution, Dal Bó & Dal Bó (2011) incorporate a richer characterization of the productive side of the economy. They replace the simple one-sector model with a model consisting of two production sectors. Labor flows freely between three sectors––the two production sectors and the conflict sector (in which members contest rents)—in response to wage differentials. One production sector is capital intensive, and the other one is labor intensive. Using a simple trade model, the authors analyze the consequences of economic shocks to the productive sectors on the intensity of predation, as determined by the relative returns to labor allocated to conflict. Their main result is that

3

In addition to impacting the onset of interstate war and different types of domestic war (e.g., secession and centrist conflict), natural resources may also affect other types of conflict. Esteban et al. (2010) study strategic mass killings of civilians—the deliberate reduction in the group size of the opponent—focusing on the material interests of potential perpetrators. In the context of proportional sharing rules, mass killings enable groups to obtain a larger share of (future) resource rents. This gain is balanced against potential losses, such as the reduction in (taxable) output resulting from the loss of labor.

www.annualreviews.org




Natural Resources and Violent Conflict

73

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

positive shocks to the labor-intensive sector diminish conflict as they raise the return to labor in production, drawing effort out of the appropriation sector. This result is a direct consequence of the Stolper-Samuelson theorem. In contrast, a positive shock to the capital-intensive sector lowers wages and intensifies predation. The implication is that, depending on the (relative) labor intensity of the resource sector, positive shocks may accentuate or attenuate conflict. Acemoglu et al. (2010) introduce additional dimensions to the problem. They focus on the case of a rebel group contesting an elite-dominated government. This government can invest in a strong military that can quickly defeat the rebels with certainty, but the government also has to consider the risk that a strong military may wage a coup and oust the elite. This scenario presents a so-called political moral hazard problem: Rent-grabbing governments face the choice between (a) a weak army and persistent war and (b) a strong army and the risk of a coup. Two equilibrium outcomes are possible. When elite rents are not too badly affected by ongoing conflict (for example, because conflict is located in a remote corner of the territory without important stocks of natural resources), then tolerating persistent conflict may be a strategy yielding the greatest elite payoffs. In contrast, when conflict is costly to the elite, the elite will seek to crush the rebels and to appease the army. The strategy, then, will be to overinvest in the army by paying for more military material and soldiers than necessary to safeguard the monopoly of violence and by possibly allowing for some army interference in government business. This strategy is an effort to commit to not reforming the military after the rebels’ threat has been neutralized, minimizing the risk of a coup. This focus on the persistence of wars explicitly introduces a temporal dimension to the basic conflict model. Several other papers examine the dynamics of conflict, and not surprisingly, such analysis sets the stage for an even richer set of results. For example, Grossman & Mendoza (2003) examine the implications of a model that includes a positive intertemporal effect of current consumption on the probability of survival, which in turn affects the discounted value of future consumption (and leisure). The authors assume that this survival effect becomes stronger as the resource becomes scarcer, so rational agents will fight harder as scarcity increases. The authors’ main result confirms Hirshleifer’s (1995) hypothesis that, “as Malthusian pressures depress per capita income, it comes to a choice between fighting and starving,” and essentially turns the main prediction of the standard contest model—that more effort is allocated to conflict as the stakes increase—upside down. Acemoglu et al. (2012) obtain a related result in the context of an international context with two countries: The first is resource rich, and the second is resource poor. The first country exports resources, drawing down its in situ resource stock over time. If demand for the resource is inelastic, the value of the remaining resource stock gradually increases, which implies that the incentives for the resource-poor country to invest in arms and to invade the resource-rich country—gaining control over the resource base—also rise over time. If competitive firms extract resources in the resource-rich country, such firms will not internalize the impact of extraction on war incentives of the other country. If there is inelastic demand and if the resource-poor country has sufficient military strength, this external effect makes conflict inevitable. In response, the government of the resource-rich country may try to internalize the conflict risk and to regulate prices or quantities in an effort to slow down extraction. However, Acemoglu et al. (2012) demonstrate that, in the presence of limited commitment, such a strategy may backfire and actually invite conflict. The resource-poor country has an incentive to invest in arms in an effort to secure beneficial terms of trade, but the discounted costs associated with such a prolonged investment strategy may exceed the short-term costs of conflict. The idea that scarcity increases over time is logical in the context of exhaustible resources (even if exploration and technological progress may mitigate scarcity). More complex dynamics may emerge when renewable resources such as fish and trees are contested and extracted. Renewable 74

Nillesen




Bulte

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

resources have the potential to replenish naturally, and resource growth is typically represented by a logistic model, i.e., an inverted-U-shaped relation between stock size and growth. Reuveny & Maxwell (2001) develop a model of two income-maximizing groups, whereby the groups allocate effort to harvesting from a shared renewable resource and to fighting over the total harvest. Assuming that the renewable resource is essential for human reproduction, and that groups respond myopically to profit differentials, the authors demonstrate that a potentially complex and nonlinear dynamic interaction between the resource and the groups may emerge (see also Maxwell & Reuveny 2005). Reuveny et al. (2011) consider an extension in which conflict outcomes contain a stochastic element, the winner of the contest takes everything, and today’s earnings in the contest affect tomorrow’s power or strength on the battlefield. Obviously, the last assumption creates additional incentives for fighting if the warring groups have foresight. Garfinkel & Skaperdas (2000) and Mehlum et al. (2006) also explore this last issue. If warring parties realize that their position in future conflict is affected by the outcome of current conflict, they have an incentive to fight harder now and to gain an extra edge in the future. The implications of such foresight may reverse some of the results of static conflict models or of repeated conflict models based on myopic decision making. In sum, theory provides ample suggestions for empirical work. Although the simplest models of conflict unambiguously predict that greater resource rents translate into more intense conflict (or into an increased probability of conflict), several of the extensions provide important caveats and qualifications. Moreover, most of the theory discussed above is rooted in the so-called greed perspective: Profit-maximizing groups contest resource rents that may be grabbed. The situation becomes even more complex when we consider that resources may also affect conflict via alternative channels. Notably, resource extraction may fuel grievances if, for example, large-scale mining activities destroy the homelands of ethnic groups. Grievances may also be fed if some social groups perceive that resource rents are unfairly distributed. Resource rents may also, via the rentier-state dynamics mentioned above, be used to attenuate grievances and to buy off the opposition. In light of the embarrassing richness of theoretical results, therefore, we must consider the empirical evidence and do so next.

3. CROSS-COUNTRY EMPIRICS Much of the recent interest in the relation between resources and conflict is the result of influential research by Collier (2000) and by Collier & Hoeffler (1998, 2002, 2004). These papers suggest a (nonlinear) relation between a measure of resource dependence (the ratio of primary exports divided by income) and violent conflict. Specifically, an inverted-U-shaped relation exists between resources and conflict, and the risk of civil war is highest when resource dependence equals approximately 35%. The main argument is as follows: Resources motivate greedy rebels to seek a share of the rents, but for sufficiently large resource exports, state revenues will deter rebellion because the state can buy popular support or increase military spending.4 The cross-country work of Collier & Hoeffler (1998, 2002, 2004) attracted many followers who extended or qualified the earlier results. However, the methodological approach also invites comments and concerns. For example, various analysts critique the resource measure used; the

4

Elbadawi & Sambanis (2000) argue that, because of this quadratic relationship, dispersion rather than mean values of resource dependence should be compared. They report much lower (close to 50%) standard deviations for countries within the African continent than for elsewhere, implying that many African states are closer to the peak of natural resource dependence that maximizes the risk of civil conflict.

www.annualreviews.org




Natural Resources and Violent Conflict

75

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

measure does not map squarely onto the resource curse narrative. Primary commodity dependence includes agricultural exports (which most analysts would not classify as resource output) but typically does not include illegal commodity exports (such as blood diamonds) that are typically implicated in narratives of the resource-conflict curse. Also, the inclusion of reexports may create misleading measures of dependence (Humphreys 2005). Moreover, dependency ratios are arguably endogenous to a country’s conflict risk, which invites concerns about reverse causality. As mentioned above, resources may be overextracted if leaders anticipate political turmoil, increasing resource dependence. Similarly, resource dependence may go up if the resource sector is the only sector operational in times of conflict (as more footloose industries leave the country and new investments in other sectors halt). Brunnschweiler & Bulte (2008) empirically illustrate the importance of such reverse-causality issues. When the resource dependence variable is instrumented, the correlation between resource dependence and conflict weakens or disappears. As suspected, resource dependence enters endogenously in most conflict models, and its dynamics are shaped by weak institutions and by the history of conflict. Using the arguably more exogenous measure of resource abundance (defined as the “discounted value of expected resource rents for a future period of 20 to 25 years, calculated in per capita US dollars for 1994”) changes the conventional outcome. Higher levels of resource abundance are associated with a modest decrease in conflict. The mechanism may be an income effect associated with an inflow of rents. Significant correlations between resource and conflict measures may also result from data artifacts. The paucity of annual cross-country statistics often forces researchers to be creative and to either focus on long average time periods (a common approach is to use five years, although some studies use three years) or interpolate missing values. Fearon (2005) replicates Collier & Hoeffler’s (2002) model but uses country years instead of five-year intervals as the unit of analysis. He finds that the conflict-inducing effect disappears in all but one model. Other qualifications include (a) the use of logs (rather than a quadratic measure) to capture the nonlinearity in the data and (b) a distinction between different types of natural resources. Oil exports seem more robustly associated with conflict onset than are exports of other commodities (see also Lei & Michaels 2011). Dependence on oil exports is often associated with weak states due to the rentier effect (eroding incentives to invest in a functional bureaucracy for domestic taxation). Some evidence suggests that oil exports are associated with low state capacity (see also Fearon & Laitin 2003). As discussed above, and as emphasized by Collier & Hoeffler (2004), resource revenues may also increase state capacity, as these revenues may be used to defend the state, to increase control, and to buy support. Snyder & Bhavnani (2005) explicitly focus on states and their capacity to deter conflict. In a special issue of the Journal of Conflict Resolution on the role of natural resources in civil conflict, they theorize that whether resource-rich countries experience conflict depends on what they refer to as the resource profile of the economy––the quality of economic institutions and the governments’ spending of revenues. The central idea is that if governments generate sufficient revenues, through either direct taxation of resources or other income-generating activities, they will be able to deter rebellion. In addition, the allocation of resource revenues across expenditure categories, such as investments in the quality of the bureaucracy, coercive capacity, and social welfare, may influence state strength as well as citizens’ grievances. Hence, spending patterns indirectly impact a country’s risk to plunge into war (see also Basedau & Lay 2009). Snyder & Bhavnani’s argument is supported by case studies of Sierra Leone, Ghana, and Guinea. In a related vein, Olsson (2007) presents a link between diamond rents, conflict, and growth on the basis of theoretical predictions from a two-period model in which rebel activity depends on the quality of institutions, on formal sector productivity, and on appropriable rents. Cross-country results demonstrate a negative effect of natural resources on a country’s growth rate if institutions are weak (see also Hamilton & Clemens 1999 on negative savings rates as often documented in 76

Nillesen




Bulte

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

mineral-rich Sub-Saharan African countries). Dunning (2005) also focuses on state capacity and develops a two-period game-theoretic model in which elites can decide to invest in public goods and nonelites can revolt to take power. Three case studies are used to illustrate equilibrium predictions: (a) no revolt, (b) no investment and continuous revolt, and (c) revolt when the resource price is sufficiently low. The case studies illustrate how variation in economic and political outcomes is explained by variation in elite incentives to contribute to public goods, diversifying the economy. Variation in the latter is arguably caused by volatility in resource revenues, by the extent of societal opposition, and by prior development of nonresource sectors. Collier & Hoeffler (2005) respond to the challenges to their resource dependence measure by empirically testing whether resource rents (rather than resource dependence) affect conflict risk. Substituting primary commodity dependence with a rent measure, they continue to find the quadratic relationship to hold (albeit less significantly than before). Moreover, the dependence variable trumps the rent variable when both are included in the same regression. Collier & Hoeffler conclude that the rent effect exists but argue that it is indirect: Elites steal resource rents rather than investing them in public goods that facilitate long-term growth. Stagnant growth rates in turn increase the risk of conflict. Collier & Hoeffler consider four cases in which looting may be the preferred strategy: (a) short time horizons, (b) small ethnic ruling group, (c) low opportunity costs of income growth, and (d) democratic elections degenerating into patronage politics (which reduces the need for accountability; see also Collier et al. 2009). Brückner & Ciccone (2010) test the opportunity cost mechanism more directly by looking at international commodity prices and export demand as determinants of war risk. They test whether a drop in countries’ main export commodities hampers economic development and increases the risk of the onset of civil war. For this purpose, they construct a single, country- and year-specific commodity price index based on a mix of renewable and nonrenewable natural resources and agricultural commodities. Using a data set of 39 Sub-Saharan African countries, they find that a 20% decrease in the commodity price index increases the probability of war onset by 2.8 percentage points (given an average civil war onset probability during 1980–2006 of 2.8%). Their additional analysis supports the hypothesis that slow growth may be a channel explaining conflict. Using commodity price growth and OECD growth as instruments for Sub-Saharan Africa GDP per capita growth, Brückner & Ciccone report a strong, negative impact of growth on civil war onset. However, as mentioned above, lumping different types of commodities together into a single price index (or in a single measure of resource wealth) is not an uncontested method. For example, Ross (2004b) reviews 14 cross-country econometric studies and concludes that oil exports are associated with the onset of conflict, whereas agricultural commodities or broader measures of primary commodities are not. By contrast, Humphreys (2005) reports that agriculture-led countries reflect lower levels of domestic trade, which puts them at a higher level of conflict risk (independent of their resources). Others, like Lujala et al. (2005), also demonstrate that the type of resource matters. Distinguishing between primary (nonlootable) diamonds and secondary (lootable) diamonds, Lujala et al. find opposing effects. Although primary diamond production, measured by a dummy variable that is 1 for the first year in which diamonds were discovered and every year thereafter, decreases the onset and incidence of ethnic wars by not less than 80%, the effect of secondary diamonds dwarfs that of primary diamonds. The production of secondary diamonds increases the probability of engaging in ethnic wars by 200%. Again, these results appear to be sensitive to the measure of resource wealth––no effects are found when deposits are used instead of production, so the presence of diamonds is insufficient to initiate a war. Instead, the study suggests that modes of resource extraction matter. This conclusion is consistent with that of Ross (2004a), who reports case study evidence suggesting that easy-to-loot commodities such as gemstones, drugs, and timber www.annualreviews.org




Natural Resources and Violent Conflict

77

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

may affect conflict duration. The evidence regarding the effect on duration is, however, ambiguous. The length of conflicts may depend on incentives for fighting parties to sign a peace agreement and on the ability to fund ongoing fighting. The study of Ross also reveals other mechanisms by which resources may affect conflict—lootable resources may give rise to so-called booty futures, in which warring factions trade future exploitation rights of resources for weapons. However, this effect is contested by Humphreys (2005), who argues that “weak states and grievance seem to matter much more than the idea that rebels perceive the state as a prize or booty future.” More recently, Cotet & Tsui (2013) also find little support for the idea that “greedy forward looking rebels [can] sell future exploitation rights.” As discussed in the previous section, the spatial distribution of resources within a country also determines whether resource revenues are available to fighting parties to start or sustain fighting (see also Le Billon 2001). Earlier research (e.g., Ross 2004a,b; Collier & Hoeffler 2005; Ross 2006) suggests that resources play a role in secessionist wars, but these studies lack adequate data to test the influence of a within-country location(s) of natural resources in conflict. Lujala (2009) uses a fine-grained global data set that includes spatial information about primary and secondary diamond deposits (see Gilmore et al. 2005). Considering three types of activities (gemstone mining, hydrocarbon extraction, and drug cultivation) in relation to conflict intensity and geographic location, Lujala finds a positive correlation between gemstone mining in the conflict zone and combat-related mortality. Hydrocarbon extraction is also correlated with more intense and longer conflicts, but similar results are not obtained for drug cultivation. A key implication is that the relation between resource measures and conflict may be obscured when the analyst uses aggregate data––the above-mentioned results are not found when country-level data on conflict are used. Moreover, conflict mortality declines if production takes place outside of conflict zones (or offshore). This finding supports the hypothesis that elites may be uninterested in investing in a strong military (and risking a coup) when the conflict does not threaten the flow of resource revenues. Buhaug & Rød (2006) further emphasize the importance of a resource’s spatial distribution. Using 100 3 100–km grid cells as the unit of analysis (in contrast to country years) and distance to secondary diamond and petroleum deposits as resource variables, these researchers find that the influence of local (resource) conditions varies with the type of conflict. In particular, a (weak) negative association exists between proximity to diamonds (petroleum) and conflict. But diamonds seem to have the opposite effect in conflicts over state control, consistent with observations from recent conflicts in West Africa (e.g., Liberia, Sierra Leone) and Central Africa (e.g., Angola, Democratic Republic of Congo). A recent contribution is that of Bazzi & Blattman (2013), who try to overcome many of the issues discussed above. Akin to the approach by Brückner & Ciccone (2010), Bazzi & Blattman study how commodity price shocks affect conflict onset, duration, and intensity. They argue that commodity price shocks affect actors’ motivations to get involved in conflict differently, depending on who is affected most by the shock (i.e., households or the state). They differentiate between commodity classes to discriminate between (a) the opportunity cost mechanism (i.e., higher values of commodities raise the costs of fighting and hence reduce effort allocated to conflict) and (b) the state-as-a-prize effect (i.e., higher commodity prices increase the value of controlling the state, which creates incentives for more conflict effort). Moreover, they use a variety of conflict measures and different model specifications to probe the robustness of their results. In contrast to Brückner & Ciccone, Bazzi & Blattman find little robust evidence supporting either the opportunity cost mechanism or the state-as-a-prize mechanism as a factor explaining the onset of conflict. Results are somewhat more consistent when the cessation of conflict is analyzed. Higher prices across all commodity classes are associated with an increased likelihood of stopping civil 78

Nillesen




Bulte

wars. This finding runs counter to the state-as-a-prize argument. Instead, this finding supports the idea that resource rents increase state capacity, which makes the end of conflict war more likely. Higher resource prices are also associated with less intense conflict (albeit not very robustly); the effect is most pronounced for annual agricultural goods and extractive commodities. This finding is consistent with the opportunity cost mechanism but also supports the state capacity mechanism for the case of extractive commodities. These cross-country findings have inspired scholars to examine the resource-conflict nexus in greater detail at the meso or micro level. The major results of these endeavors are discussed in the next section.

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

4. MICRO EMPIRICAL EVIDENCE The rich and varied cross-country literature on resources and conflict stands in contrast to a relatively small and consistent micro-level literature. Micro-level studies adopt a variety of perspectives and often seek to directly test a theoretical prediction. Several studies are based on powerful identification strategies, adopting a natural or quasi-natural experimental perspective. Such findings tend to be less susceptible to endogeneity bias than most of the cross-country work. Among the first studies, building on existing cross-country approaches, is a pioneering study by Angrist & Kugler (2008). The authors look at the impact of an illegal resource (coca) boom in Colombia on rural incomes and on the perpetuation of the country’s civil conflict. They use the ending of an air bridge in 1993 to transport coca base from Peru and Bolivia to Colombia as a source of exogenous variation in coca prices. In response to the ending of the air bridge, Peru and Bolivia were unable to sell their coca base, and Colombia stepped into this niche and started to produce its own coca. The authors analyze how increased coca production is related to income and conflict by using household data from 1992 to 2000 in a difference-in-differences framework (comparing growing regions with nongrowing regions, before and after the ending of the air bridge). A first result is that higher prices made coca farmers better off, which, secondly, provided new opportunities for informal taxation benefiting guerrillas and paramilitaries. A third result concerns the dynamics of other economic sectors. The results also show a modest gain in income from self-employment and suggest an increase in labor supply of teenage boys. There is no robust effect on the probability to be self-employed or on the (log) hours and wages. Yet, more importantly, violence increased as a result of enhanced coca production, with much stronger effects in rural areas than in urban areas. This study supports the idea that coca provides revenues for insurgents (whose activities are based largely in rural areas) and contributes to perpetuating the conflict. Also focusing on Colombia, Dube & Vargas (2013) demonstrate how changes in different commodity prices affect conflict. This work is most closely related to the theoretical work of Dal Bó & Dal Bó (2011) and investigates the opportunity cost effect as well as the state-as-a-prize effect. As in other studies, the net effect of prices on conflict depends on which effect dominates. If the price of labor-intensive goods increases, the opportunity cost effect is likely to dominate (i.e., conflict becomes less likely), and the reverse is true for capital-intensive goods. Colombia’s two main export commodities are a labor-intensive crop (coffee) and a capital-intensive good (oil). Consistent with the theory, a decrease in coffee prices increases all types of violence. An increase in oil prices causes more paramilitary attacks but does not affect other forms of violence. To further investigate the opportunity cost channel, the authors analyze whether changes in coffee and oil prices affect labor market outcomes differentially across coffee-producing and non-coffeeproducing regions. They find a strong significant effect for coffee, but not for oil. Increasing oil prices are associated with increased government revenues and political kidnappings, which is in www.annualreviews.org




Natural Resources and Violent Conflict

79

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

line with the hypothesis that different commodity shocks affect violence through different channels. Similar resource-specific effects are obtained for other types of resources, including coal, gold, and additional agricultural crops. Dube & Vargas’s study, using more-fine-grained data on conflict-related violence at the municipality level, also replicates the above-mentioned findings by Angrist & Kugler (2008). A different strand of literature approaches the resource-conflict nexus from the perspective of (prospective) rebels. An influential study by Weinstein (2005) argues that variation in economic endowments (including natural resources) attracts different types of rebels––the presence of resources crowds out high-commitment rebels with a long-term focus on alleviating grievances associated with social or ethnic factors and replaces them with opportunistic insurgents focused on consuming short-term material rewards without much commitment to long-term goals. Also, rebel leaders who manage to utilize resources can create rebel groups more quickly than those who recruit on the basis of identities or shared ideologies. Theoretical predictions are supported by case study evidence from Uganda, Mozambique, Sierra Leone, and Eritrea. Humphreys & Weinstein (2008) test these hypotheses more systematically by using data from Sierra Leone within an econometric framework. Using data from a survey of fighters of rebel and government forces (plus a control group of nonfighters), Humphreys & Weinstein find that material (selective) incentives (i.e., money and diamonds) are important predictors of joining fighting groups (either the rebels or the counterinsurgents defending the status quo), although they conclude that motivations to participate in Sierra Leone’s civil war are diverse and that explanations of material rewards may coexist with those of grievances and social pressure. Bellows & Miguel (2009) corroborate the importance of diamonds in the Sierra Leonean conflict, demonstrating that chiefdoms with diamond mines experienced more attacks and battles (see also Olsson & Siba 2013 for weak evidence of more attacks close to fertile soils in Darfur). A third category investigates the reverse issue: how conflict affects the (private) resource sector. A notable example is that of Lind et al. (2009), who study how conflict facilitates the production of an illegal commodity (opium) in Afghanistan. Conflict destroys existing lines of production and weakens law enforcement, giving rise to the development of new illegal opportunities. An interesting point in this paper is the departure from traditional explanations, which are based on (illegal) resources controlled by rebel groups or the government and used for financing military actions. Although Lind and coauthors do not dispute that drug money facilitates and invites conflict, they argue that the reverse story is also true––increased opium production is caused by conflict. The reason is that violence destroys infrastructure, including roads, storage, and irrigation facilities––factors important for wheat production but less important for the alternative occupation of opium production. Moreover, conflict promotes institutional chaos, implying low government control and potentially changing local power structures (e.g., providing local protection) and moral norms conducive to the production of illegal commodities. Using plausibly exogenous variation in Western casualties as a proxy for conflict intensity, Lind et al. report a strong, positive effect of conflict on opium production by using district-level panel data on casualties and opium production from 2001 to 2007. To strengthen their causal claim, the authors use the timing of the planting season––if violence causes opium production, they should find significant effects of conflict on opium production during the planting season (i.e., the period in which the farmer decides whether to grow opium), but not thereafter. Indeed, their finding confirms this hypothesis. The authors report even stronger effects when institutions are weak (i.e., further away from the capital), and they rule out several alternative explanations, such as the presence of Western soldiers being a proxy for easy drug smuggling because they were not allowed to intervene in drug trade, changes in measurement of opium production in some districts as of 2002, and opium price changes over time. 80

Nillesen




Bulte

Guidolin & La Ferrara (2007) also study the consequences of conflict for the private sector and consider how conflict affects businesses in the resource sector. Using an event-study approach, they observe that the death of the Angolan rebel leader Jonas Savimbi and the ending of the civil war decreased returns to stocks for diamond-producing firms active in Angola. There was no effect on stocks for similar companies not operating in Angola. To explain these findings, the authors speculate how, in the postwar era, the government was better able to implement regulation to the detriment of firm profits. In addition, transparency along the diamond value chain increased, which reduced bargaining power for international diamond firms.

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

5. DISCUSSION Explaining the incidence, onset, or duration of civil war has become an important research topic in economics and the political sciences domain. Although many studies focus on the conflict-growth nexus, a parallel literature considers the role of natural resources in starting or perpetuating armed violence. This literature features theoretical, macro, and micro empirical contributions or combinations thereof. In this review, we bring these strands together, discussing regularities but also divergent results. A number of findings stand out. For example, most theoretical and empirical studies focus largely on greed-based explanations of conflict, in which resource rents provide an incentive for fighting. However, insofar as resources cause conflict, the mechanisms linking resources to conflict are ill understood. In particular, grievances may also matter. Additional channels linking resources to conflict are state capacity (for better or worse) and resource-dependent opportunity costs of conflict effort. Few studies can credibly discriminate between these channels. The cross-country evidence suggests that there is no unconditional relation between resource wealth and conflict. If anything, the impact of resource wealth depends on various mediating factors, including income or poverty levels, institutional quality, ethnicity, and geography. However, there is little consensus about the relative importance of each of these factors, and several contributions even call the role of natural resources as a catalyst of conflict into question. The micro literature offers a more consistent perspective. These studies unambiguously demonstrate that natural resource wealth is harmful for peace and stability (we should realize that case studies have been picked nonrandomly––Colombia and Afghanistan feature prominently in this literature). Also, these studies go a long way toward ruling out alternative interpretations and extensively probe the robustness of their findings. In sum, the relation between natural resources and civil conflict has been a lively and fruitful field of research in recent decades, generating many useful insights. Yet uncertainties remain about the conditions under which natural resource wealth accentuates or attenuates civil conflict and about the channels via which such processes happen. Ingenious micro-level studies have pointed the way toward the future research that is necessary to understand these mechanisms.

DISCLOSURE STATEMENT The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review. LITERATURE CITED Acemoglu D, Golosov M, Tsyvinski A, Yared P. 2012. A dynamic theory of resource wars. Q. J. Econ. 127(1):283–331

www.annualreviews.org




Natural Resources and Violent Conflict

81

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Acemoglu D, Vindigni A, Ticchi D. 2010. Persistence of civil wars. J. Eur. Econ. Assoc. 8(2–3):664–76 Angrist JD, Kugler AD. 2008. Rural windfall or a new resource curse? Coca, income, and civil conflict in Colombia. Rev. Econ. Stat. 90(2):191–215 Aslaksen S, Torvik R. 2006. A theory of civil conflict and democracy in rentier states. Scand. J. Econ. 108(4):571–85 Basedau M, Lay J. 2009. Resource curse or rentier peace? The ambiguous effects of oil wealth and oil dependence on violent conflict. J. Peace Res. 46(6):757–76 Bazzi S, Blattman C. 2013. Economic shocks and conflict: evidence from commodity prices. Work. Pap., Sch. Int. Aff., Columbia Univ. Bellows J, Miguel E. 2009. War and local collective action in Sierra Leone. J. Public Econ. 93(11–12):1144–57 Bjorvatn K, Naghavi A. 2011. Rent seeking and regime stability in rentier states. Eur. J. Polit. Econ. 27(4):740–48 Brückner M, Ciccone A. 2010. International commodity prices, growth and the outbreak of civil war in SubSaharan Africa. Econ. J. 120(544):519–34 Brunnschweiler CN, Bulte EH. 2008. Linking natural resources to slow growth and more conflict. Science 320(5876):616–17 Buhaug H, Rød JK. 2006. Local determinants of African civil wars, 1970–2001. Polit. Geogr. 25(3):315–35 Butler CK, Gates S. 2012. African range wars: climate, conflict, and property rights. J. Peace Res. 49(1):23–34 Caselli F, Coleman WJ. 2013. On the theory of ethnic conflict. J. Eur. Econ. Assoc. 11(Suppl. 1):161–92 Caselli F, Morelli M, Rohner D. 2013. The geography of inter-state resource wars. NBER Work. Pap. 18978 Collier P. 2000. Economic Causes of Civil Conflict and Their Implications for Policy. Washington, DC: World Bank Collier P, Hoeffler A. 1998. On economic causes of civil war. Oxf. Econ. Pap. 50(4):563–73 Collier P, Hoeffler A. 2002. On the incidence of civil war in Africa. J. Confl. Res. 46(1):13–28 Collier P, Hoeffler A. 2004. Greed and grievance in civil war. Oxf. Econ. Pap. 56(4):563–95 Collier P, Hoeffler A. 2005. Resource rents, governance, and conflict. J. Confl. Res. 49(4):625–33 Collier P, Hoeffler A, Rohner D. 2009. Beyond greed and grievance: feasibility and civil war. Oxf. Econ. Pap. 61(1):1–27 Cotet AM, Tsui KK. 2013. Oil and conflict: What does the cross country evidence really show? Am. Econ. J Macroecon. 5(1):49–80 Dal Bó E, Dal Bó P. 2011. Workers, warriors, and criminals: social conflict in general equilibrium. J. Eur. Econ. Assoc. 9(4):646–77 Dube O, Vargas J. 2013. Commodity price shocks and civil conflict: evidence from Colombia. Rev. Econ. Stud. 80(4):1384–421 Dunning T. 2005. Resource dependence, economic performance, and political stability. J. Confl. Res. 49(4):451–82 Elbadawi E, Sambanis N. 2000. Why are there so many civil wars in Africa? Understanding and preventing violent conflict. J. Afr. Econ. 9(3):244–69 Esteban J, Morelli M, Rohner D. 2010. Strategic mass killings. Work. Pap. 486, IEW Fearon JD. 1995. Rationalist explanations for war. Int. Organ. 49:379–414 Fearon JD. 2005. Primary commodity exports and civil war. J. Confl. Res. 49(4):483–507 Fearon JD, Laitin DD. 2003. Ethnicity, insurgency, and civil war. Am. Polit. Sci. Rev. 97(1):75–90 Garfinkel MR, Skaperdas S. 2000. Conflict without misperceptions or incomplete information: how the future matters. J. Confl. Res. 44(6):793–807 Gilmore E, Gleditsch NP, Lujala P, Rød JK. 2005. Conflict diamonds: a new dataset. Confl. Manag. Peace 22(3):257–72 Gleditsch NP, ed. 2012. Special Issue: Climate Change and Conflict. J. Peace Res. 49(1) Gleditsch NP, Wallensteen P, Eriksson M, Sollenberg M, Strand H. 2002. Armed conflict 1946–2001: a new dataset. J. Peace Res. 39(5):615–37 Grossman HI, Mendoza J. 2003. Scarcity and appropriative competition. Eur. J. Polit. Econ. 19(4):747–58 Guidolin M, La Ferrara E. 2007. Diamonds are forever, wars are not: Is conflict bad for private firms? Am. Econ. Rev. 97(5):1978–93

82

Nillesen




Bulte

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Hamilton K, Clemens M. 1999. Genuine savings rates in developing countries. World Bank Econ. Rev. 13(2):333–56 Hegre H, Karlsen J, Nygård HM, Strand H, Urdal H. 2013. Predicting armed conflict, 2010–2050. Int. Stud. Q. 57(2):250–70 Hirshleifer J. 1995. Anarchy and its breakdown. J. Polit. Econ. 103(1):26–52 Hodler R. 2006. The curse of natural resources in fractionalized countries. Eur. Econ. Rev. 50(6):1367–86 Humphreys M. 2005. Natural resources, conflict, and conflict resolution: uncovering the mechanisms. J. Confl. Res. 49(4):508–37 Humphreys M, Weinstein JM. 2008. Who fights? The determinants of participation in civil war. Am. J. Polit. Sci. 52(2):436–55 Lacina B, Gleditsch NP. 2005. Monitoring trends in global combat: a new dataset of battle death. Eur. J. Popul. 21(2–3):145–66 Le Billon P. 2001. The political ecology of war: natural resources and armed conflicts. Polit. Geogr. 20(5):561–84 Lei YH, Michaels G. 2011. Do giant oilfield discoveries fuel internal armed conflicts? Work. Pap. 6934, CEPR Lind JT, Moene KO, Willumsen F. 2009. Opium for the masses? Conflict-induced narcotics production in Afghanistan. Work. Pap. 2573, CESifo Lujala P. 2009. Deadly combat over natural resources: gems, petroleum, drugs, and the severity of armed civil conflict. J. Confl. Res. 53(1):50–71 Lujala P, Gleditsch NP, Gilmore E. 2005. A diamond curse? Civil war and a lootable resource. J. Confl. Res. 49(4):538–62 Maxwell JW, Reuveny R. 2005. Continuing conflict. J. Econ. Behav. Organ. 58(1):30–52 Mehlum H, Moene K, Torvik R. 2006. Cursed by resources or institutions? World Econ. 29(8):1117–31 Montalvo JG, Reynal-Querol M. 2005. Ethnic polarization, potential conflict, and civil wars. Am. Econ. Rev. 95(3):796–816 Morelli M, Rohner D. 2010. Natural resource distribution and the multiple forms of civil war. Presented at Annu. Meet. Am. Polit. Sci. Assoc., Seattle Olsson O. 2007. Conflict diamonds. J. Dev. Econ. 82(2):267–86 Olsson O, Siba E. 2013. Ethnic cleansing or resource struggle in Darfur? An empirical analysis. J. Dev. Econ. 103:299–312 Reuveny R, Maxwell JW. 2001. Conflict and renewable resources. J. Confl. Res. 45(6):719–42 Reuveny R, Maxwell JW, Davis J. 2011. On conflict over natural resources. Ecol. Econ. 70(4):698–712 Ross ML. 2004a. How do natural resources influence civil war? Evidence from thirteen cases. Int. Organ. 58(1):35–67 Ross ML. 2004b. What do we know about natural resources and civil war? J. Peace Res. 41(3):337–56 Ross ML. 2006. A closer look at oil, diamonds, and civil war. Annu. Rev. Polit. Sci. 9:265–300 Snyder R, Bhavnani R. 2005. Diamonds, blood, and taxes: a revenue-centered framework for explaining political order. J. Confl. Res. 49(4):563–97 Themnér L, Wallensteen P. 2011. Armed conflict, 1946–2010. J. Peace Res. 48(4):525–36 van der Ploeg F, Rohner D. 2012. War and natural resource exploitation. Eur. Econ. Rev. 56(8):1714–29 Weinstein JM. 2005. Resources and the information problem in rebel recruitment. J. Confl. Res. 49(4):598–624 Wick K, Bulte EH. 2006. Contesting resources—rent seeking, conflict and the natural resource curse. Public Choice 128(3–4):457–76

RELATED RESOURCE Cederman L-E, Gleditsch KS, eds. 2009. Special Issue: Disaggregating Civil War. J. Confl. Res. 53(4)

www.annualreviews.org




Natural Resources and Violent Conflict

83

Annual Review of Resource Economics

Contents

Volume 6, 2014

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Autobiographical A Conversation with Irma Adelman Irma Adelman, David Zilberman, and Eunice Kim . . . . . . . . . . . . . . . . . . . 1 Resources Measuring the Wealth of Nations Partha Dasgupta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Optimal Control in Space and Time and the Management of Environmental Resources W.A. Brock, A. Xepapadeas, and A.N. Yannacopoulos . . . . . . . . . . . . . . . 33 Natural Resources and Violent Conflict Eleonora Nillesen and Erwin Bulte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Regime Shifts in Resource Management Aart de Zeeuw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Fiscal Rules and the Management of Natural Resource Revenues: The Case of Chile Luis Felipe Céspedes, Eric Parrado, and Andrés Velasco . . . . . . . . . . . . . 105 Energy Oil Price Shocks: Causes and Consequences Lutz Kilian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 The Economics of Energy Security Gilbert E. Metcalf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Auctioning Resource Rights Kenneth Hendricks and Robert H. Porter . . . . . . . . . . . . . . . . . . . . . . . . 175

ix

Carbon Markets: Past, Present, and Future Richard G. Newell, William A. Pizer, and Daniel Raimi . . . . . . . . . . . . . 191 Environment What Do We Know About Short- and Long-Term Effects of Early-Life Exposure to Pollution? Janet Currie, Joshua Graff Zivin, Jamie Mullins, and Matthew Neidell . . . 217 Valuing Morbidity in Environmental Benefit-Cost Analysis Trudy Ann Cameron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

The Long-Run Discount Rate Controversy Christian Gollier and James K. Hammitt . . . . . . . . . . . . . . . . . . . . . . . . . 273 Consumption- Versus Production-Based Emission Policies Michael Jakob, Jan Christoph Steckel, and Ottmar Edenhofer . . . . . . . . . 297 Economic Experiments and Environmental Policy Charles N. Noussair and Daan P. van Soest . . . . . . . . . . . . . . . . . . . . . . 319 The Economics of Environmental Monitoring and Enforcement Jay P. Shimshack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 Payment for Ecosystem Services from Forests Jennifer Alix-Garcia and Hendrik Wolff . . . . . . . . . . . . . . . . . . . . . . . . . 361 Agriculture Consumer Acceptance of New Food Technologies: Causes and Roots of Controversies Jayson L. Lusk, Jutta Roosen, and Andrea Bieberstein . . . . . . . . . . . . . . . 381 The Economics of Voluntary Versus Mandatory Labels Brian E. Roe, Mario F. Teisl, and Corin R. Deans . . . . . . . . . . . . . . . . . . 407 Limitations of Certification and Supply Chain Standards for Environmental Protection in Commodity Crop Production Kurt B. Waldman and John M. Kerr . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Theory and Application of Positive Mathematical Programming in Agriculture and the Environment Pierre Mérel and Richard Howitt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451 Development Agriculture in African Development: Theories and Strategies Stefan Dercon and Douglas Gollin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471 x

Contents

Trade Liberalization and Poverty: What Have We Learned in a Decade? L. Alan Winters and Antonio Martuscelli . . . . . . . . . . . . . . . . . . . . . . . . 493 The Intersection of Trade Policy, Price Volatility, and Food Security Kym Anderson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513 The Power of Information: The ICT Revolution in Agricultural Development Eduardo Nakasone, Maximo Torero, and Bart Minten . . . . . . . . . . . . . . 533 Errata

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

An online log of corrections to Annual Review of Resource Economics articles may be found at http://www.annualreviews.org/errata/resource

Contents

xi

Annual Reviews It’s about time. Your time. It’s time well spent.

New From Annual Reviews:

Annual Review of Statistics and Its Application Volume 1 • Online January 2014 • http://statistics.annualreviews.org

Editor: Stephen E. Fienberg, Carnegie Mellon University

Annu. Rev. Resour. Econ. 2014.6:69-83. Downloaded from www.annualreviews.org by ${individualUser.displayName} on 10/23/14. For personal use only.

Associate Editors: Nancy Reid, University of Toronto Stephen M. Stigler, University of Chicago The Annual Review of Statistics and Its Application aims to inform statisticians and quantitative methodologists, as well as all scientists and users of statistics about major methodological advances and the computational tools that allow for their implementation. It will include developments in the field of statistics, including theoretical statistical underpinnings of new methodology, as well as developments in specific application domains such as biostatistics and bioinformatics, economics, machine learning, psychology, sociology, and aspects of the physical sciences.

Complimentary online access to the first volume will be available until January 2015. table of contents:

• What Is Statistics? Stephen E. Fienberg • A Systematic Statistical Approach to Evaluating Evidence from Observational Studies, David Madigan, Paul E. Stang, Jesse A. Berlin, Martijn Schuemie, J. Marc Overhage, Marc A. Suchard, Bill Dumouchel, Abraham G. Hartzema, Patrick B. Ryan

• High-Dimensional Statistics with a View Toward Applications in Biology, Peter Bühlmann, Markus Kalisch, Lukas Meier • Next-Generation Statistical Genetics: Modeling, Penalization, and Optimization in High-Dimensional Data, Kenneth Lange, Jeanette C. Papp, Janet S. Sinsheimer, Eric M. Sobel

• The Role of Statistics in the Discovery of a Higgs Boson, David A. van Dyk

• Breaking Bad: Two Decades of Life-Course Data Analysis in Criminology, Developmental Psychology, and Beyond, Elena A. Erosheva, Ross L. Matsueda, Donatello Telesca

• Brain Imaging Analysis, F. DuBois Bowman

• Event History Analysis, Niels Keiding

• Statistics and Climate, Peter Guttorp

• Statistical Evaluation of Forensic DNA Profile Evidence, Christopher D. Steele, David J. Balding

• Climate Simulators and Climate Projections, Jonathan Rougier, Michael Goldstein • Probabilistic Forecasting, Tilmann Gneiting, Matthias Katzfuss • Bayesian Computational Tools, Christian P. Robert • Bayesian Computation Via Markov Chain Monte Carlo, Radu V. Craiu, Jeffrey S. Rosenthal • Build, Compute, Critique, Repeat: Data Analysis with Latent Variable Models, David M. Blei • Structured Regularizers for High-Dimensional Problems: Statistical and Computational Issues, Martin J. Wainwright

• Using League Table Rankings in Public Policy Formation: Statistical Issues, Harvey Goldstein • Statistical Ecology, Ruth King • Estimating the Number of Species in Microbial Diversity Studies, John Bunge, Amy Willis, Fiona Walsh • Dynamic Treatment Regimes, Bibhas Chakraborty, Susan A. Murphy • Statistics and Related Topics in Single-Molecule Biophysics, Hong Qian, S.C. Kou • Statistics and Quantitative Risk Management for Banking and Insurance, Paul Embrechts, Marius Hofert

Access this and all other Annual Reviews journals via your institution at www.annualreviews.org.

Annual Reviews | Connect With Our Experts Tel: 800.523.8635 (us/can) | Tel: 650.493.4400 | Fax: 650.424.0910 | Email: [email protected]