ABSTRACT

This work conducts a comparative analysis on how diamonds and petroleum produce differing types of economic inequality in resource curse states, contributing to institutional entropy. By arguing for the causal primacy of resources in types of eventuated curses, this approach posits the concept of a “resource class” in diamond- and petroleum-producing resource curse states. Strength tests of resource classes against a variety of independent variables finds that petroleum-based resource classes funnel revenues to fewer, more powerful individuals than diamond-based classes, at the expense of currency stability, equal opportunity for women and minorities, and competing interest groups. Conversely, diamond resource classes tend to be more egalitarian, yet a negative correlation is observed between the market economy and diamond production among cursed states.

Resource curse states are characterized by degrees of economic underdevelopment, political dysfunction, and civil unrest. Scholars, attracted by the puzzling nature of the resource curse, attempt to explain why some resource-rich states become cursed while others remain immune to this paradox. Instead of engaging in such ontological debates concerning the existence of the resource curse, this work decouples economic differences among already-cursed states, hypothesizing that resources are causal agents in eventuated curses, specifically concerning degrees of economic inequality. By singling out two resources, diamonds and petroleum, this work develops the concept of a “resource class,” establishing a link between resource production and outcomes in economic inequality. Differences in resource class strength are pinned against the Bertelsmann Stiftung (2018) Transformation Index (BTI), measuring various aspects of institutional capacity, such as currency stability and market organization. The results from the cross-sectional analyses suggest varying degrees of institutional entropy contingent upon the resource extracted, lending credence to studies prioritizing material underpinnings in policy decision-making. Results show that petroleum tends to concentrate wealth in the top 10% and 1% of share earners, thereby marginalizing competing political interest groups, diminishing equal opportunity for women and ethnic minorities, and lowering overall economic robustness. Diamonds, conversely, are found to have a negative correlation with the status of the market economy, yet a stronger positive correlation with decreases in socioeconomic barriers. The overall picture painted by the scatterplots and multivariate regressions below illustrates how petroleum production in resource curse states develops different types of institutional entropy from those in diamond-producing cursed states.

Economic inequality in resource curse states

Throughout most of the twentieth century, resource-rich states were just as poor and (un)democratic as resource-poor ones. Yet surprisingly, since the 1970s, scholars have noticed either similar or slower economic development among resource-rich states (Ross 2012, 63). Since then, a large body of literature has asked how resources influence development and growth (e.g., Collier 2010; Humphreys et al. 2007; van der Ploeg 2011; Wick & Bulte 2009). Instinctively, analysts may expect that possessing large swaths of petroleum deposits or diamond mines would benefit industrialization efforts and enrich communities. Ideally, states would capture resources, export them for profit, and reinvest those monies domestically. However, the opposite phenomenon often occurs, with resources sometimes legitimizing authoritarian regimes or incumbent governments (Andersen and Aslaksen 2013; Cuaresma et al. 2010), leading to increases in corruption (e.g., Shaxson 2007), or suffering from Dutch Disease (e.g., Corden 1984). Scholars have labeled this scenario the “paradox of plenty,” but Richard Auty’s (1993) pithy term, “resource curse,” has stood the test of time.

Recently, scholars such as Menaldo (2016, 11) have considered institutions to be primary agents in actuating resource curse outcomes, with curses resulting as direct and joint consequences of both resources and bad institutions. Menaldo considers policies as determinants of outcomes, but these curses are institutional rather than pinned to resources. Since resources exist but are not agential, policy becomes the barometer by which scholars analyze curses, as resources can only (dis)incentivize agents to react to their presence. State-developed policies, therefore, become dimmer switches for curses; they can dull effects enough for prosperity, or intensify them to the point of institutional atrophy. As Castañeda et al. (2018) expect, resource-rich states often mimic policies of others at subsequent levels of development, hoping to capitalize upon newfound discoveries. However, such “stepwise” policy duplication is resource-contingent, leading the present article to consider the primacy of resources in institutional entropy among cursed states. From an economic perspective, Castañeda et al.’s (2018) policy duplication is a pathway states may take to become cursed, which is evidenced by deteriorating institutional capacities, specifically economic inequality among diamond- and petroleum-producing resource-cursed states.

Other scholars, such as Engerman and Sokoloff (2003), claim that different institutional structures may be reasonable substitutes for each other, discussing how systematic patterns of institutional evolution weaken the concept that exogenous change in institutions powers economic development. They partially borrow Douglass North’s definition of institutions as “encompassing the specific organizations or rules that constrain and influence human behavior” (Engerman and Sokoloff 2003, 7; North 1990). Yet, if institutional entropy, to use Gopalakrishnan’s (2005) term, penetrates and corrodes cursed states, then scholars ought to examine root causes atrophying institutional capacity, with this article suggesting that analysts may find answers in differences among resources themselves. Thus, this work deviates slightly from Menaldo’s (2016) scholarship affording institutions primary agential causality in resource curse outcomes; rather, this work suggests that resources influence institutional strength and/or entropy.

This approach borrows North’s (1990) definition of institutions, Gopalakrishnan’s (2005) concept of entropy, and Menaldo’s (2016) scholarship, but challenges Menaldo’s “resource-less resource curse,” whereby institutions are primary agents in eventuated curses with resources relegated as epiphenomenal variables. Rather, this work adopts a middle road, partially agreeing with Boschini et al.’s (2007) analysis that resources impact institutional capacity in nonmonotonic ways. Borrowing heavily from Acemoglu, Johnson, and Robinson (2002), who argue that initial conditions shaped by colonialism favor certain institutional arrangements over others, Boschini et al. (2007) emphasize the filtering of resources through matrices of institutions to produce outcomes. By drawing upon Robinson, Torvik, and Verdier (2002) and Mehlum, Moene, and Torvik (2002) in assuming resources to be nonmonotonic in institutional quality, Boschini et al. (2007) insinuate that resources hold potential influence over institutional behavior, rendering good institutions better and bad ones worse. This article parts from Boschini et al. (2007) in that institutions do not simply become “better or worse,” but rather are (de)formed by resource extraction to such an extent that analysts may consider such states as suffering from different types of institutional entropy. This is evident in economic institutional capacity among diamond- and petroleum-rich cursed states, especially in the formation of an elite class bedrocked by resource extraction, which this work calls the “resource class.”

In this sense, resources are agentless organizing forces incentivizing and restricting states to act in peculiar (and oftentimes suicidal) manners. To a great extent, the presence of resources constricts governments, as states are hardly motivated to leave diamonds or petroleum in the ground. And when resources are extracted, cursed states develop resource classes, potentially corroding economic institutions. Yet not all resource classes are created equal. Institutional entropy does not creep into extraction economic structures evenly, with diamond states experiencing different types of entropy than petroleum states.

As discerned from the introduction, analysts tend to agree that some resource-rich states experience political and economic dysfunction, but disagree on which pathways contribute to institutional entropy, be it resources, institutions, or a combination thereof. Some offer compelling quantitative evidence that the resource curse is a statistical artifact created by endogenous variables (e.g., Brooks and Kurtz 2016; Haber and Menaldo 2011). Others argue that export-based economies are better positioned to spur domestic economic growth in the short, but not long term (e.g., Collier and Goderis 2008). Common to most studies are comparisons between “blessed” and “cursed” states. Yet, if analysts are to study degrees of institutional entropy among already cursed states, writings ought to pit resources against each other to define differences in outcomes. Only from this perspective, and not from evaluating cursed states against blessed states, can analysts flesh out patterned variances of institutional strength among diamond- and petroleum-rich states. In this way, this work poses a different question than much of the previous literature: Do diamonds and petroleum manifest themselves differently in institutional capacity vis-à-vis economic inequality? This work champions such a possibility, arguing that among cursed states, causal priority may be afforded to resources, and not institutions, in explaining degrees of entropy.

Barring discoveries of new deposits, resources themselves are static. Therefore, intervening variables are institutional reactions to the presence of resources; this factor is always fluctuating, as resources do not compel states to act in specific ways once, but do so continuously over time. Specific qualities unique to both diamonds and petroleum influence policymakers’ preferences, shape agendas, and act as filters for potential curse outcomes, but do so in different manners. These pathways, in turn, determine how a state experiences a curse. One such pathway acting as a variable determining institutional entropy is the creation and fueling of a “‘resource class.”

The resource class

Both the petroleum and diamond trades conjure images of wealthy executives benefiting from mineral extraction businesses. The differences in the two markets reside in magnitude—as the petroleum industry is much larger than the diamond trade. However, as there are fewer individuals involved in diamond trading, it is uncertain which industry generates more inequality among workers, executives, and the neighboring industries, such as jewelry stores or paint products. This work hypothesizes the emergence of a strong, wealthy, and elite resource class revolving around the petroleum industry, especially if the government nationalizes facets of the petro-trade, and conversely, a weaker and less-abled resource class in diamond-states. The importance of the resource class would be magnified in states with strong nationalized oil companies (NOCs), which preserve revenues in the hands of government officials rather than among the civilian populace, leaving such states prone to a double exposure of resource class ills with few benefits to nonresource classes. Not only do the majority of civilians not see benefits of petroleum rents, but also petrodollars may be kept in governmental coffers and used as salaries for industry loyalists, rather than used to fund projects benefiting nonresource classes. Such expectations are not ipso facto resource curses, but pose political representation and economic inequality dilemmas for states with already atrophying institutions. Thus, possessing a strong resource class can be considered a pathway, not an outcome, leading to institutional entropy along resource-specific lines.

Before testing resource class strength in cursed states, it is important to predicate that diamonds are not plentiful or valuable enough to raise an entire state out of poverty. Assuming perfect distribution of resources—a theoretical construct, to be sure—this work will make use of per capita values of diamonds and petroleum, theoretically supposing resources were evenly distributed throughout the population. With this in mind, Figure 1 depicts revenue differences among diamonds and petroleum, with the two states enjoying the highest monetary gain from their resource at the right of the scatterplot. This dataset’s resource curse state with the most petroleum per capita is Kuwait, with the average Kuwaiti theoretically “earning” $25,004 annually from petroleum. The story is very different for diamonds. Angolans, on an annual per capita basis, make only $48.94 off diamonds, though they are this dataset’s cursed state with the most diamonds per capita. In discussing macroeconomic curses, it makes little sense to consider diamond-rich states to be on par with petroleum-rich states, as even assuming perfect distribution, diamond-generated revenue pales in comparison to petro-rents. Unlike their petroleum-rich counterparts, diamond-rich states cannot hope to guarantee the economic well-being of their populace simply by tapping into diamond revenues. Diamonds are valuable enough to fund rebellious activity and spark civil unrest, yet not lucrative enough to raise an entire state out of poverty. This is the peculiar curse burdening diamond states.

Figure 1

Per capita value of diamonds and petroleum in cursed states (2014 U.S. dollars).

Figure 1

Per capita value of diamonds and petroleum in cursed states (2014 U.S. dollars).

Close modal

Economics-minded readers will likely point to fiscal disparities found in Figure 1 as a reason for differences in resource class strength. However, simply enjoying higher per capita values of resources does not buffer against institutional entropy, as policymaking remains the intervening variable (read: pathway) in determining resource curse outcomes. This is evidenced by states, such as Venezuela, boasting large oil reserves, yet failing to feed their people. Other states, such as Equatorial Guinea, boast very high per capita oil revenues, yet remain economically disadvantaged, with their under-five mortality rates worsening by 20% between their discovery of oil in 1990 and when oil production reached 350,000 barrels per day in 2005 (Shaxson 2007, 1123). Conversely, if diamond mines do not escape the government’s clutches, this could lead to licit extraction, such as in Ghana (Snyder and Bhavnani 2005, 587), which could translate into institutional strength, rather than entropy.

Institutional problems may be complicated by how the resource class is fueled, as NOCs, wielded by every petroleum state in the following dataset, comprise a large portion of the resource class. Though NOCs may allow the extraction industry to act unilaterally, petroleum rents may be unusually high, unpredictable, and easily hidden. As Ross notes, “if foreign companies were the problem, then nationalization should have been the cure” (2012, 6). Yet other variables are at play in determining resource curse pathways, such as whether the state develops a strong resource class, siphoning revenues from other potentially profitable industries.

Testing resource class strength

Strength testing the resource class segregates a state’s income inequality using Gini coefficients and Piketty et al.’s (2015) World Inequality Database (WID) indicators from that of the resource—diamonds or petroleum—to triangulate the financial concentration of a state’s resource class. There is an assumption inherent in this logic: the greater financial capabilities the resource class enjoys directly correlate with the political power it wields. Given the state of world affairs, the present work views this as a risk worth taking.

Testing resource class strength dissects income inequality at the product level while accounting for a state’s general income inequality along with the aggregate value of the resource produced. The model would have to be recalibrated for states that refine petroleum to account for the added expenditure of purifying processes. To quantify the resource class, this work adapts three equations inspired by Hartmann et al.’s (2017, 92) Product Gini Index (PGI), which the Observatory of Economic Complexity often adapts as a measure of economic complexity.

Hartmann et al. (2017) did not create the PGI ex nihilo, and in fact do not discuss the term “resource class” at all. Rather, they are analyzing how resource exports contribute to inequalities in economic structures. Hartmann et al. (2017) develop the PGI from pioneers emphasizing economic structural transformations that resources may produce in states (Hirschman 1958; Rosenstein-Rodan 1943; Singer 1950). More recently, scholars have also introduced measurements of economic complexity, expanding analysts’ abilities “to quantify a country’s productive structure and have revived interest in the macroeconomic role of structural transformations” (Hausmann et al. 2007; Hartmann et al. 2016, 2; Cristelli et al. 2013). As Hartmann et al. (2017) note, these measures can be predictive of future economic growth, are relevant for social welfare programs, and measure poverty prevention (Ravallion 2004). Furthermore, the combination of products produced or extracted from a state determines available occupations, strength of unions, and advanced learning opportunities (Hartmann et al. 2017, 75), potentially contributing to a new middle class in emerging economies (Milanovic 2013). Interestingly, as Hartmann et al. (2017, 75–76) note, this may have the opposite effect in industrialized economies, leading to deindustrialization and deunionization in highly competitive markets. Many industrial workers in these states may be forced to work in lower-paying jobs or may decrease unions’ capabilities of negotiating for higher wages (Gustafsson and Johansson 1999; Acemoglu, Aghion, and Violante 2001).

While Hartmann et al.’s (2017) PGI is a valuable tool, certain variables must be altered to unearth resource class strength, as the PGI is determined by state exports and not national production. If this work were to embrace Hartmann et al.’s (2017) PGI in its equations, it would overestimate states exporting large percentages of their petroleum or diamonds while underestimate states consuming much of their own resource. To avoid this dilemma, this work alters Hartmann et al.’s (2017) formula, accounting for oil and diamond production, rather than exports. As such, it will measure three different types of inequality in state economies. First, it will incorporate an Industrial Age measure of inequality, the Gini coefficient, while cognizant of its weaknesses. It is clearly a dated measure of inequality reminiscent of an era when most industrialized states held few laws on labor rights. Gini coefficients also measure inequality in the form of quintiles, thereby not providing the granularity necessary for a more robust analysis. However, Gini coefficients are largely slow- moving and ubiquitous evaluators of wealth inequality. Data abound for resource curse states across longitudinal lines, rendering Gini coefficients a valuable benchmark against which to measure economic inequality stemming from resource production. To this end, this work develops an equation, motivated by Hartman et al.’s (2015) approach, to define each state’s Resource Class (RCc), incorporating the Gini coefficient as follows:

        

where Ginic is the Gini coefficient of state c. The annual resource per capita value (Vp/Dc) is included to measure wealth magnitude derived from resource classes, and SGDPcp is the share of a state’s product production divided by GDP. It should be noted that state GDPs are partially composed of resource revenues, as production p ultimately is counted as a sometimes-large fraction of a state c’s GDP. This is not problematic in testing economic inequality, as it does not seek to determine what states look like without diamonds or petroleum, but future scholars may find it useful to subtract the resource values from state GDPs in analyses when measuring others aspects of institutional entropy. Finally, Mcp is equal to 1 if state c’s production of product p has a revealed comparative advantage (RCA), determined by equation (3) below, greater than or equal to 1, and is 0 if it does not, as depicted below:

        

The binomial operator requires adding a synthetic control method outside of the summation of equation (1), where ΣcSGDPcp is equal to the sum of each state’s share of production of petroleum or diamonds compared with its overall GDP. In order to calculate Mcp, the RCAcp, adopted from Balassa’s (1965) index, calculates the relative (dis)advantage of a state for exporting a specific resource. The RCAcp is altered to fit the production, and not export, conditions. We can define RCAcp and hence Mcp as follows:

        

where Scp is a state’s resource production p and ΣScp measures the state’s entire market value of all goods and services (GDP) in 2014 U.S. dollars. The numerator allows analysts to determine the percentage of a state’s production stemming from the resource in question, while the denominator adopts the same formula for world resource production in comparison with world GDP. This ratio illustrates the RCA of a state per resource. If there is a comparative advantage, meaning that RCAcp is greater than or equal to 1, then equation (1) is determined by additional inputs defined above; otherwise, the result will be zero. Accounting for the RCA is necessary because it filters out economically disadvantaged states not producing enough diamonds or petroleum to warrant a bona fide resource class, as the rest of the equation would equal zero.1 A score of zero indicates no matter how strong a state’s resource class may be, it would not make a statistically significant difference in their country because their RCA does not meet the threshold as described above.

As previously stated, Gini coefficients only separate economic inequality into quintiles, and therefore do not provide the degree of specificity necessary to measure extreme inequality in economic structures. Thus, this work incorporates state data from Piketty et al.’s (2015) World Inequality Database (WID) to further triangulate economic inequality in cursed states. WID wealth indicators may be more accurate and modern calculators of inequality, but data are limited, especially among diamond-rich states, rendering statistical analyses difficult without using Gini coefficients as a benchmark. This work will replace the Ginic variable in equation (1) with Wc variables for equations (4) and (5). One can define each state’s Resource Class (RCc) in terms of WID scores as follows:

        

        

where Wc10 indicates the WID score of the top 10% of share earners and Wc1 denotes the top 1%. These two indicators allow varying degrees of specificity not found in Gini coefficients. Though accounting for fewer states, equations (4) and (5) allow the analyst to understand whether wealth becomes concentrated in each state’s top 10% or 1% of share earners, or permeates to (at least) the top quintile.

Naturally, as only one variable denoting inequality is different among the three equations, there will be some multicollinearity among the results. However, this is done by design. Equation 1 determines the strength of the top 20% of the resource class; equation (4) measures the top 10%, and equation (5) the top 1%. By comparing and contrasting the results from these three equations pinned against BTI indicators, scholars may better triangulate money flow vectors in cursed states. The intensity of economic inequality stemming from resource production would help scholars diagnose vulnerability of institutional entropy in resource curse states.

State selection

Thirty-five states demarcated into two categories will be evaluated. The first category comprises 13 diamond states: Angola (D), Cameroon, the Republic of the Congo, the Democratic Republic of the Congo, Côte d’Ivoire, Ghana, Guinea, Guyana, Liberia, Russia (D), Sierra Leone, Tanzania, and Zimbabwe. The second category of 24 countries comprises the group dubbed petroleum states: Algeria, Angola (P), Bahrain, Brunei, Ecuador, Equatorial Guinea, Gabon, Iran, Iraq, Kazakhstan, Kuwait, Libya, Malaysia, Mexico, Nigeria, Oman, Qatar, Russia (P), Saudi Arabia, Sudan, Syria, the United Arab Emirates, Venezuela, and Yemen. As Russia and Angola have both diamonds and petroleum, they will be coded as such throughout this work.

Many diamond states were selected because they are signatories of the Kimberley Process, which provides one of the few public and reliable datasets regarding diamond production. Yet, in recent years, many diamond states have begun to discover petroleum. Revenue from this newfound resource has easily outstripped (reported) diamond revenue. Liberia, for example, auctioned off offshore blocks for petroleum production, with many contracts sold to international corporations (Dean 2004). Offshore oil was also discovered in the Ivory Coast in the late 1970s, with production commencing a few years later. Some states, such as the Republic of the Congo, while diamond-rich, and coded as diamond states, produce and export enough petroleum to be considered petrostates as well. Nonetheless, the acute reason that the Republic of the Congo and others are documented as diamond producers is that blood diamonds have wrought havoc in these states while petroleum, though lucrative, has led neither to civil and gang violence on such scale nor to the illicit export of resources (e.g., Deibert 2007; Goldman 2008). Other states, such as Cameroon, are not great producers of diamonds, in terms either of value or of carats. Yet, due to the location of diamond mines on the Cameroon–Central African Republic border, diamonds have been smuggled across state boundaries, leading to the state’s failure to implement Kimberley Process goals (Jamasmie 2016). Still other states, such as the Democratic Republic of the Congo (DRC), enjoy enormous reserves of gold, copper, cobalt, timber, and uranium, on top of diamonds. Nevertheless, as Jones (2004) notes, diamonds are by far the DRC’s most valuable resource, acting as a pillar for its struggling economy.

Among petroleum states, dictatorships may result from political institutional curses (e.g., Kazakhstan, Equatorial Guinea, Libya under Gaddhafi, Iraq under Saddam Hussein). Furthermore, populist-leaning states legitimized by petro-rents blur lines of democratically elected governments (e.g., Venezuela, Russia, Angola, and to lesser extents, Mexico and Ecuador). Finally, analysts (e.g., Menaldo 2016) consider resource-rich monarchies fueled by petro-wealth as cursed in their analyses. These states include Oman, Saudi Arabia, Qatar, Kuwait, the UAE, and, to a lesser extent, Bahrain. Oil-rich monarchies are not peculiar to the Gulf; Brunei and Malaysia also have monarchs ruling over their petro-kingdoms, marginalizing democratic advances, though, at least in Malaysia, providing their citizenry with good-enough, albeit undemocratic, government. Still other states may suffer not only from political malformation but also from economic woes despite resource wealth (e.g., Yemen, Algeria, Sudan, and Syria). Regarding oil states, either a single factor, or a combination thereof, may expose a state to experience some type of resource curse. As the reader will note: Oil buys many things, but not institutional capacity.

Analysis

Once each state’s statistics are run through equation (1) denoting resource class strength at the state’s Gini coefficient, the results are quite striking. First, as illustrated in Figure 2, every single petroleum state has a stronger resource class than every diamond state, suggesting the primacy of resources in determining resource class strength. Additionally, Cameroon, Côte d’Ivoire, Ghana, and the Republic of the Congo do not hold enough diamonds to reach the RCAcp threshold. Therefore their score is zero, indicating that these states’ resource classes are not strong enough to make a statistically significant difference in institutional entropy. Interestingly, the resources’ per capita values, and not resource total values, point the strongest finger toward this end. The scatterplot below is a static snapshot of resource class strength in diamond- and petroleum-rich resource curse states. To guard against major outliers among petroleum states, this work logs the value per capita to display more subtle differences in resource classes among both types of states. Figure 2 depicts the correlation between resource per capita values (Vp/Dc) and the resource class in 2014, the last year this work has complete data for both diamonds and petroleum.

Figure 2

Per capita value (log) and resource class (log).

Figure 2

Per capita value (log) and resource class (log).

Close modal

With both per capita values and resource classes logged, the scatterplot above illustrates the disparity between diamond- and petroleum-producing states, with states below the RCAcp threshold remaining at a score of zero along the x-axis. Petroleum states, largely due to the per capita value of their resource, are likely to possess stronger resource classes than diamond states. Of particular note, there seems to be a gap in the middle of Figure 2 between Angola (D) and Syria, dividing diamond and petroleum states’ resource classes, further suggesting that the separation in resource class strength is contingent upon policymaking revolving around resource presence in state economies.

Figure 2 is a fixed assessment of resource class strength in 2014. To bring these static measures to life, this work will conduct multivariate regression analyses for the resource class from 2005 to 2015, separated into petroleum states, diamond states, and both petroleum and diamond states. The ordinary least squares regression in Table 1 borrows the results on institutional capabilities from the Bertelsmann Stiftung (2018) Transformation Index (BTI) as independent variables, evaluating the quality of market economies in developing states. In addition to the individual variables, this work aggregates indicators into one score, called “economic composite,” at the end of Table 1.

Table 1

Cross-sectional analysis—resource class measured by Gini coefficients (2005–2015).

VariablesGini PetroleumGini DiamondsGini Both
Currency stability —63,618** 45.21 —53,619*** 
 (25,451) (118.2) (16,092) 
Equal opportunity —90,341*** 52.71 —59,079*** 
 (20,763) (93.64) (13,103) 
Interest groups —58,747*** —121.4 —50,476*** 
 (18,788) (110.8) (12,852) 
Market economy status —136,360* —811.6** —99,524** 
 (69,006) (327.1) (47,217) 
Market organization 30,661 —66.38 32,968** 
 (23,740) (126.8) (16,156) 
Socioeconomic barriers 32,181 -38.29 19,779 
 (30,844) (104.6) (17,762) 
Economic composite 32,649*** 96.90* 25,215*** 
 (11,791) (52.58) (7,681) 
Constant 65,226 659.4 14,583 
 (89,317) (434.2) (57,976) 
Observations 102 61 163 
R-squared 0.442 0.180 0.419 
VariablesGini PetroleumGini DiamondsGini Both
Currency stability —63,618** 45.21 —53,619*** 
 (25,451) (118.2) (16,092) 
Equal opportunity —90,341*** 52.71 —59,079*** 
 (20,763) (93.64) (13,103) 
Interest groups —58,747*** —121.4 —50,476*** 
 (18,788) (110.8) (12,852) 
Market economy status —136,360* —811.6** —99,524** 
 (69,006) (327.1) (47,217) 
Market organization 30,661 —66.38 32,968** 
 (23,740) (126.8) (16,156) 
Socioeconomic barriers 32,181 -38.29 19,779 
 (30,844) (104.6) (17,762) 
Economic composite 32,649*** 96.90* 25,215*** 
 (11,791) (52.58) (7,681) 
Constant 65,226 659.4 14,583 
 (89,317) (434.2) (57,976) 
Observations 102 61 163 
R-squared 0.442 0.180 0.419 

Note: Standard errors in parentheses.

*** p < 0.01.

** p < 0.05.

* p < 0.1.

This multivariate regression depicts a statistically significant correlation between the resource class and multiple BTI independent variables. As determined from Table 1, there is a strong negative correlation between resource class strength and currency stability among petroleum states, at p < 0.05, while there are statistically nonsignificant positive associations among diamond states. Currency stability refers to institutional and political practices to control inflation, and procedures governing appropriate fiscal and monetary policy (Bertelsmann Stiftung 2018). Clearly, petroleum has a greater impact than diamonds on currency stability, as evidenced by some petrostates overrelying on petroleum during oil booms and suffering economic busts when prices drop. Such volatility ultimately leads to rapid inflation and deflation in petro-economies, which tend to hurt petroleum classes less than nonresource classes. Anecdotally, in Venezuela, the upper echelons of the elite class (read: resource class) paradoxically benefit from the bolivar’s hyperinflation and the downward economic spiral—they use the dollar (Brodzinsky 2015).

Statistically significant negative correlations at p < 0.01 between resource class strength and equal opportunity, which measures women and minority integration into economic structures, exist among petroleum states, but diamond states show nonsignificant positive correlations. The arrows flow in the same direction for interest groups, though significantly more strongly for petroleum states (p < 0.01). The Bertelsmann Stiftung (2018) interest groups marker denotes the capacity for underrepresented groups to successfully petition governments, and the negative correlation between this independent variable and resource classes among (specifically) petroleum states reflects a strong resource class with narrow political interests, potentially marginalizing competing political interest groups. The negative trend continues regarding the market economy indicator in both diamond- and petroleum-rich resource curse states at the p < 0.05 and p < 0.1 levels, respectively. This indicator scores levels of socioeconomic development, market organization and competition, price stability, private property, welfare, economic performance, and sustainability (Bertelsmann Stiftung 2018). These results suggest that while there may be poignant differences between diamond- and petroleum-producing resource classes, when taken in aggregate and accounting for Gini coefficients, there ultimately are similarities between the two types of resource classes.

Finally, this work took a composite score of Bertelsmann Stiftung (2018) economic variables, except for the market economy variable, which is itself a composite score, finding a strong, positive correlation for petroleum states (p < 0.01) and less robust for diamond states (p < 0.1), suggesting that petroleum plays a bigger role in defining resource classes than diamonds. Predictably, the r-squared for petroleum states at 0.442 is significantly higher than that for diamond states at 0.180, signifying that large revenues generated from petroleum rents account for higher levels of variability in petrostates’ results. Table 1 depicts correlations between different types of economic inequality and the resource class when accounting for Gini coefficients, finding that arrows flow in the same direction for interest groups, market economy status, and the economic composite score. However, diamond- and petroleum-producing states are correlated differently in currency stability, equal opportunity, and, though not at statistically significant levels, in market organization and socioeconomic barriers.

This work will now compare and contrast outputs between equation (1) and equations (4) and (5) to discern if resource class strength remains proportionally strong among the top 10% and top 1% of share earners. Table 2 displays the results.

When currency stability is analyzed for the top 10% and top 1% of share earners, there are negative correlations at the p < 0.1 levels across both diamond- and petroleum-producing states, but not individually. This may be contrasted with Table 1’s results, showing stronger correlations between the resource class and currency stability at p < 0.05 levels for petroleum states and at p < 0.01 levels for both types of states. This implies that, while the resource class may influence currency stability, this correlation is stronger for the top quintile than for the top 10% or top 1% of share earners. A different situation exists with the Bertelsmann Stiftung (2018) equal opportunity indicator, measuring market participation of women and minority ethnic/racial groups. There is a correlation at the p < 0.01 level for petrostates in equation (1) and at the p < 0.05 level for equations (4) and (5), further suggesting a compartmentalization of the resource class in the top 10% or 1% of share earners rather than the top quintile.

Table 2

Cross-sectional analysis—resource class at top 10% and top 1% (2005–2015).

VariablesTop 10% petroleumTop 10% diamondsTop 10% bothTop 1% petroleumTop 1% diamondsTop 1% both
Currency stability -427.3 0.473 —388.7* —133.0 0.228 —124.7* 
 (360.8) (0.330) (213.9) (120.4) (0.160) (71.56) 
Equal opportunity —675.2** —0.398 —464.3*** —230.9** —0.189 —158.9*** 
 (294.3) (0.261) (174.2) (98.26) (0.127) (58.27) 
Interest groups —447.6* —0.495 —307.0* —150.5* —0.231 —102.4* 
 (266.3) (0.309) (170.9) (88.91) (0.150) (57.15) 
Market economy status —2,301** —1,544* —1,565** —794.7** —0.611 —539.3** 
 (978.1) (0.913) (627.7) (326.6) (0.442) (210.0) 
Market organization 170.0 0.416 166.7 56.07 0.161 55.24 
 (336.5) (0.354) (214.8) (112.3) (0.171) (71.84) 
Socioeconomic barriers 248.8 0.771** 113.6 93.84 0.358** 43.65 
 (437.2) (0.292) (236.1) (146.0) (0.141) (78.99) 
Economic composite 392.3** 0.161 289.6*** 132.2** 0.0678 97.83*** 
 (167.1) (0.147) (102.1) (55.80) (0.0711) (34.15) 
Constant —298.1 —2.156* —382.9 —93.55 —1.058* —129.9 
 (1,266) (1.212) (770.8) (422.7) (0.587) (257.8) 
Observations 102 61 163 102 61 163 
R-squared 0.206 0.537 0.175 0.213 0.529 0.180 
VariablesTop 10% petroleumTop 10% diamondsTop 10% bothTop 1% petroleumTop 1% diamondsTop 1% both
Currency stability -427.3 0.473 —388.7* —133.0 0.228 —124.7* 
 (360.8) (0.330) (213.9) (120.4) (0.160) (71.56) 
Equal opportunity —675.2** —0.398 —464.3*** —230.9** —0.189 —158.9*** 
 (294.3) (0.261) (174.2) (98.26) (0.127) (58.27) 
Interest groups —447.6* —0.495 —307.0* —150.5* —0.231 —102.4* 
 (266.3) (0.309) (170.9) (88.91) (0.150) (57.15) 
Market economy status —2,301** —1,544* —1,565** —794.7** —0.611 —539.3** 
 (978.1) (0.913) (627.7) (326.6) (0.442) (210.0) 
Market organization 170.0 0.416 166.7 56.07 0.161 55.24 
 (336.5) (0.354) (214.8) (112.3) (0.171) (71.84) 
Socioeconomic barriers 248.8 0.771** 113.6 93.84 0.358** 43.65 
 (437.2) (0.292) (236.1) (146.0) (0.141) (78.99) 
Economic composite 392.3** 0.161 289.6*** 132.2** 0.0678 97.83*** 
 (167.1) (0.147) (102.1) (55.80) (0.0711) (34.15) 
Constant —298.1 —2.156* —382.9 —93.55 —1.058* —129.9 
 (1,266) (1.212) (770.8) (422.7) (0.587) (257.8) 
Observations 102 61 163 102 61 163 
R-squared 0.206 0.537 0.175 0.213 0.529 0.180 

Note: Standard errors in parentheses.

*** p < 0.01.

** p < 0.05.

* p < 0.1.

Of particular note, resource class strength in petrostates is negatively correlated with capacities for interest groups to mediate between society and political systems at the p < 0.01 level for equation (1) and at the p < 0.1 level for equations (4) and (5), suggesting that large numbers of social interests remain sidelined, but this underrepresentation does not penetrate into the top 10% or 1% of share earners. This work finds a negative correlation with the Bertelsmann Stiftung (2018) composite variable “market economy status” among petroleum states at the p < 0.05 level for equations (4) and (5) but only at the p < 0.1 level for diamond states in equation (4_. However, for equation (1), the correlations were stronger for diamond states and weaker for petroleum states, hinting that, at least from the perspective of the Bertelsmann Stiftung (2018) market economy composite variable, the resource class in diamond states is more robust at the top quintile than at the top 10% or 1%. However, the opposite is the case with petroleum, suggesting a stronger resource class at the top 10% and 1% of share earners than in the top quintile.

As with equation (1), this work took a composite score of the Bartelsmann Stiftung (2018) economic variables, except for market economy, finding, for petroleum states, positive correlation between the strength of the resource class and state economic institutions at p < 0.01. Conversely, this work only found positive correlation with these variables among diamond states for equation (1) at p < 0.05, but negative correlation for equations (4) and (5), suggesting changes in direction and magnitude among diamond states as inequality increases.

Conclusion

There is a gap in the resource curse literature. Too often, resource curse is treated as a singular, multipronged monolith affording institutions primacy and relegating resources as epiphenomenal in institutional entropy. Rather, these results show that resources are capable of influencing institutional capacity, specifically regarding currency stability, equal opportunity, interest groups, and the status of market economies. This paper forwards the hypothesis that diamond- and petroleum-producing resource curse states develop different types of resource classes, which siphon revenues into the top echelons of society. This hypothesis was purposefully picked to illustrate how diamonds and petroleum may lead to different types of inequality among resource curse states, with petroleum generally leading to stronger resource classes than diamonds along a variety of BTI variables. Contrary to this work’s initial hypothesis, resource classes are not simply “stronger” or “weaker.” Rather, they are (de)formed by resources fueling them, with petroleum correlating negatively with some variables, such as currency stability, and diamonds correlating differently with others, as with decreases in socioeconomic barriers.

More importantly, results illustrate how petroleum production benefits the top 10% and top 1% of share earners more than diamond production, suggesting that petroleum-producing resource classes funnel money beyond the top quintile and toward pockets of the top 10% and 1%. Conversely, diamond resource classes tend to be more egalitarian, but still suffer from significant institutional problems, specifically negative correlations with the status of their market economies. With these results, this work suggests two policy recommendations each for diamond- and petroleum-rich cursed states, based on Gopalakrishnan’s (2005) typology. These recommendations are designed to strengthen institutional adaptability and accountability, as shown in Table 3.

Table 3

Policy implications for strengthening institutional capacity.

DiamondsPetroleum
Institutional adaptability Diamond-producing states rarely enjoy economic linkages their resources offer, as locals tend not to polish diamonds domestically. To spur economic development, diamond states ought to support linkages between extractors, polishers, and jewelers. Petroleum-rich governments ought to use rents to diversify economies into petroleum-linked manufacturing and industrial sectors, rather than for paternalistic and short-term goals. 
Institutional accountability Diamond-rich states ought to support artisanal diamond mining for low-income communities by enforcing property rights to ameliorate disparities in their resource class. National governments ought to support decentralization of petroleum rents to provincial and local governments to avoid concentration of petro-wealth in one national resource class. 
DiamondsPetroleum
Institutional adaptability Diamond-producing states rarely enjoy economic linkages their resources offer, as locals tend not to polish diamonds domestically. To spur economic development, diamond states ought to support linkages between extractors, polishers, and jewelers. Petroleum-rich governments ought to use rents to diversify economies into petroleum-linked manufacturing and industrial sectors, rather than for paternalistic and short-term goals. 
Institutional accountability Diamond-rich states ought to support artisanal diamond mining for low-income communities by enforcing property rights to ameliorate disparities in their resource class. National governments ought to support decentralization of petroleum rents to provincial and local governments to avoid concentration of petro-wealth in one national resource class. 

The results of this work, and subsequent policy recommendations, run against contemporary scholarly trends suggesting a fundamentally social underpinning for institutional capacity. Whereas the current narrative suggests a “turtles all the way down” theme in resource curse literature, these results imply a fundamentally material underpinning for levels of inequality among cursed states. This has important implications for scholarly analyses of natural resource politics and for future policymaking, as the latter must be congruent with the resources in question. While policy implications are specific to each resource curse state, the guidelines postulated above offer viable avenues for policymakers among diamond- and petroleum-rich producers to avoid institutional entropy in cursed states.

Note

1.

This work adopts GDP data from the CIA World Factbook. Petroleum data are taken at the end of 2016, with world oil production reaching 3.3 billion barrels. This figure is then pinned to the price of Brent Crude Oil, largely understood as the world’s benchmark for crude oil prices, trading at $43.74 per barrel, equating to $147 billion in world production for 2016 (British Petroleum 2017). As for diamonds, the average price per carat is $110, while total world production of diamonds in 2017 is estimated to be 142.29 million carats, equating to $15.6 billion in revenue (Zimnisky 2017).

References

Acemoglu, D., P. Aghion, and G. L. Violante (2001). “Deunionization, technical change and inequality,” SSRN Scholarly Paper No. 267264. Rochester, NY: Social Science Research Network.”
Acemoglu, Daron, Simon Johnson, and James A. Robinson (2002). “Reversal of fortune: Geography and institutions in the making of the modern world income distribution.” Quarterly Journal of Economics, 117(4): 1231–94.
Andersen, Jørgen and Silje Aslaksen (2013). “Oil and political survival.” Journal of Developmental Economics, 100(1), 89–106.
Auty, Richard (1993). Sustaining development in the mineral economies: The resource curse thesis. London: Routledge.
Balassa, Bela (1965). “Trade liberalisation and ‘revealed’ comparative advantage, 1.” Manchester School, 33(2): 99–123.
Bertelsmann Stiftung (2018). “BTI 2016: Transformation index.” Available at http://www.bti-project.org/en/index/overview.
Boschini, Anne D., Jan Pettersson, and Jesper Roine (2007). “Resource curse or not: A question of appropriability.” Scandinavian Journal of Economics, 109(3): 593–617.
British Petroleum (2017). BP Statistical Review of World Energy, 66th ed., June 2017. Available at http://large.stanford.edu/courses/2018/ph241/kuet2/docs/bp-2017.pdf.
Brodzinsky, Sibylla (2015). “Venezuela’s high life elite hope hard-hit poor will abandon Chávez’s legacy.” The Guardian. Available at https://www.theguardian.com/world/2015/dec/06/venezuela-elite-hope-poor-abandon-chavez-legacy-election.
Brooks, Sarah M., and Marcus J. Kurtz (2016). “Oil and democracy: Endogenous natural resources and the political ‘resource curse.’” International Organization, 70(2): 279–311.
Castañeda, Gonzalo, Florian Chávez-Juárez, and Omar A. Guerrero (2018). “How do governments determine policy priorities? Studying development strategies through spillover networks.” Journal of Economic Behavior and Organization 154: 335–61.
Collier, Paul (2010). The plundered planet: Why we must—and how we can—manage nature for global prosperity. Oxford: Oxford University Press.
Collier, Paul, and Benedikt Goderis (2008). “Commodity prices, growth, and the natural resource curse: Reconciling a conundrum.” Available at http://dx.doi.org/10.2139/ssrn.1473716.
Corden, Warner Max (1984). “Booming sector and Dutch disease economics: Survey and consolidation.” Oxford Economic Papers 36(3): 359–80.
Cristelli, Matthieu, Andrea Gabrielli, Andrea Tacchella, Guido Caldarelli, and Luciano Pietronero (2013). “Measuring the intangibles: A metrics for the economic complexity of countries and products.” PloS one 8(8): e70726.
Cuaresma, Jesus Crespo, Harald Oberhofer, and Paul Raschky (2010). “Oil and the duration of dictatorships.” Public Choice, 148(3–4): 505–30.
Dean, F. Musah (2004). “NOCAL 2004 Offshore Bid Announcement.” Business Wire. Available at https://www.businesswire.com/news/home/20040202005192/en/NOCAL-2004-Liberia-Offshore-Bid-Announcement.
Deibert, M. (2007). “TRADE: Blood diamonds no longer Congo-Brazzaville’s best friend.” International Press Service. Available at http://www.ipsnews.net/2007/11/trade-blood-diamonds-no-longer-congo-brazzavillersquos-best-friend/.
Engerman, Stanley L., and Kenneth L. Sokoloff (2003). “Institutional and non-institutional explanations of economic differences.” In Claude Ménard and Mary M. Shirley (Eds.), Handbook of new institutional economics (pp. 639–65). Boston: Springer.
Goldman, Haley Blaire (2008). “Between a ROC and a hard place: The Republic of Congo’s illicit trade in diamonds and efforts to break the cycle of corruption.” University of Pennsylvania Journal of International Law, 30: 359–97.
Gopalakrishnan, Chennat (2005). “Water allocation and management in Hawaii: A case of institutional entropy.” In Chennat Gopalakrishnan, Cecilia Tortajada, and Asit K. Biswas, Water institutions: Policies, performance and prospects (pp. 1–23). Berlin/New York: Springer.
Gustafsson, Björn, and Mats Johansson (1999). “In search of smoking guns: What makes income inequality vary over time in different countries?” American Sociological Review 64(4): 585–605.
Haber, Stephen, and Victor Menaldo (2011). “Do natural resources fuel authoritarianism? A reappraisal of the resource curse.” American Political Science Review, 105(1): 1–26.
Hartmann, Dominik, Miguel R. Guevara, Cristian Jara-Figueroa, Manuel Aristarán, and César A. Hidalgo (2016). Linking economic complexity, institutions, and income inequality. Working draft. Available at https://arxiv.org/pdf/1505.07907.pdf.
Hartmann, Dominik, Miguel R. Guevara, Cristian Jara-Figueroa, Manuel Aristarán, and César A. Hidalgo (2017). “Linking economic complexity, institutions, and income inequality.” World Development, 93: 75–93.
Hausmann, Ricardo, Jason Hwang, and Dani Rodrik (2007). “What you export matters.” Journal of Economic Growth, 12(1): 1–25. Available at http://doi.org/10.1007/s10887-006-9009-4.
Hirschman, Albert O. (1958). The strategy of economic development. New Haven, CT: Yale University Press.
Humphreys, Macartan, Jeffrey D. Sachs, and Joseph E. Stiglitz. (2007). Escaping the resource curse. New York: Columbia University Press.
Jamasmie, C. (2016). “Cameroon involved in Central African ‘conflict diamonds’ trade—Report.” Mining.com. Available at http://www.mining.com/cameroon-involved-in-central-african-conflict-diamonds- trade-report/
Jones, Luis. (2004). Preventing the export of conflict diamonds in the Democratic Republic of the Congo. E297 Winter. Civil Wars in Africa: Southern Africa, B. Stanford University. Available at https://web.stanford.edu/class/e297a/CIVIL%20WARS%20IN%20AFRICA.htm.
Mehlum, Halvor, Karl Moene, and Ragnar Torvik (2002). Institutions and the resource curse. Mimeo.
Menaldo, Victor (2016). The institutions curse: Natural resources, politics, and development. Cambridge: Cambridge University Press.
Milanovic, Branko (2013). All the Ginis Dataset. Available at http://data.worldbank.org/data-catalog/all-the-ginis.
North, Douglass. (1990). Institutions, institutional change and economic performance. Cambridge, UK: Cambridge University Press.
Piketty, Thomas, Facundo Alvaredo, Lucas Chancel, Emmanuel Saez, and Gabriel Zucman. (2015). World Inequality Database (WID). Available at https://wid.world.
Ravallion, M. (2004). Pro-poor growth: A primer. Policy Research Working Paper No. ID 610283. World Bank, Washington, DC.
Robinson, James A., Ragnar Torvik, and Thierry Verdier (2002). Political foundations of the resource curse. CEPR Discussion Paper 3422.
Rosenstein-Rodan, Paul N. (1943). “Problems of industrialisation of eastern and south-eastern Europe.” Economic Journal, 53(210/211): 202–11.
Shaxson, Nicholas (2007). “Oil, corruption and the resource curse.” International Affairs, 83(6): 1123–40.
Singer, Hans W. (1950). “The distribution of gains between borrowing and investing countries.” American Economic Review, 40(2): 473–85.
Snyder, Richard & Bhavnani, Ravi. (2005). “Diamonds, blood, and taxes: A revenue-centered framework for explaining political order.” Journal of Conflict Resolution, 49(4): 563–97.
Van der Ploeg, Frederick (2011). “Natural resources: Curse or blessing?” Journal of Economic Literature, 49(2): 366–420.
Wick, Katharina, and Erwin Bulte (2009). “The curse of natural resources.” Annual Review of Resource Economics, 1(1): 139–56.
Zimnisky, P. (2017). “2017 Global Natural Diamond Production Forecasted at 142M Carats Worth $15.6B,” Paul Zimnisky Diamond Analytics. Available at http://www.paulzimnisky.com/2017-global-natural-diamond-production-forecasted-at-142m-carats-worth-15-6b.