The political economy of oil and gas in Southeast Asia: heading towards the natural resource curse?

Transcription

1 The Pacific Review, Vol. 23 No. 2 May 2010: The political economy of oil and gas in Southeast Asia: heading towards the natural resource curse? Benjamin K. Sovacool Abstract The notion of the resource curse suggests that countries with large caches of natural resources often perform worse in terms of economic growth, social development, and good governance than other countries with fewer resources. The theory posits that countries depending on oil or other extractive industries for their livelihood are among the most economically troubled, socially unstable, authoritarian, and conflict-ridden in the world. This article explores whether the resource curse is occurring in relation to oil and gas production in Southeast Asia, where investments in oil and gas infrastructure are expected to increase significantly. The article begins by conceptualizing the resource curse before explaining the factors believed to cause it. It then proposes metrics that can be used to identify the presence of the resource curse before testing these metrics on the five Southeast Asian countries with the largest rates of oil and gas production and reserves Brunei, Indonesia, Malaysia, Myanmar, and Thailand from 1987 to The article compares the performance of these Southeast Asian countries with the five largest producers in the Organization of Petroleum Exporting Countries as well as Brazil, China, India, the Russian Federation, and South Africa. The article concludes that the resource curse is not occurring in any of these countries, and that the theory may be too simplistic and deterministic to fully explain why some countries appear to be cursed with resources while others are blessed. Keywords Resource curse; crude oil; natural gas; energy security. 1. Introduction Emil Salim, the President of the Extractive Industries Review, an attempt by the World Bank to improve transparency and accountability in the Dr Benjamin K. Sovacool is Assistant Professor at the Lee Kuan Yew School of Public Policy at the National University of Singapore. He is also a Research Fellow in the Energy Governance Program at the Centre on Asia and Globalization. Address: 469C Bukit Timah Road, Singapore, bsovacool@nus.edu.sg The Pacific Review ISSN print/issn online C 2010 Taylor & Francis DOI: /

2 226 The Pacific Review mining and energy sectors, once began a conference on governance with the following statement. Not only have the oil, gas, and mining industries not helped the poorest people in developing countries, they have often made them worse off (Steiner 2007: 2). Juan Pablo Perez Alfonzo, one of the founders of the Organization of Petroleum Exporting Countries (OPEC), went even further and called oil the devil s excrement (Karl 2005: 21). Both statements imply that for many countries with relatively rich endowments of natural resources ranging from gems and copper to gas and crude oil, those very commodities can become curses to the government and its citizens. Indeed, a long line of political economists and theorists have stated different variants of the resource curse argument, which explains that natural resource endowments can incite, prolong, and intensify government failure and violent conflict. Stevens (2004) argues that large windfall revenues from oil or gas production can change government behavior, damage economic growth, and imperil development. Karl (1998, 2005) postulates that the consolidation and extraction of natural resources can create and solidify asymmetries in wealth that then contribute to rising income gaps between the rich and poor, institutionalize corruption, and enable oppressive regimes to maintain their political power. De Soysa (2000) notes that civil strife and social instability are strongly associated with natural resource abundance, particularly with mineral, gas, and oil exports. Silberfein (2004) states that resource based conflicts can become some of the most intractable because participants often benefit from unstable conditions, which can facilitate access to resources, smuggling, and covert trade. Beneficiaries not only grow in their power but also become reluctant to terminate bloody conflagrations. Le Billon (2004) suggests that resources can be used to finance conflicts (such as the civil war in Sierra Leone over diamonds), or can be a motive for other countries to go to war (such as the Iraqi invasion of Kuwait in 1990). Omorogbe (2006) expresses that a majority of resource-rich countries have performed well below expectations and capability and with little to show after decades of resource extraction. Esanov et al. (2006) comment that Azerbaijan, Kazakhstan, Turkmenistan, and Uzbekistan have performed worse than other countries in the Caspian Sea and Central Asia because their energy resources have been appropriated by elites and stunted economic development. For confirmation of the resource curse thesis, consider that in Saudi Arabia, proven crude oil reserves are the greatest in the world but per capita income has plunged from $28,600 in 1981 to $6,800 in 2001 (Baer 2003). In Nigeria and Venezuela, also rich in crude oil and natural gas, real per capita income has regressed to levels of the 1960s. Algeria, Angola, the Congo, Ecuador, Gabon, Iran, Iraq, Kuwait, Libya, Qatar, and Trinidad and Tobago, countries with substantial endowments of minerals, gold, timber, and diamonds, have seen real per capita income fall back to levels of the 1970s and 1980s as the extraction of these resources has increased. In late 1980s

3 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 227 Afghanistan, Ahmad Shah Massoud was able to earn $50 million annually for his United Front campaign against the Russians by selling locally harvested emeralds and lapis lazuli, and the civil wars in Zaire and the Congo were directly exacerbated by gold and diamond mining. These examples are not anecdotal, and O Lear (2004) has identified 10 serious civil wars and conflicts from 1990 to 1999 directly fueled by natural resources: copper facilitated a secessionist revolt in Bougainville, Papa New Guinea; oil funded five separate civil wars in Angola, East Timor/Indonesia, the Kurkuk region of Iraq, Southern Sudan, and the Xinjiang province of China; natural gas enabled a conflict in Aceh, Indonesia; nickel a secessionist movement in New Caledonia, France; gold financed a war in West Papua, Indonesia; and phosphates contributed to conflict in the Western Sahara and Morocco. Controlling for structural attributes, Humphreys et al. (2007) estimate that resource-rich countries as a whole have shown less economic development than resource-poor countries from 1975 to Auty (2006) compared the economic performance of developing countries rich in resources with those poor in resources and found that resource rich countries had a median gross domestic product (GDP) per capita of about $200 less than resource poor ones. The theory of the resource curse is worrisome and alarming for Southeast Asia for at least two reasons. First, resource extraction is anticipated to increase dramatically throughout the region. Southeast Asia is where significant investments in oil and natural gas are being called for to satisfy robust regional economic growth, diversify oil and gas supply outside of the Middle East, and satisfy demand for imported fuel in China, Japan, and South Korea. Regulators have rushed to invest in oil and gas, and by 2030 some analysts expect that Southeast Asia will become the Persian Gulf of Gas and be responsible for one-quarter of the world s gas production (Tan 2008). The International Energy Agency (2008) anticipates the need to invest trillions of dollars in electricity and energy supply in Southeast Asia by 2030, and the Asian Development Bank and World Bank expect to spend more than $50 billion in infrastructure lending between 2005 and 2015 (Bapna 2006). Second, corruption and poor governance are prevalent in Southeast Asia. One World Bank survey recently detailed that large-scale infrastructure projects in Southeast Asia, because of their capital intensive nature, practically invite corruption. The survey documented that in Indonesia, a physical audit of roads built under the oversight of village heads uncovered missing materials worth 24 per cent of total expenditures; in Cambodia and Laos, up to one-third of electricity generated was stolen; and that an average firm working in Asia expects to lose 11 per cent of its budget for infrastructure projects on bribes (Kenny 2007). If rapid resource extraction threatens to fundamentally alter the political and economic structure of exporting countries, and investments in oil and gas infrastructure will only accelerate resource depletion and increase

4 228 The Pacific Review corruption, then it is essential to explore the dynamics of the resource curse in Southeast Asia. This article begins by conceptualizing the resource curse before explaining the factors believed to cause it. It then proposes metrics that can be used to identify the presence of the resource curse before testing these metrics on the five Southeast Asian countries with the largest rates of oil and gas production and reserves Brunei, Indonesia, Malaysia, Myanmar, and Thailand from 1987 to The article then compares the performance of these Southeast Asian countries with the five largest producers in the Organization of Petroleum Exporting Countries as well as Brazil, China, India, the Russian Federation, and South Africa. Such an investigation adds important depth to discussions of the resource curse because so far exemplars of scholarship have mostly ignored Southeast Asia (with the exception of Malaysia and Indonesia) and instead focused on Africa and the Middle East. Moreover, most debates about the resource curse frame it in somewhat abstract terms that make it difficult to comprehend and even more challenging to measure. One value to this article is that it proposes ways of defining and quantifying the presence of the resource curse (or lack of it) that may be useful to policy-makers and other scholars studying extractive industries around the world. The article also provides a way of concretely testing the validity of the resource curse as a concept 30 years after its inception for the performance of 15 countries from 1987 to Conceptualizing the natural resource curse Put simply, the resource curse suggests that countries with large caches of natural resources often perform worse in terms of economic growth, social development, and good governance than other countries with fewer resources (Humphreys et al. 2007). The theory posits that countries depending on oil or other extractive industries for their livelihood are among the most economically troubled, socially unstable, authoritarian, and conflict-ridden in the world (Karl 2005). These resource-rich but practically-poor countries have slower than average economic growth and little industrial diversification. The curse or mismanagement of their resources expands outside the extractive industries sector and slowly degrades social welfare. Resource cursed countries also display rising levels of poverty and inequality, deteriorating environmental quality, institutionalized corruption, and an increased frequency of conflict and war (Karl 2005). Ross (1999) elaborates that three separate strands of the resource curse theorem exist. The cognitive strand implies that windfalls from resource extraction produce myopic disorders among policy-makers. The social strand suggests that windfalls empower recipients that then favor growth-stunting and self-serving policies. The statist strand argues that windfalls weaken the state institutions necessary for long-term economic development. Many investigations

5 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 229 weave all three strands together. The inverse of the resource curse also appears justifiable: lack of natural resources has not prevented Hong Kong, Singapore, South Korea, and Taiwan from achieving booming export industries based on solid manufacturing sectors and rapid economic growth (Humphreys et al. 2007). The problem, in essence, is that many of the most valuable resources such as diamonds, gold, oil, and gas are worth so much money that their extraction can be severely mismanaged but still create profits for companies and governments (Maassarani et al. 2007: 139). Resource cursed governments tend to manage natural resources opportunistically rather than strategically. That is, government officials view them as a mechanism to create immediate wealth rather than a long-term but depletable asset that should be carefully managed over time. Resource depletion therefore tends not to serve the public interest, and instead consolidates wealth among a very narrow class of bureaucrats and managers. Moreover, the immense value of those resources can mask mismanagement and inequity. As one high ranking executive for a large transnational oil company told the author, high prices [for natural resources] cover a multitude of sins. The easiest resources to control the ones most germane to the resource curse are point resources, resources spatially concentrated in small areas, relatively easy to monitor, and extracted through capital intensive technologies. Point resources contrast diffuse resources, which are spatially distributed and therefore more lootable, exploited by less capital intensive measures, and more difficult to consolidate (Le Billon 2004; O Lear 2004). Point and diffuse resources can contribute to different patterns of capital ownership and distribution of resource benefits. While both can be corrupted, or rapidly extracted or pillaged to serve limited ends or means, point resources are more prone to the resource curse. Because of their spatially concentrated nature, point resources are well attuned to highly centralized control. Ruling elites can extract them by controlling small geographic areas, and the revenues generated by even models levels of extraction can be significant. Karl (2005) argues that point resources are easily the worst managed, for they can be extracted from a narrow geographic or economic base and have the strongest correlation with poor growth. Diffuse resources such as coal, charcoal, wood, diamonds, gems, cobalt, gold, silver, sapphires, rubies, and other small, strategic, or valuable metals are more difficult to track and govern due to their distributed nature. While sapphires and rubies provided the Khmer Rouge in Cambodia and Karen in Burma with significant revenues during 1990s, their extraction could only occur by controlling vast amounts of land through almost complete authoritarian control. Five distinct and interrelated conditions appear to explain the structural social, political, and economic factors behind the resource curse (Humphreys et al. 2007; Bapna 2006; Stiglitz 2005; Karl 2005; Stevens 2004; Sachs and Warner 1999; Karl 1998).

6 230 The Pacific Review Figure 1 Real and nominal global crude oil prices, (in US dollars per barrel). Source: U.S. Energy Information Administration (2008a). First, those countries most susceptible to the resource curse have economies that are not diversified and remain heavily dependent on a few extractive industries. When the price of these few commodities fluctuates widely, the sudden price gyrations and boom and bust cycles erode budgetary discipline, decrease expenditures for public sector investment, and complicate efforts at state planning. Oil prices, for example, have been incredibly volatile from 1980 to 2008 (see Figure 1). Mikesell (1997) calculated that between 1972 and 1992, regions with high primary mineral and oil exports experienced periods of trade volatility two to three times greater than average industrialized economies over the same period. The volatility of any mineral commodity is exacerbated by banks and financial institutions. When commodity prices fall or interest rates rise, lenders are quick to call in their loans (a general maxim of bankers is that they always prefer to lend to those that do not need their money). When prices fall, resourcedependent countries also need more money, but this is usually when banks want their money back. Capital flows and the decline of oil prices, for example, tend to be pro-cyclical, worsening the fluctuations brought about by the initial fall in the price of oil. Falling prices also tend to incentivize increased amounts of resource production to maintain stable levels of revenue, precipitating greater environmental damage, economic concentration, and depletion of the resource. When prices are high, leaders face expectations that they should increase spending, embark on capital intensive projects, and reward productive components of the economy with subsidies (Tsalik 2003).

7 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 231 Second, mineral and commodity exports can cause economic consolidation and the Dutch disease, or rapid inflation and increases in the exchange rate of local currency, rendering other exports less competitive (Corden and Neary 1982). (Economists and historians call this the Dutch disease after North Sea oil and gas production had rapid negative effects on the exchange rate and inflation in the Netherlands in the 1960s). In essence, the primary resource being exported then crowds out other sectors of the economy, causing even more economic concentration. In response, regulators often adopt policies that protect that very sector to a greater degree, reinforcing dependence on the resource and worsening the cycle. The risk of economic consolidation becomes greater the more valuable the natural resource, for unlike other sources of wealth and manufactured goods, lucrative minerals need only be extracted rather than produced to attain revenue. This creates enclaved spheres of economic activity without major linkages to other sectors and can take place without large-scale participation of the labor force. Third, most extractive industries rely on sophisticated technology not owned or licensed by the host countries. Oil and gas production and mining, for instance, are very capital- and technology-intensive, meaning they create fewer jobs per dollar invested and require unique skills not well suited for other areas of economic activity. Resource-rich countries often send highly skilled labor abroad to train, and foreign workers are typically sent in to do the local extracting, minimizing the chance that exporting countries will learn by doing or own the intellectual property and technology associated with extraction. Fourth, since most extractive industries and technologies are large-scale, capital intensive and foreign owned, few productive links exist between resource extraction and the rest of the economy. Generally, the revenues flow directly to the government in the form of royalties, taxes, and profits. This consolidates revenue among the existing political parties, worsens the gap between rich and poor, and also breaks the link between taxation, representation, and accountability, since most resource-rich regimes do not need to tax their own people. It also increases the changes of regulatory capture and rent seeking, since only a small number of officials need to be captured to facilitate large transfers of wealth. The concentration of wealth from resource revenues sometimes enables rulers to spend their money on subsidies to friends, family, and political supporters. Karl (2000), for example, estimates that from 1974 to 2000 OPEC states spent 65 to 75 per cent of their income in this manner. Fifth and finally, large extractive projects frequently fail to benefit minorities and poor communities because these stakeholders have not participated in project selection and financing. The decision making processes in resource cursed countries tend to lack transparency and accountability, so vested interests determine what gets built and who benefits. Local communities and indigenous populations must then contend with the damages from resource extraction while the political and corporate elite reap

8 232 The Pacific Review the benefits. Moreover, when cost benefit analyses are conducted by those in power, benefits are usually overestimated publicly while costs underestimated. In the absence of effective oversight and regulation, companies lower their social and environmental standards to reduce costs and maximize short-term returns. In the hazardous extractive industries sector, the consequence of poor oversight can be severe and include acute and catastrophic spills, explosions, accidents, and disasters along with chronic and gradually deleterious exposure to noxious emissions, noise, degraded habitat, and food and water contamination (Steiner 2007; Sovacool 2008). Perhaps the extraction of oil and gas resources in Angola, Columbia, Nigeria, Saudi Arabia, and OPEC states as a whole bests exemplifies how the resource curse can occur at multiple scales of governance, in many degrees, and to a collection of different countries. In Angola, which has extracted some of the world s largest oil and gas reserves for 40 years, average life expectancy has plunged to 36.8 years in 2006, the infant mortality rate has risen to 19 per cent, the country imports most of its food, and almost three-quarters of the population live below the poverty line (Omorogbe 2006: 45). Proceeds from oil and gas production have nourished oppressive tendencies within the state. Oil revenues have enabled elites to capture virtually all economic enterprises, and encouraged violent contests for political power from a dissatisfied population. The United Nations has accused Angola of using state used oil revenues to fund blood diamond operations, and the state has admitted to using oil revenues to fund illegal cocaine operations throughout the Andean region (Auty 2004). In Columbia, oil and gas profits have contributed to acts of terrorism and intensified military conflict. Guerilla groups reap an oil war tax and earn $140 million per year from oil-related extortion and kidnappings. The National Liberation Army (ELN) and Revolutionary Armed Forces of Columbia (FARC), the two largest rebel groups, have extracted war taxes from oil companies and local contractors through threatened sabotage and murders, extortions, and bombings. When their threats have not gotten them the leverage they desire, ELN and FARC guerillas have dynamited state oil and gas pipelines more than 1,000 times in 13 years, spilling 2.9 billion barrels of crude oil and greatly damaging ecosystems and local water resources. The economic losses alone from explosions on the Cano Limon- Covenas pipeline amounted to $1 billion from 1990 to 1995, or seven per cent of Columbia s entire export revenue over the same period. Paramilitary groups have also built a cottage industry by stealing natural gas and gasoline by drilling holes in pipelines. Ecopetrol, one state oil company, estimates they lose $5 million per month. Finally, oil and gas revenues provide the Columbian Army with significant income needed to purchase weapons and train soldiers. A military tax of roughly $1 per barrel (or $12 to $30 million per year) allows the army to increase their troop presence and prolong their struggle against the rebels (Dunning and Wirpsa 2004).

9 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 233 In Nigeria, the thirteenth largest producer of petroleum in the world, oil and gas accounts for 80 per cent of all government revenues, 95 per cent of exports, and 90 per cent of foreign exchange earnings. Oil production has generated billions of dollars in government revenues, but the country s average GDP in the past decade was less than its GDP in the 1960s and the poverty rate increased from 27 per cent in 1980 to 70 per cent in 1999 (Omorogbe 2006: 44). Rather than bringing peace and stability, oil production has exacerbated political flashpoints. A collection of smaller Niger Delta states have attempted to expand their access and control over oil and gas resources, struggling with minority groups clamoring for sovereignty. Militant youth movements have grown from the resulting insecurity and promote intensified campaigns of interethnic violence (Watts 2004). Saudi Arabia is heavily dependent on oil production, and derives 85 per cent of its total export revenue from the oil industry, so much that every dollar increase in the price of a barrel of oil means a gain of about $3 billion to the Saudi treasury. The Saudis have used their oil wealth not to improve social welfare, however, and have instead consolidated it among a very few number of families. Five extended families in the Middle East own about 60 per cent of the world s oil, and the House of Saud controls more than a third of that amount. Per capita income in Saudi Arabia has fallen significantly from 1981 to 2001, and seven out of 10 jobs in the country are filled by foreigners (and close to nine out of every 10 private sector jobs). The disparity in income between the rich and poor has generated significant resentment towards the royal family, a resentment sharpened by the belief that the Saudis have failed to protect Muslims in Palestine and elsewhere and have been corrupted by their money. Adding to the social instability, Saudi Arabia has one of the highest birth rates in the world (37.25 births for every 1,000 citizens, compared to 14.5 in the United States), half the population is under 18 and another half is unemployed (Baer 2003). Finally, Karl (1998) analyzed the impact of oil production on the political economies of Venezuela, Nigeria, Mexico, Iran, Algeria, and Norway and found that in only one, Norway, did oil extraction correlate with positive development. In each of the other OPEC countries, bottlenecks in production, declines in other economic activity (especially in the agricultural and industrial sectors), capital flight, double digit hyperinflation, and sudden declines in living standards occurred. Karl hypothesizes that these countries have become addicted to oil, creating lavish subsidies for oil companies and pursuing oil profits over the public good. She concludes that production and dependence on oil has changed the entire political economy of each country. The per capita GDP for all OPEC members from 1965 to 1998, for instance, decreased by an average of 1.3 per cent each year whereas per capita GDP for all other developing countries that did not rely on oil exports as a whole grew by an average of 2.2 per cent over the same period (Karl 2005: 23).

10 234 The Pacific Review While these examples may be convincing, the resource curse is by no means inevitable. Australia, Canada, Chile, and the United States are all world leaders in mineral economies and manufacturing and have steadily improved per capita gross national product and living standards (Karl 2005). Sierra Leone and Botswana are both rich in diamonds, but the latter has had an average growth rate of 5.2 per cent between 1974 and 2002 while the former has plunged into civil war and negative growth (Stiglitz 2005: 13). In 1973, Indonesia and Nigeria had comparable per capita incomes, and both were dependent on oil revenues, but by 2002 Indonesia s per capita income was four times that of Nigeria s (Nigerian per capita income actually fell from $ in 1973 to $ in 2002) (Stiglitz 2005: 13). Clearly, political economy and governance play a role in how susceptible countries are to the resource curse. Bevan et al. (1999) explain that Indonesia fared much better than Nigeria in developing its resources and successfully minimized the resource curse because of three political and economic factors. Indonesian President Sukarno s articulation of a clear vision of social cohesion and fighting poverty in the 1950s and 1960s engendered expectations among the Indonesian population. This created a political environment where poverty alleviation could not be ignored, yet no such expectations took root in Nigeria. The occurrence of hyperinflation in the 1960s also made avoiding inflation an important social issue for Indonesians but not in Nigeria, and the rapid economic growth in the Asian region precipitated greater trade and investment for Indonesia while Nigeria s neighbors struggled economically. Moreover, the presence of five interrelated factors volatility, inflation, knowledge transfer, economic concentration, and lack of participation illustrate that the resource curse cannot be reduced to institutional structure, politics, distribution of natural resources, or standards of living. A combination of each plays a role, meaning the presence of the resource curse seems to be highly contingent and contextualized (Rosser 2006). 3. Measuring the resource curse in Southeast Asia Given its complex and contextual roots, how can the resource curse be measured? And is it occurring among Southeast Asian countries that have the largest oil and gas reserves and rates of production? The resource curse, according to the logic presented above, is ultimately connected to geologic factors such as a country s resource endowment; political factors such as transparency, participation, and corruption; economic factors such as diversification and wealth consolidation; and social factors such as life expectancy and standards of living. Countries suffering from the resource curse would be blessed with certain mineral resources, tend to have less transparency, more corruption, a less diversified economy and more concentration of wealth, and undergo deteriorating social indicators

11 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 235 Table 1 Oil reserves for Southeast Asia Country Total proven reserves (billion barrels) Per capita reserves (barrels) Cambodia 0 0 Brunei Indonesia Laos 0 0 Malaysia Myanmar Philippines Singapore 0 0 Thailand Vietnam Total/Average Source: U.S. Energy Information Administration (2007). Per capita reserves calculated by dividing total proven reserves by the total population listed in World Bank (2008). including higher rates of malnutrition, child mortality, and poor educational performance. In the case of Southeast Asia, the five countries with the largest endowment of combined oil and natural gas reserves (both per capita reserves and absolute reserves) and production are Brunei, Indonesia, Malaysia, Myanmar, and Thailand (see Tables 1 and 2). Thirteen metrics can be used to determine the extent that these countries resources have cursed them by measuring three political dimensions (transparency, authorized levels of oil and gas production), six economic dimensions (composition of exports, composition of GDP, government revenues, per capita income, rates of poverty, and rates of inflation), and four social dimensions (education, infant mortality, food security, and life expectancy) (see Table 3). For these countries to be suffering from the resource curse, we would expect transparency to decrease and corruption to increase from 1987 to The amount of oil and gas as a percentage of exports would also increase, along with overall oil and gas production (which would be harvested at faster rates to increase royalties, taxes, and profits). Oil and gas would be anticipated to contribute to more GDP as economic activity becomes concentrated, government revenues from the oil and gas sector would increase along with inflation and poverty, and per capita income would decrease. Finally, we would predict that social indicators such as literacy and life expectancy would deteriorate, while infant mortality and undernourishment would rise. Contrary to what the resource curse posits, none of the five countries examined exhibited deteriorating trends in most of these metrics. While natural gas and crude production did increase for all five countries, crude oil production remained almost flat in Brunei and exports actually remained

12 236 The Pacific Review Table 2 Natural gas reserves for Southeast Asia Proven reserves Possible reserves Probable reserves Total (trillion Per capita total reserves Country (trillion cubic feet) (trillion cubic feet) (trillion cubic feet) cubic feet) (million cubic feet) Cambodia Brunei Indonesia Laos Malaysia Myanmar Philippines Singapore Thailand Vietnam Total/Average Source: U.S. Energy Information Administration (2007). Per capita reserves calculated by dividing total (trillion cubic feet) by the total population listed in World Bank (2008).

13 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 237 Table 3 Metrics for measuring the resource curse Dimension Metric Explanation Political Transparency International Rating Measures the amount of corruption within a given country. The resource curse posits that corruption will increase along with resource extraction over time Political Natural gas production Measures the amount of domestic natural gas production in billion cubic feet. The resource curse predicts that gas production will increase over time so governments can extract more rents Political Oil production Measures the amount of domestic oil production in thousand barrels per day. The resource curse predicts that oil production will also increase over time so governments can extract more revenue Economic Fuels as % of all exports Measures the percentage of oil and gas in overall exports. The resource curse suggests that oil and gas will grow to constitute greater exports over time Economic % GDP in oil and gas Measures the extent that oil and gas contribute to gross domestic product. The resource curse implies that oil and gas will crowd out other areas of economic activity Measures the percentage of government revenues attributable to oil and gas taxes, Economic Oil and gas as % of government revenues royalties, and state-owned production. The resource curse posits that governments will become more dependent on oil and gas revenues over time Economic GNP/GNI per capita Measures the amount of gross national product or gross national income distributed to Economic % population below $2a day Economic Average annual rate of inflation each person of the country per year. The resource curse foretells a lowering of GNP/GNI per capita Measures the amount of people living in poverty (or less than US$2 per day). The resource curse would expect poverty to increase as knowledge and labor become concentrated in the oil and gas sector Measures the rate of inflation for each year. The Dutch disease aspect of the resource curse argues that inflation will increase over time (Continued on next page)

14 238 The Pacific Review Table 3 Metrics for measuring the resource curse (Continued) Dimension Metric Explanation Social Adult literacy rate Provides a crude measurement of the government s ability to provide education. The resource curse posits that governments will become less able to provide basic services such as education over time Social Infant mortality (per Offers a rough measurement of health care, which would also be expected to decline in thousand live births) Social Prevalence of undernourishment (% population) quality as governments become resource cursed Indicates a general measurement of the availability of food within a country. The resource curse suggests that undernourishment will increase as governments become corrupt and less competent Social Life expectancy (age/yrs) Perhaps the most general of all indicators, it measures the overall quality of life by tracking how long people live. The resource curse suggests that life expectancy will deteriorate

15 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 239 Figure 2 Bruneian natural gas and crude oil production and exports, 1980/ roughly constant or fell from 1987 to 2007 for Indonesia and Thailand (see Figures 2 6). The 13 metrics, furthermore, quite clearly show that most political, economic, and social indicators improved for all five countries over the same period (see Table 4). While Myanmar performed the worst of all five countries analyzed, for it was most reliant on oil and gas revenues and the least diversified, even it displayed improvements in education, life expectancy, and per capita GDP. When given a score for how well the group of five Southeast Asian countries performed according to each metric, with a 1 indicating a deteriorating metric and a +1 signifying an improvement, they collectively showed progress in 47 indicators and regressed in only 17 (data for two metrics did not exist). However, the resource curse is not only about whether resource-rich countries will perform poorly in absolute terms, but also how well they do

16 240 The Pacific Review Figure 3 Indonesian natural gas and crude oil production and exports, 1980/ /2007. compared to other countries, both resource rich and average. To typify resource rich countries, Table 5 presents data relating to the same political, economic, and social metrics for the five largest oil producers of OPEC: Saudi Arabia, Iran, Venezuela, Kuwait, and United Arab Emirates from 1987 to These five countries produced about 70 per cent of OPEC s daily output of crude oil in 2007, and they also represented 43 per cent of the world s total daily output. To exemplify average developing countries, Table 6 presents equivalent data for five countries that are still developing but also relatively industrialized: Brazil, China, India, Russia, and South Africa. These countries are often grouped together as newly industrializing and framed as definitively developing (Bozyk 2006). Interestingly, relying on the same scoring technique, the five OPEC countries as a whole progressed in 43 metrics and regressed in 17 (data for

17 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 241 Figure 4 Malaysian natural gas and crude oil production and exports, 1980/ /2007. five metrics was not available), and the group of developing countries progressed in 56 metrics and regressed in only nine (see Figures 7 and 8). When metrics are compared between groups of countries, a few trends emerge: In terms of transparency and political openness, Southeast Asian countries performed better than both OPEC countries and developing countries, scoring the highest average as a group on Transparency International s corruptions perception index; Southeast Asian average production of crude oil remained relatively flat (slightly less than 420,000 barrels per day in both 1987 and 1997) compared to massive increases in production among OPEC members (a jump of more than 1.7 million barrels per day over the same period), yet natural gas production remained comparable to OPEC s (off

18 242 The Pacific Review Figure 5 Myanmari natural gas and crude oil production and exports, 1980/ /2007. by only about 200 billion cubic feet per day for two decades) and both were dwarfed by production from the developing countries, mostly due to Russia, which produced more natural gas than the OPEC and Southeast Asian countries combined; A significant amount of GDP and government revenues among Southeast Asian countries came from the oil and gas sectors, making them similar to OPEC countries (both sets of countries average more than one-quarter of GDP and revenues from oil and gas) but unlike the developing countries (which received less than 10 per cent); The most remarkable gains in gross national product (GNP) per capita, fighting poverty, and keeping annual interest rates low occurred in the OPEC countries, which saw incomes rise over the period by $8,844

19 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 243 Figure 6 Thai natural gas and crude oil production and exports, 1980/ /2007. compared to $3,660 for Southeast Asian countries and only $2,746 for the developing countries; All three groups of countries made progress on social indicators but the developing countries had higher rates of child mortality (one-third higher than the combined rates for Southeast Asia and OPEC) and lower life expectancy (9 per cent lower). While this type of scoring is admittedly crude, it suggests that Southeast Asian countries performed slightly better than the OPEC countries but slightly worse than the developing countries in the metrics related to the resource curse, although each group of countries made significant strides towards improving various political, economic, and social metrics.

20 244 The Pacific Review Table 4 Resource curse indicators for Brunei, Burma/Myanmar, Indonesia, Malaysia, and Thailand, Confirm resource curse? (Y/N) Average transparency rating 57/85 108/180 N Brunei Burma/Myanmar 178/180 Indonesia 80/85 126/180 N Malaysia 29/85 47/180 N Thailand 61/85 80/180 N Average fuels as % of all exports 34% 29% 35% N Brunei 97% 88% 96% N Burma/Myanmar 3% 0% 36% Y Indonesia 49% 39% 28% N Malaysia 18% 17% 13% N Thailand 3% 1% 4% N Average natural gas production (billion cubic feet) Y Brunei Y Burma/Myanmar Y Indonesia 1,300 2,400 2,600 Y Malaysia 520 1,400 2,250 Y Thailand Y Average oil production (thousand barrels per day) N Brunei N Burma/Myanmar N Indonesia 1,400 1,500 1,100 N Malaysia N Thailand Y

21 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 245 Average % GDP in oil and gas 27% 21% 26% N Brunei 80% 34% 40% N Burma/Myanmar 10% 15% 19% Y Indonesia 8% 8.6% 10.8% Y Malaysia 31% 39% 51% Y Thailand 4% 6% 11% Y Average oil and gas as % of government revenues 30% 44% Y Brunei 95% 75% 94% N Burma/Myanmar 37% 54% Y Indonesia 56% 20% 25% N Malaysia 10% 35% Y Thailand 10% 11% N Average GNP/GNI per capita 3,702 5,564 7,362 N Brunei $12,170 $16,310 $26,740 N Burma/Myanmar $485 $451 $616 N Indonesia $762 $1,429 $1,349 N Malaysia $3,359 $6,025 $5,237 N Thailand $1,732 $3,605 $2,866 N Average % population below $2 a day 38% 36% Y Brunei 2% 1% N Burma/Myanmar 30% 91% Y Indonesia 59% 52% N Malaysia 27% 9% N Thailand 24% 25% N Average annual rate of inflation 8% 10% 15% Y Brunei 11% 9% 1% N Burma/Myanmar 13.8% 25% 54% Y Indonesia 8.5% 11% 13% Y Malaysia 1.3% 2.7% 2.3% N Thailand 3.1% 4% 2.3% N (Continuedonnextpage)

22 246 The Pacific Review Table 4 Resource curse indicators for Brunei, Burma/Myanmar, Indonesia, Malaysia, and Thailand, (Continued) Confirm resource curse? (Y/N) Average adult literacy rate 72% 88% 92% N Brunei 73% 91% 95% N Burma/Myanmar 49% 83% 90% N Indonesia 74% 84% 91% N Malaysia 73% 84% 89% N Thailand 91% 94% 93% N Average child mortality (per thousand life births) N Brunei N Burma/Myanmar N Indonesia N Malaysia N Thailand N Average prevalence of undernourishment (% population) 11% 8% N Brunei 10% 4% 4% N Burma/Myanmar 10% 5% N Indonesia 9% 6% N Malaysia 3% 3% N Thailand 30% 22% N Average life expectancy (age/yrs) N Brunei N Burma/Myanmar N Indonesia N Malaysia N Thailand N Source: World Bank (1990); World Bank (1999); International Monetary Fund (2006); World Bank (2008); Transparency International (2008); U.S. Energy Information Administration (2008b). All monetary figures have been updated to $2007. Average column computed as the mean of the five numbers for each column.

23 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 247 Table 5 Resource curse indicators for five OPEC countries, Confirm resource curse? (Y/N) Average transparency rating 93/180 Saudi Arabia 79/180 Iran 131/180 Venezuela 77/85 162/180 N Kuwait 60/180 United Arab Emirates 34/180 Average fuels as % of all exports 78% 78% 83% Y Saudi Arabia 81% 89% 88% N Iran 90% 86% 85% N Venezuela 60% 80% 91% Y Kuwait 88% 89% 95% Y United Arab Emirates 73% 45% 58% N Average natural gas production (billion cubic feet) 607 1,173 1,946 Y Saudi Arabia 946 1,601 2,680 Y Iran 565 1,660 3,952 Y Venezuela N Kuwait Y United Arab Emirates 682 1,282 1,723 Y Average oil production (thousand barrels per day) 2,288 3,926 4,027 Y Saudi Arabia 4,265 8,362 8,722 Y Iran 2,298 3,664 3,912 Y Venezuela 1,752 3,280 2,433 N Kuwait 1,585 2,007 2,464 Y United Arab Emirates 1,541 2,316 2,603 Y (Continuedonnextpage)

24 248 The Pacific Review Table 5 Resource curse indicators for five OPEC countries, (Continued) Confirm resource curse? (Y/N) Average % GDP in oil and gas 33% 32% 34% N Saudi Arabia 60% 41% 32% N Iran 8% 20% 33% Y Venezuela 18% 22% 30% Y Kuwait 42% 45% 50% Y United Arab Emirates 36% 33% 25% N Average oil and gas as % of government revenues 65% 69% 67% N Saudi Arabia 90% 81% 75% N Iran 70% 65% 58% N Venezuela 45% 47% 52% Y Kuwait 43% 75% 80% Y United Arab Emirates 76% 75% 70% N Average GNP/GNI per capita 9,406 10,436 18,250 N Saudi Arabia $7,670 $8,410 $15,470 N Iran $3,320 $1,770 $3,540 Y Venezuela $3,160 $3,370 $7,550 N Kuwait $16,000 $18,560 $38,420 N United Arab Emirates $16,880 $20,070 $26,270 N Average % population below $2 a day 16% 19% 13% N Saudi Arabia 15% 8% 7% N Iran 7% 3% 8% N Venezuela 30% 61% 31% N Kuwait 3% 1% 1% N United Arab Emirates 25% 22% 19.5% N Average annual rate of inflation 15% 10% 14% N Saudi Arabia 4% 2% 4% N Iran 23% 14% 20% N

25 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 249 Venezuela 33% 38% 14% N Kuwait 11% 5% 18% N United Arab Emirates 3% 0% 12% N Average adult literacy rate 72% 81% 89% N Saudi Arabia 79% 80% 83% N Iran 51% 76% 85% N Venezuela 87% 92% 96% N Kuwait 70% 79% 93% N United Arab Emirates 71% 79% 89% N Average child mortality (per thousand life births) N Saudi Arabia N Iran N Venezuela N Kuwait N United Arab Emirates N Average prevalence of undernourishment (% population) 10% 6% 7% N Saudi Arabia 3% 4% 4% N Iran 7% 3% 4% N Venezuela 11% 15% 18% Y Kuwait 24% 5% 5% N United Arab Emirates 4% 3% 3% N Average life expectancy (age/yrs) N Saudi Arabia N Iran N Venezuela N Kuwait N United Arab Emirates N Source: World Bank (1990); World Bank (1999); International Monetary Fund (2006); World Bank (2008); Transparency International (2008); U.S. Energy Information Administration (2008b). All monetary figures have been updated to $2007. Average column computed as the mean of the five numbers for each column.

26 250 The Pacific Review Table 6 Resource curse indicators for five average developing countries, Confirm resource curse? (Y/N) Average transparency rating 54/85 80/180 N Brazil 46/85 72/180 N China 52/85 72/180 N India 66/85 72/180 N Russia 76/85 143/180 N South Africa 32/85 43/180 N Average fuels as % of all exports 7.3% 8.4% 12.8% N Brazil N China 3% <1% N India <1% N Russia 33.5% 42% 64% Y South Africa <1% N Average natural gas production (billion cubic feet) 4,486 4,376 5,416 N Brazil Y China ,446 Y India ,119 Y Russia 21,604 20,168 23,064 N South Africa N Average oil production (thousand barrels per day) 2,318 2,164 3,212 N Brazil ,748 Y China 2,690 3,200 3,729 Y India Y Russia 7,724 6,102 9,874 N South Africa N

27 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 251 Average % GDP in oil and gas 5% 4% 4% N Brazil <1% N China 4% <1% N India <1% N Russia 20% 22% 20% N South Africa <1% N Average oil and gas as % of government revenues 5% 5% 8% N Brazil <1% N China 2% <1% N India <1% N Russia 21% 25% 40% Y South Africa <1% N Average GNP/GNI per capita $1,740 $2,514 $4,486 N Brazil $2,040 $5,060 $5,860 N China $320 $750 $2,370 N India $360 $420 $950 N Russia $3,310 $2,660 $7,530 N South Africa $2,670 $3,680 $5,720 N Average % population below $2 a day 62% 43% 31% N Brazil 45% 23% 18% N China 68% 47% 18% N India 81% 75% 73% N Russia 53% 25% 12% N South Africa 61% 45% 34% N Average annual rate of inflation 49% 8% 8% N Brazil 204% 8% 4% N China 5% 2% 7% N India 9% 6% 5% N Russia 14% 15% 14% N South Africa 14% 8% 9% N (Continuedonnextpage)

28 252 The Pacific Review Table 6 Resource curse indicators for five average developing countries, (Continued) Confirm resource curse? (Y/N) Average adult literacy rate 77% 81% 87% N Brazil 80% 85% 90% N China 81% 83% 91% N India 48% 54% 66% N Russia 96% 99% 100% N South Africa 79% 84% 88% N Average child mortality (per thousand life births) N Brazil N China N India N Russia N South Africa N Average prevalence of undernourishment (% population) 16% 11% 9% N Brazil 11% 10% 6% N China 15% 12% 9% N India 44% 24% 21% N Russia 5% 5% 4% N South Africa 5% 5% 5% N Average life expectancy (age/yrs) N Brazil N China N India N Russia N South Africa Y Source: World Bank (1990); World Bank (1999); International Monetary Fund (2006); World Bank (2008); Transparency International (2008); U.S. Energy Information Administration (2008b). All monetary figures have been updated to $2007. Average column computed as the mean of the five numbers for each column.

29 B. K. Sovacool: Political economy of oil and gas in Southeast Asia 253 Figure 7 Performance evaluation of average resource curse metrics for Southeast Asian, OPEC, and developing countries. Notes: SEA = Brunei, Burma, Indonesia, Malaysia and Thailand; OPEC = Saudi Arabia, Iran, Venezuela, Kuwait and UAE; DEV = Brazil, China, India, Russia and South Africa. 4. Explaining Southeast Asian performance The factors underpinning the success of the key Southeast Asian countries do imply that certain social and political variables shape economic development and welfare outcomes. With this in mind, the following section suggests that three factors, some unique to Southeast Asia, may explain why they have avoided the resource curse so far: diffusion of resources, political institutions, and cooperative production. First, point resources such as oil fields tend to be easily dominated by oligarchs and elites, but in Southeast Asia oil and gas reserves are more diffused over a broader geographic area than in other regions. Indonesia s oil and gas fields, for example, are spread across Riau province in Central Sumatra, the Java Sea, parts of the western Pacific Ocean, East Kalimantan, and Natuna, reserves that occupy more than 433,000 square kilometers of onshore and offshore land. Other oil and gas fields, like those in the Gulf of Thailand, are jointly extracted by Malaysia and Thailand. The exceptions

30 254 The Pacific Review Figure 8 Average resource curse metrics compared between Southeast Asian, OPEC, and developing countries. Source: Figures taken from Tables 1, 2, and 3. Notes: Transparency rating is out of 85 countries for 1997 and 185 countries in 2007 (the higher the ranking the poorer the score). Natural gas production is in billion cubic feet. Crude oil production is in thousand barrels per day. GNI/GNP per capita has been updated to $2007. Poverty rate is defined as the percentage of people living on less than $2 per day. Child mortality refers to number of children under the age of 5 that die every year out of 1,000.