The volatility of hydrocarbon prices

8The sharp fluctuations in world oil prices, causing both boom and bust cycles, make effective management of the black gold by those in power uncertain. The volatility of prices on the international market has a negative effect on budgetary discipline, state planning and sustainable development.

The “Dutch Disease”

9Countries dependent on hydrocarbons suffer from “Dutch Disease”. Diagnosed following the negative effects on the manufacturing industry in the Netherlands caused in the 1970s by the rise in gas exports, the disease can be described as follows: the massive flow of oil revenues causes a rise in the real exchange rate of the country’s currency, subsequently making exports from the non-hydrocarbon sectors uncompetitive. Consequently, by marginalising the non-oil related production activities such as the agricultural and manufacturing sectors, oil and gas exports are completely detrimental to the diversification of the economy.

10In order to support these uncompetitive sectors of economic activity, those in power finance protectionist policies, via the financial transfer of petrodollars. This brings the whole cycle full circle: the more agriculture and manufacturing are dependent on the oil sector, the greater the oil dependency and the lower the competitiveness. [32]

Wealth concentration and “rent-seeking”

11The oil industry, which requires a very high level of capital, expertise and technological skills, creates only a small amount of jobs. Weakly linked to other production activities, the oil sector is an “enclave” which keeps itself apart from the rest of the economy. [33] Development founded on mineral and oil resources, in generating significant profits, marginalises other sources of revenue and encourages “rent-seeking” – the concept referring to efforts aimed at accessing or controlling the rents. In these states, where administrative structures are experiencing enormous expansion, clientelism governs the distribution of resources. [34] This form of exchange, while fostering the concentration of wealth in the hands of the few, at the same time heavily penalises production activities.

“Rentier state”

12States dependent on hydrocarbons and drawing their resources from rent revenues do not need to develop an effective fiscal bureaucracy. Consequently, they have an obvious institutional weakness, notably with regard to fiscal accountability and political representation. By loosening the vital ties between the government and the population, development built on rentier wealth promotes authoritarianism, with oil revenues giving governments the means of financing a tough repressive system.

The biases of correlational studies

14Correlational analysis, the basis for a large number of explanations, occupies an increasingly important place in sociology and political science. Based on the assumption that correlation is at the heart of causality, many view the theories put forward by correlational analyses – the statistical models of which are, it is true, highly sophisticated – as offering, by definition, guarantees of rigour and objectivity. [37] Such is the case for the “resource Curse” theory. The correlational analysis at the heart of this has, however, numerous biases. The most important of them relates to the very epistemic foundations of the approach: establishing a correlation between two variables is one thing, to then impute a causal relationship is another. On closer inspection, can a correlation not be read in one way, just as much as in another? In this instance, is it the rent that causes authoritarianism or the authoritarian government that generates dependency on natural resources? The seminal study by Michael Ross and those which have succeeded it, [38] apparently overlook this fundamental point of method: a correlation is not a causality. A second bias undermines these canonic studies: it concerns the conceptual confusion that they foster between “abundance” of natural resources and “dependency” on resources. [39] There are numerous countries that are both wealthy in natural resources and democratic. Australia, South Africa (post-Apartheid), Canada, the United States of America and Great Britain are some of the many examples. [40] The third bias which compromises the quantitative studies of the “law” of the “resource curse” concerns the limits of the indices employed. Their authors unanimously use indicators such as the Polity index, the ratio of military spending in relation to GDP. [41] These indicators, taking the state to be a holistic and homogeneous entity, struggle to grasp the social and political dynamics, or even succeed in providing a static view of the country in question. These indices fail, in this respect, to capture the underground logics, such as factionalism and ethnicity, which relate to the segmentation of the state and society. [42] Moreover, the indices come up against a seldom discussed theoretical problem: that of conceptualisation. [43] Take the “democracy” variable for example. If one integrates – as would a maximalist definition of the concept – the market economy as an attribute within the definition of democracy, the links between these two categories cease to be a research problem. Conversely, the Polity index’s omission of the “participation” criterion in the quantitative assessment of democracy is an issue. The index does not de jure take account of the restrictions on the right to vote in force before 1945. Nor does it measure the obstacles which de facto encumber the effective use of the formal right to vote. [44]

15The correlational analyses which have made the “resource curse” theory so successful have another weighty omission: the key variable of time is absent. The omission is all the more prejudicial to the analysis given that temporality is crucial in order to establish causality, here as elsewhere. Based on a database arbitrarily established between 1971 and 1972, these studies do not reflect the anteriority of authoritarianism. Moreover, some regressive quantitative analyses which go back a century before the first oil boom contradict the causal relationship accepted by supporters of the “resource curse” theory. [45]

16These biases are enough to call into question the relevance of correlational analyses, and argue in favour of going beyond them. [46]

Does rent lead to authoritarianism?

17The “rentier state” theory aims to contribute to a political economy approach. Supporters of the paradigm, in giving in however a little too easily to economic determinism, manage to gloss over the political aspect. Thus, they spend little or no time on the question of political legitimacy [47] – the acute nature of which is clearly demonstrated today by the Arab popular uprisings. How did the elite seize power? How are hydrocarbon revenues distributed? How do groups and individuals view the distribution of rights and resources? Popular wisdom states that hydrocarbons shaped the “petro-states” [48] and the “hydrocarbon societies”, [49] and ignores a variable as decisive as the rules of the political game, when the latter, often predating the oil era, pre-empt how the rent is used. [50] The avoidance of institutional and temporal variables produces an error in interpretation that has far-reaching consequences: this being making rent per se the cause – fons et origo – of authoritarianism. There is no shortage of examples to the contrary, which show that authoritarianism often pre-dates oil dependency. Take the case of Iran’s contemporary history which has served as a founding model. When Mohammad Réza Pahlavi came to power in 1941, the Iranian state was only a very modest oil producer, with hydrocarbon exports accounting for no more than 1% of government revenue. [51] The Shah with his absolutism, however, did not wait until the first oil boom before cracking down. [52] The same can be said for Oman and Yemen: authoritarianism was present there prior to the dawn of the oil era – which occurred in the 1970s for the Qaboos Sultanate and during the 1980s for Saleh’s republic. [53] It is the same story in Algeria where authoritarianism did not wait for oil dependency – which moreover did not appear until the early 1970s, [54] or a decade after the installation of the praetorian regime – before establishing itself. [55] Gabon, Equatorial Guinea, Kazakhstan, Turkmenistan and Tajikistan are included here: authoritarianism is to be found in these countries and was established in each one prior to the dawn of the oil era. [56]

18Reductio ad absurdum is a good way of testing the validity of a theory. In the case under discussion, a similar undertaking would consist of demonstrating that those countries not dependent on natural resources are governed by democratic regimes. The test here again proves inconclusive: from Tunisia to Uzbekistan, via Eritrea, there is no shortage of examples to the contrary. Let’s take this argument to the extreme and envisage a counterfactual hypothesis. The argument that makes hydrocarbons the cause of authoritarianism implicitly suggests that those countries dependent on natural resources would have been democratic had they not been endowed with these riches. So would the monarchies of the Arabian Peninsula, where the per capita income is comparable to that of Scandinavian countries, have had political institutions similar to those of Denmark and Sweden if they had not had oil resources? Would Libya, Saudi Arabia and Bahrain, have become democratic without the oil? To reply in the affirmative, as certain advocates of the “resource curse” theory do implicitly, [57] is to ignore the fact that undemocratic forms of government have, in the area to which these countries belong, a historical depth that substantially pre-dates the first oil boom. If the historical legacy on its own cannot justify contemporary authoritarianism, nor can the latter be reduced to the dependent variable of an abundance of natural resources.

No representation without taxation?

19“Rentier state” theorists, referring to the course followed by Western European history, make representation the inevitable outcome of taxation. A rigorous review of the European history of political representation, however, invites this logic to be challenged. [58] Indeed, the supporters of the thesis according to which, in Europe, “taxation leads to representation” [59] in truth mix up three distinct historical phases: the emergence, the consolidation and the longevity of representative institutions. These authors – who adopt the “Whig” interpretation of the history of representation, [60] claiming to show the necessary and universal evolution of an institution towards progress – transpose the achievements of the elected parliaments of the nineteenth century (restriction of monarchical absolutism, institutionalisation of parliamentarism and of governmental accountability) onto those of the representative institutions of the eighteenth century. This view, adopted by the “rentier state” therorists, makes at least two mistakes: the first consists of mistaking the search by the powerful medieval representative institutions for autonomy as accountability; the second is to make a propensity for conflict the sole parameter capable of defining the relationships between representative institutions and monarchies, to the exclusion of other variables such as co-operation and cooptation. Great historians of the pre-modern assemblies, such as Francis Carsten and Russell Major nevertheless clearly demonstrated this: taxation contributed to the emergence and the longevity of representative institutions principally where these had a direct role in the collection of taxes. [61] The representative institutions set up in the period between the French Revolution and the proclamation of the Second French Republic did not enjoy this bargaining and negotiating power that the pre-modern assemblies were able to derive from taxation, the executive being, in the modern state, the sole institution that fixes and levies tax through the fiscal bureaucracy. And so, ultimately, taxation only played a “minor” role in the emergence, consolidation and longevity of modern representative institutions. [62]

20The central thesis on which the “rentier state” theory is founded is all the more uncertain given that the review of the twentieth century does not provide any additional elements in support of the argument that democracy arises solely from the bargaining that takes place between the government and the governed around taxation (the representative bargain). In a landmark text, John Waterbury updated the anomalies in the “rentier state” theory, [63] which I shall now review. In India, the oldest and largest democracy in the Third World, indirect contributions usually account for three quarters of the total federal state revenue. [64] In Turkey, the proportion of the GNP that fiscal revenue represented fell from 17% in 1982 to 14% in 1988. It was, however, during this period that the country successfully undertook its (third) democratic transition. [65]

21If it is undeniable that the majority of states in North Africa and the Middle East rely more on rents than on taxes, it would, on the other hand, be wrong to think that the Arab world is undertaxed in relation to other developing regions. Figures from the World Bank show that between 1975 and 1985 taxes accounted for 12% of the GNP in Latin America as opposed to 25% in the Middle East. [66] It is, however, in South America that processes of “transition to democracy” were underway during this period and not in the Maghreb or Machrek regions. Clement M. Henry also calls into question the central thesis of the “rentier state” theory. Doubting the effectiveness of the mechanism by which “taxation leads to representation”, he points to the fact that in Tunisia, fiscal revenues represented close to 26% of the country’s GDP. However, this high rate of taxation under the Ben Ali regime was associated with more authoritarianism and not at all with greater governmental accountability. [67] To suggest that taxation (in other words the absence of the “rentier state”) was the reason for the recent “Jasmine Revolution”, as one would if just skim reading the situation, would be to ignore the high level of Tunisian household debt: credit had been at the heart of the political economy of the Ben Ali autocracy. [68] Waterbury, who rightly doubts the “alchemy of taxation”, concludes: “The point is that the relative absence of conventional institutions of accountability in the Middle East cannot be attributed to a weak tax effort. This effort has not been, in historical or comparative terms, weak, but neither historically nor in the twentieth century is there much evidence that taxation has evoked demands that governments account for their use of tax monies. Predatory taxation has produced revolts, especially in the countryside […], but there has been no translation of tax burden into pressures for democratization.” [69]

22While the aphorism “no representation without taxation” supposes that political representation is the necessary condition for taxation, supporters of the “rentier state” theory suggest the reverse logic. To make taxation the condition of possibility for representation amounts to supporting the idea that there is ultimately only one path, the one best way, an “ideal” or even “unique” path, [70] which is capable of leading to democracy: this being that of taxation bargaining (the “representative bargain”) between governments and the governed. [71] By postulating, on the basis of a simplistic and erroneous interpretation of the history of political representation in Europe, that “the fact is that there is ‘no representation without taxation’ and there are no exceptions to this version of the rule”, [72] the supporters of the “rentier state” theory fall foul of determinism and linearity. This preconception, which is reminiscent of the one that affects the tireless prospecting for the economic and social “prerequisites” of democracy, is characteristic of work on authoritarianism in the Arab world; [73] it “leads researchers to focus on that which is missing, on that which has not happened, and to chase after a non-fact (the absence of an element which, all things being equal, “should have” been there) which would be the central analytical issue to resolve”. [74] Many different paths however have been taken to build a democracy, such as “imitation” (India); the “cascades” effect (the Eastern Europe countries in 1989); “collapses” (Russia in 1991); “revolutions” (Portugal); “external imposition” (West Germany and Japan); and “negotiated settlements” (South Africa). [75]

Is rent inconsistent with democracy?

23The advocates of the “resource curse” thesis concede an “exception” to the “law” that they updated, namely Norway, [76] where the discovery in 1962 of oil in the North Sea did not bring about the fall of democracy and the subsequent establishment of authoritarianism. [77] Botswana is also considered by some as an “African anomaly”: the rent derived from diamond mining has not altered the institutionalisation of participation and opposition, which has been a constant element since the country’s independence in 1966. [78] However, there is more. Indeed, examples to the contrary are far more numerous and significant than the promoters of the “resource curse” theory will (be made to) admit to. There are many counter-examples in South America. Venezuela, for example, long presented as both the oldest “petro-state” and the most stable democracy on the South American continent, invalidates this “law” that “rent promotes authoritarianism”: the government’s dependency on oil revenues has in no way prevented the institutionalisation of a democratic regime. [79] Several interpretations are put forward by experts to account for Venezuela’s “oil democracy”. Daniel Levine emphasizes the importance of two factors: firstly, the political apprenticeship undergone by the parties (notably AD and COPEI) after the end of the “Trienio” democratic interlude (1945-1948); and secondly, the robustness of the political institutions set up during the second democratic transition – by the “Pacto de Punto Fijo” of 1958 and the Constitution of 1961. [80] Terry Lynn Karl highlights, with regard to this oil democracy, the structural changes introduced by the rentier economy, foremost among which is the emergence of a counter-elite, the new bourgeoisie, which has become attached, in its own economic interests, to the democratic institutions. [81] Others suggest that rent has encouraged the introduction of a positive sum game in which there are no losers. [82] Thad Dunning provides an additional explanation: in the highly inegalitarian societies of South America, rent – which constitutes a substantial share of the state’s revenues while at the same time accounting for a limited portion of the gross domestic product [83] – helps to dampen the redistributive conflict and promote democratisation. [84] While taxing the non-rent related wealth (produced by the capitalist elite) tends – as was the case in Chile and Argentina – to push the oligarchy into supporting military coups, in Venezuela, the distribution of the rent revenues within public services (health, education, etc.) has, on the other hand, greatly helped to amortise the redistributive cost of democracy, to the benefit of the capitalist bourgeoisie. [85] The theory developed in Dunning’s Crude Democracy aims not only to account for the stabilisation but also the destabilisation of Venezuelan democracy. For Dunning, the oil bust in the years between 1985 to 2002, by hastening the re-emergence of redistributive taxation as a salient dimension of polarisation, intensified class conflict and exacerbated the crisis of the traditional political parties. The election of Hugo Chavez as president in 1998 was consistent with this context of crisis within the political party system. The new president – an anti-establishment candidate elected by a majority of the working class and part of the middle class – initially positioned his economic and redistributive policy as representing a “substantial continuity” with the previous periods. The combined effect of the decline in oil revenues and the fall in popularity of the new president forced the latter to radicalise, from 2001, the content of his economic policy. The return to redistributive taxation, in a context of low oil revenues, was perceived as an economic threat by the private sector elite, some of whom went as far as to support the aborted coup of 2002. The oil boom which occurred at the beginning of 2003 changed the situation however. Three salient aspects can be identified. The influx of rent resources into the state’s coffers has enabled the government to finance a vast programme of public spending, essentially allocated to social services; [86] the state-owned oil company too is part of this effort: charged with financing social expenditure, PDVSA now supports several programmes – or Misiones (Missions) – which are focused on education, the subsidising of food products (via the famous Mercal co-operative), assisting indigenous communities, etc. The second aspect is less well-known: the reduction, in the shadow of chavismo, of the cost of redistributive taxation. Lastly, if the oil boom has undeniably increased the centralisation of political power and encouraged the decline of accountable institutions, it has not, despite this, caused the collapse of the democratic regime. The rejection, by referendum, of the constitutional amendments proposed by Hugo Chavez in December 2007 – not least the removal of the restriction on the number of presidential terms – is proof of this: the result, immediately accepted by the government, reflects the commitment of the electorate to the democratic regime and the increasing level of electoral competition. [87]

24Venezuela is not an idiosyncratic case of “rentier democracy”. Chile, Bolivia and Ecuador prove that they too have an equally positive relationship between rent and democracy. Chile, with two distinct sources of rent, sodium nitrate and copper, exploited in different periods – the first between 1880 and the mid-1920s, the second since the Second World War – is a state that some would readily qualify as “rentier”, [88] where electoral politics are relatively competitive and stable, and in which military intervention – until the military coup of 1973 – played a somewhat limited role. [89] Nitrate, which between 1889 and 1990 represented more than 45% of the state’s revenues, enabled the government to finance social services such as education, health, transport, and urbanisation. While social spending grew continuously during the nitrate boom – increasing from 7.5% of overall central government expenditure between 1865 and 1880 to 12.28 between 1890 and 1910 – ordinary taxation declined considerably, plunging from 23.1% in 1845 to 0.5% in 1900. [90] Rent, in reducing the level of taxation for the oligarchy of big landowners, defused the redistributive conflict and stabilised the social order. The nitrate mining sector crisis in the 1920s, however, forced governments to return to redistributive taxation, with the share of government revenue represented by ordinary taxation increasing from 3.5% in 1910 to 30% in 1930. [91] The decline of the nitrate rent, in provoking the rise of social movements on one hand, and the exacerbation of the redistributive conflict on the other, pushed the traditional property-owning elite to favour the authoritarian solution. The role of the military during this period (1924-1931) is complex however: it enabled, on one hand, the labour movement to be controlled, and on the other, the implementation of tax reforms that the elite did not want. The return to democracy in the context of the Great Depression, while benefitting the Popular Front – the coalition uniting the communists, socialists and the centre-left – was characterised by a “policy of moderation”. [92] Copper rent, which supplanted that of nitrate from the Second World War, played an important role here, reflected in – nolens volens – the rise of the public sector, which, prior to the Allende administration, represented 40% of national production and 60% of total investment; [93] the financing of major social policies; and the predominance, until the mid-1960s, of the centre parties. [94] From the second half of the 1960s, the decline in rentier revenues served to exacerbate the conflicts inherent in the redistribution of nonrentier wealth. Implemented under the Allende presidency, the nationalisation of copper – of which one quarter of the world’s reserves are found in Chile – did not manage to halt the decline in rentier revenues. Land reform, launched by the Christian Democrat Eduardo Frei (1964-1970) and furthered by the Socialist Salvadore Allende, exacerbated the elite’s opposition to the redistributive policies. Thus class conflict was at its most intense when the military junta decided to overthrow the democratic regime via the orchestration of the explicitly anti-redistributive coup of 11 September 1973. [95]

25Ecuador is also an extremely interesting counter-example, in as much as it undermines a key argument of the “rentier state” and “resource curse” theories: the argument that rent gives a considerable advantage to those holding power (incumbency advantage). [96] The Ecuadorian state, taking advantage of the Arab oil embargo, did indeed substantially increase its oil exports between 1973 and 1976, with the rise in oil revenue increasing its national budget by 32% over this period. [97] However, while the military coup – staged on 15 February 1972 with the support of the economic elite in order to prevent the potential election of a left-wing “Marxist” candidate as president [98] – preceded the first boom by a year, the oil did not, contrary to this “rule” that “rent promotes authoritarianism”, perpetuate the praetorian regime. In fact, it was in 1978, in other words at the moment where the state’s “autonomy” was at its peak, that the ruling elite decided – after two years of preparations – to return to democracy. [99]

26These counter-examples, far from being anomalies, invalidate the “rule” established by the advocates of the “rentier state” and “resource curse” theories: rent is not in itself inconsistent with democracy; the effects of rent on democracy depend on the rules of the game, and on the established political institutions. [100]

Can rent “buy” legitimacy?

27Supporters of the “rentier state” theory repeat the refrain that the allocation of the profits of rent to the population enables the state to “buy” “popular consent”. In erasing the institutional foundations of regimes prior to the oil era, they forget, for example, that the legitimacy of the Algerian ruling elite is derived from the glory of national liberation after 132 years of colonisation, and not from the repercussions of the 1973 oil boom. Moreover this explanation, in limiting the expression of popular consent to the topic of rent benefit distribution, takes no account of any other ways of speaking out: for example the refusal to allow a foreign military presence on national soil; denunciation of the governmental attitude towards the Palestinian question; condemnation of large-scale corruption, etc. Gwenn Okruhlik, in an article that received a lot of attention, demonstrated that the distribution of oil wealth to the population, far from protecting the Saudi sovereigns against social pressures, had in fact strengthened opposition and dissidence instead, including during the oil boom phases. [101] According to Okruhlik, two reasons explain this paradox. Firstly, there is the inequality that prevails in the distribution of the oil wealth to the population, in which the allocation of the oil money does not happen automatically but according to the “overriding solidarities” (of religious denomination, provincial, tribal and family affiliations) of groups, and to political contexts (Islamic revolution in Iran, the Gulf War, rise in domestic opposition, etc.); and secondly, the fact that oil wealth provides potential dissidents, like the Sunni Islamist movement, private company owners, or Shiite opposition groups from Eastern Province, the resources necessary for a hostile mobilisation against the regime. [102]

28In order to explain the puzzling growth of opposition in the reputedly peaceful rentier states, Giacomo Luciani offers the following argument.

29

“In such a state, there is always an opposition, but the opposition will not be any more democratic than the ruler. Democratic methods will not appear as the most promising means to achieving the desired goal, simply because all groups will have a particularistic agenda, which is not conducive to the organization of consensus and majority support.” [103]

30Again, the reasoning here is based on a very debatable presupposition. The fiscal policy of Western democracies, cited here as a reference point, is rarely consensual and tends to favour, in order to satisfy the electoral interests and political agenda of the government majority, a particular social class to the detriment of another. [104] Algeria, a “rentier state” if the canonic consensus is to be trusted, moreover refutes this premise. Indeed, it was not the “pursuit of a particularistic agenda” [105] but rather the praetorian policy of the hardliners, concerned about the survival of the authoritarian and corrupt system, that prevented a political compromise being struck at the last minute between the National Liberation Front (FLN) of the reformers Mehri and Hamrouche, the Socialist Forces Front (FFS) of the democrat Aït Ahmed, and the Islamic Salvation Front (FIS) of the moderate Hachini, via a coup orchestrated on 11 January 1992, between the two rounds of the legislative elections. Refuting rentier determinism, after several weeks of debates held within the Sant’Egidio Catholic community, these parties managed to strike a political deal. The “national contract” signed on 13 February 1995 in Rome by the principal opposition actors – namely the FIS, FLN and FFS – demanded the “respect of the Universal Declaration of Human Rights”, the “rejection of violence to gain or hold on to power”, the “non-involvement of the army in political affairs”, and the “respect for the changeover of political power between parties through universal suffrage”. [106] It was the praetorian government that chose to obstruct this democratic pact, [107] despite its fiscal crisis and the support provided by several important Western capitals to the Algerian opposition’s agenda.

31The “rentier state” theory, by stating that the distribution of rentier wealth enables governments to “buy” political legitimacy, makes the sustainability of regimes dependent on the rent. Some representatives of this theory, although they suggest that the state’s fiscal crisis “encourages” and “stimulates” democratisation, note at the same time that: 1/ the durability of authoritarian regimes that show themselves incapable of dealing with their fiscal crisis is reduced accordingly, as technological progress, especially with regard to communications, and criteria for international acceptability, are making the task of the bloody dictator increasingly difficult; [108] 2/ “democracies may be in demand and yet fail or be derailed”. [109] Why, then, did the long and severe fiscal crisis that struck the Arab “rentier” states between 1985 and 2002 not lead either to democratisation or to the collapse of the authoritarian regimes? If the allocation of rent profits to the population is supposedly able to guarantee if not “consent”, then at least “depoliticisation”, how can one explain then the springing up, mid-oil boom, of popular revolts in these so-called rentier states, such as Bahrain, Egypt, Libya, Syria and Yemen? Did the Iranian Revolution not arise in 1979, in other words at the time when oil taxation accounted for 66% of the government’s revenues and where income per capita reached 1,999 US dollars? [110] The Shah’s Iran was not an exception either. The Iraqi and Libyan monarchies have also been hit by “revolutions” in spite of Western support and increased oil revenues. [111] Two lessons can be drawn from this: the sustainability and the collapse of authoritarian regimes are independent of oil rent; and repression – financed by rent – is not enough to guarantee the sustainability of an authoritarian regime.

32Why do “rentier state” populations persist with their protests when they are “subject to little or no taxation”? Why does the “distribution of rent” not manage therefore to “buy” “depoliticisation” and “loyalty”? Why does oil not manage to guarantee the sustainability of an authoritarian regime? The “rentier state” paradigm, driven by its economic determinism, finds itself in an analytical impasse: it is less the rent per se than the lost variable of the political institutions which can shed light on this puzzle. Several hypotheses can be put forward in this respect:

• Protest groups often believe, in Algeria as in the countries on the Arabian Peninsula, that the “oil belongs to everyone”. Consequently, it cannot be regarded as a “gift” which, generously given by the prince, would call for “recognition” in return. [112]
• The governed compare their incomes not with those of their parents but with those, incomparably greater, of their rulers: the resentment of the governed intensifies all the more given that the latter is sorely lacking in legitimacy.
• The system of corruption in these countries that presides over the distribution of the resources, if not the implementation of development programmes, generates underdevelopment and inequalities, which serve to fuel the dissatisfaction of the excluded and ignored. [113]

33The configuration of the Bahraini opposition, mostly nationalist and left-wing, changed after the Iranian revolution. Inspired by the Islamic movement, religious Shiite groups emerged, with the more radical among them gaining higher visibility. The riots in 1996 marked a bloody turning point: the repression resulted in several deaths and mass arrests. [123] Blinded by absolutism and undermined by internal disagreements, [124] the Al Khalifa dynasty ended up driving more and more social groups into opposition. It is not surprising therefore that the “rentier state” of Bahrain was the first Gulf State to experience a large-scale popular revolt, just a few weeks after the “Tunisian Spring” began. However, while the Sunni and Shiite opposition groups called for a parliamentary monarchy and the end to discrimination, the Al Khalifa state – which hosts the US Navy’s Fifth Fleet and central command – responded with repression; the 2,000 euro cheque offered by the sovereign mid-February 2011 to 115,000 households [125] having clearly not enabled the family to “buy” their “consent”. Saudi Arabia, intent after the fall of its ally Mubarak on pushing the “contagion” of the “Arab revolution” back outside the Gulf, sent “at the request” of the Al Khalifa family, a military intervention force to suppress manu militari the Bahraini popular revolt. [126]

34Algeria provides another example: those overlooked by the clientelist and corrupt distribution of resources, [127] who denounce the “contempt” (hogra) and the “predation” (al-nahb) of those in power, demand, through protest, their share of the rent (haqna). [128] Consequently, the exluded and those on the margins of society consider that, since it is a question of rights, they do not have to express “consent” to a government that, after all, suffers from an overwhelming lack of legitimacy. In order to repel the wave of Arab uprisings, the government, in the supplementary Finance law adopted in June 2011, hastened to boost public spending by 25%, even if this meant creating public deficits (which had reached 34% of GDP). The injection of an additional 23.5 billion dollars was to finance the wage increases granted by the government since January 2011. [129] The cost of support, far from being stable, is subject to inflation. The Algerian counter-example is very significant in this respect: the uninterrupted distribution of oil rent revenues over the past ten years has not been enough, in spite of its unprecedented volume in the country’s history (200 billion dollars between 2001 and 2009, and 300 billion dollars forecast for the period 2010-2014), to protect the state against the recurrence of riots, let alone to establish some form of “consent”. [130]

Is oil a “curse”?

35In her work The Paradox of Plenty. Oil Booms and Petro-States, Karl offers a political and institutional version of the “resource curse” theory which has become authoritative. [131] Her opus, which combines “economic institutionalism”, the “dependency theory”, and the “rentier state” and “rent-seeking” [132] theses in an eclectic approach, sets out to demonstrate why the oil booms of 1973-1974 and 1979-1980 caused “economic decline” and “regime destabilization” in the majority of “petro-states”. [133] The author considers – following Martin Shafer [134] – that a dominant export sector, in favouring the emergence of a rigid framework for decision-making, conservatism and inertia, alters the state’s capacities to move away from the old development model in order to promote a new one. [135] Karl aims to demonstrate that dependence on oil shapes not only social classes and the type of regime, but also the institutions of the state and the framework for decision-making by rulers. [136] The book includes a lengthy and admirable study of Venezuela and brief analyses of the development trajectories followed by Algeria, Iran, Indonesia and Nigeria. The author asserts that these five large petro-states, between 1974 and 1992, experienced “a strikingly similar structuration of choice” and “disappointing political and economic outcomes”. [137]

36Nothing, however, could be less sure. Firstly, because Karl does not manage to establish convincingly that the “oil boom” generates, by itself, “political instability”. Moreover, there is no evidence that the five states studied are less stable than other developing countries. In fact, Venezuela, which she discusses at length, has for a long time been, as we have seen, one of the most stable states in South America. Contrary to Karl’s assertions, the ruling elites of the five big petro-states studied have not precisely followed the same “structuration of choice”: Indonesia, which unlike Algeria, opted for a continued devaluation of its national currency, the protection of its non-hydrocarbon exports, the promotion of agriculture, budgetary austerity and the diversification of its fiscal structure, has managed, as recognised by Karl, to avoid “economic deterioration” and “political instability”. [138] Lastly, the hypothetical harmful effects of “rentierism” (authoritarianism, rent-seeking, corruption, economic stagnation) are, on closer inspection, characteristic of “rentier states” as well as non-rentier states. Karl views the discovery of external rent as the starting point in the building of the state in developing countries, this simultaneity determining the state capacities. [139] She thus seizes on the discovery of oil in the 1920s as an inaugural moment in the formation of the Venezuelan state. Despite the length of the oil era, Venezuela was, according to Karl, unable to escape the “petro-state trap”. She explains the failure of Venezuela to acquire a “Norwegian” style bureaucracy by the temporal variable: the “petro-state” disease can be avoided when and only when state building has occurred prior to the introduction of the oil exporting activity. [140] The post-colonial period (1810-1920) certainly did not allow the centralisation of the Venezuelan state. This failure is not, however, inherent to the oil sector but much more to the Caudillo wars and territorial fragmentiation (Federal War of 1859-1863), wars that marked Venezuela throughout the nineteenth century. The institutional weakness of the state is not more specific to Venezuela but constitutes, during this century, a characteristic common to the whole of Latin America. [141] In arbitrarily postulating the “coincidence” of state-building with the development of the oil sector, Karl loses sight of historical depth: the building of the state and the regime precedes the oil boom. Thus, while she mentions in passing that oil taxation in Algeria remained relatively low before 1973, [142] she does not allow herself to explore the nature of the government prior to the boom in the hydrocarbons sector. [143] These are however the rules of the game – characterised in Norway by the robustness of the democratic institutions, whereas in Algeria these institutions are threatened by, nolens volens, the prevalence of the monopolistic and unaccountable use of power, clientelism, political corruption, dirty tricks politics and the institutional weakness of the state – which dictate how oil wealth is governed. [144] The avoidance of the historical (and cultural) depths of politics [145] consequently leads the author to minimise the importance of, to take the example of Nigeria, a variable as key as tribal, ethnic and religious conflicts. The Nigerian-Biafran civil war which left close to one million dead between 1967 and 1970 and increased the power of the praetorians, pre-empted the management of the oil wealth, where ethnicity had weakened the state institutions. Karl, who supports the idea that the collapse of the regime is the near-inevitable outcome of authoritarian “petro-states”, admits that the political stability of Suharto’s Indonesia represents a “notable exception” to this law. [146] However, the same applies to the Algerian “petro-state”, where the resilience of the praetorian regime contradicts the “prediction” of its model. In truth, when we consider the longevity of the Iraqi, Egyptian and Syrian regimes that Karl ranks amongst the “petro-states” without actually including them in her comparative analysis, it can be concluded that the Iranian and Nigerian cases are more the exception than the rule. [147]

37The case of Indonesia lastly invalidates the “resource curse” theory in two other important respects: economic performance and democratisation. Indeed, the economic performances registered by Indonesia are better than those of the other “petro-states”: not only did its GDp per capita not decline during the two decades subsequent to the 1973 oil boom, it moreover recorded 5% growth between 1966 and 1997. [148] The hydrocarbons that constituted 80% of exports in 1981 represented no more than 26% in 1989; manufacturing which accounted for less than 10% of the country’s domestic product in 1966, constituted a quarter of Indonesian GDP in 1996. [149] Refuting the “Dutch Disease” argument, the Indonesian economy was highly dependent on hydrocarbons when Suharto came to power in 1965-1966 but, by the time of the dictator’s fall in 1998, it had become markedly more diversified. [150] Indonesia, like the countries of South America, irrevocably refutes the “resource curse” theory: the largest Muslim country in the world, which has organised three free and fair legislative and presidential elections since the fall of the authoritarian regime in 1988, has successfully conducted its democratic transition. [151]

38*

39* *

40The popular uprisings that have taken place since January 2011 in oil countries such as Bahrain, Libya, Egypt, Syria and Yemen reveal the analytical impasse in which the “rentier state” and “resource curse” theories find themselves. In this article, I have attempted to review these theories. To conclude, I will summarise below the main criticisms made.

411/ The theorists of the “rentier state” and the “resource curse”, in adopting the “Whig” view of the European history of political representation, make taxation a necessary condition for democracy. This paradigm, by confusing the historical phases of the emergence, consolidation and longevity of the representative institutions, fails in that creates a retrospective illusion. Taxation has indeed contributed to the emergence and the longevity of representative institutions, principally where these had a direct role in the collection of taxes. The representative institutions set up between 1789 and 1848 did not enjoy this bargaining and negotiating power that pre-modern assemblies were able to derive from taxation, the executive being, in the modern state, the sole institution that fixes and levies tax through the fiscal bureaucracy. And so, ultimately, taxation only played a “minor” role in the emergence, consolidation and longevity of modern representative institutions.

422/ The theoretical mainstream of the “rentier state” and the “resource curse”, in using the wealth of natural resources as the reason for authoritarianism, does not take account of the fact that public institutions existed prior to the arrival of rent.

433/ The “rentier state” and “resource curse” theses, in yielding to economic determinism, end up glossing over politics. The key variable is not so much the presence of natural resource wealth but the manner in which it is used by those in power. Consequently, the effect of rent on the regime depends on the political institutions established – and these usually pre-date the advent of the rentier era.

444/ The theory, centred on the state, is indifferent towards social movements. Founded on the assumption that rent “buys popular consent”, it struggles to grasp the social motives of protest in a context of abundant wealth.

455/ The “political autonomy” that the “rentier state” finds in “fiscal autonomy” is contingent and not at all structural – the Iranian “Islamic revolution” and the “Arab uprisings” are clear proof of this.

466/ The “allocation” of rent profits to populations does not automatically result in their “depoliticisation”: the structured protest dynamics – fully informed about the “fiscal autonomy” of “petro-states” – demolish this key argument of the “rentier state” theory.

477/ The theoretical mainstream of the “rentier state” and the “resource curse”, which sees taxation as the sole motivation for the governed to exert democratic pressure on their rulers, struggles to grasp popular demands for accountability in situations of low taxation.

488/ Repression – the argument put forward by the “rentier state” paradigm as an answer to the failings of “depoliticisation” – is not sufficient to guarantee the survival of authoritarian regimes, as proven by the fall of autocrats Mohamed Réza Pahlavi and Zine el Abidine Ben Ali.

499/ The “resource curse” theory, based on the negative correlation between natural wealth and development, establishes the “law” of the “curse”. But a correlation, no matter how “strong” it may be, is not a causality. Moreover, a correlation can be interpreted in one sense just as much as in another. Lastly, a correlation does not indicate how the quantifiable variables interact.

5010/ While the “petro-state” thesis sentences rentier countries to the “curse” of authoritarianism, the establishment and consolidation of democracies from Norway to Peru, from Bolivia to Indonesia and from Venezuela to Botswana, invalidate the “law” that claims that natural resources and democratic regimes are incompatible.

51This critical review, whilst dismissive of the “rentier state” and “resource curse” theories, calls in particular for institutional configurations and changes to be taken more seriously in the analysis of the relationships between rent and political regimes. [152]