INTRODUCTION
Could the roots of the ‘resource curse’ be much broader than usually imagined? When social scientists think about the ‘resource curse’, images likely to come to mind are the despotic, repressive and corrupt states of the Middle East or Africa's Gulf of Guinea, states existing off the rents of ‘black gold’ with neither popular accountability nor responsibility. Or we might imagine the conflicts and wars of Africa, Asia and Central America, arising over ‘lootable resources’ like tropical timber, alluvial diamonds or illicit drugs (Snyder Reference Snyder2006). An extensive literature has shown us that countries possessing abundant extractable natural resources become ‘cursed’ by these resources, falling prone to a number of deleterious economic, social and political phenomena: they tend to be beset by high income inequalities and inflated living costs (Reed Reference Reed2006; Wantchekon Reference Wantchekon2000), weak economies (Collier & Hoeffler Reference Collier and Hoeffler2005; Ross Reference Ross1999; Sachs & Warner Reference Sachs and Warner1995), authoritarian rule and poor governance (Jensen & Wantchekon Reference Jensen and Wantchekon2004), and civil conflict (Collier & Hoeffler Reference Collier and Hoeffler1998, Reference Collier and Hoeffler2005; Ross Reference Ross1999; Snyder Reference Snyder2006). Is the ‘curse’ specific to these particular resources, or can other types of resources have the same effects?Footnote 1 Moreover, does the presence of riches inherently lead to economic and political woes, or do policy decisions make them more likely?
This essay examines what light, if any, the little-studied Republic of Djibouti might shed on these specific questions. With a highly centralised economy dominated by inflows from US, French and other European military bases, concomitant aid monies, and tariffs on trade passing through the deep-sea harbour to and from Ethiopia, does Djibouti's non-tangible resource – geo-strategic location – produce the same effects as extractive ones? If so, can Djibouti teach us anything else new?
DJIBOUTI: A TINY COUNTRY LIVING OFF BONANZA RENTS
The Republic of Djibouti is an extremely small, young country. It has a landmass of 23,200 km2 (approximately the size of Belgium or New Hampshire) and between 430,000 and 840,000 people, a large proportion of whom are immigrants fleeing the conflicts of neighbouring countries. Until very recently, most Djiboutians were subsistence pastoralists maintaining a nomadic lifestyle intertwining them with neighbouring countries. The Afar of northern Djibouti followed transhumance migration into largely Afar areas of Eritrea and Ethiopia, while the Issa Somalis grazed into present-day Somalia. Djiboutians have traded with these neighbours as well as those of the Arabian Peninsula for centuries. This proximity made Djiboutians among the first African Muslims.
The country's boundaries stem from its history as a colony and later an overseas territory of France. Beginning in the 1800s, Djibouti was known as the French Somali Coast, a name that changed in 1967 to French Territory of the Afar and Issa. The change was meant to recognise both major ethnic groups living there (not just the Somali Issa), and to distance the territory from the then-newly established Somali Republic. Independence came to Djibouti in 1977 via referendum, held after Ethiopia renounced its claim to Djibouti, and while Somalia was distracted by conflict in the Ogaden region of Ethiopia (Shehim & Searing Reference Shehim and Searing1980). For decades, Djibouti's role has alternated between buffer and pawn of these two larger neighbours.Footnote 2
Geographically, Djibouti has one of the most inhospitable, barren environments on the planet – it has virtually no arable land, no permanent fresh water source, no significant mineral resources, very little vegetation, an average daily temperature of 34°C,Footnote 3 and severe, persistent drought for the past six years. Despite these apparent difficulties, tiny Djibouti is able to generate relatively substantial incomes from the only resource it has: geo-strategic location.
The country benefits not only from its position at the gateway to the Red Sea on the shipping routes between the Indian Ocean and the Mediterranean, and its military–strategic location next to the Arabian Peninsula, but also from being the most expedient trade route for Africa's second most populous country, land-locked Ethiopia. The Government of the Republic of Djibouti (GoRD) relies on its location for the vast majority of its revenue. Money comes primarily from three sources: first, the International Port of Djibouti, 86% of whose business involves Ethiopian trade (EIU 2006). Proceeds from the port, combined with the second rental source, the Djibouti–Ethiopian Railway, account for one third of the country's GDP (FH 2007). Finally, Djibouti earns rents from several foreign military bases, especially large French and American ones; both countries pay US $30–$36m yearly, alongside considerable bilateral assistance aimed at fighting endemic poverty.
Currently 75% of Djiboutians live in poverty, with 42% chronically unable to meet their caloric intake needs. These rates have increased from 65% and 34% respectively in 1996 (RoD 2004). In rural areas, 96·7% are impoverished. Social indicators reinforce these figures, with life expectancy only forty-six years (USAID 2005). With the goal of reversing these trends, the country receives between $100m and $300m in aid annually (IRI 2005; USAID 2005), primarily from the US, Saudi Arabia and France. Depending on the population size, which is hotly contested, this averages between $130 and $700 per capita from aid alone.Footnote 4
Some of this income is not new: for centuries Djibouti has traded foodstuffs, live animals, khat and salt both formally and informally with its neighbours on the Horn and the Arabian Peninsula. Moreover, France, Djibouti's former ruler, has funded Djiboutian administration and public services since before independence. In the early 1970s, France spent $60m a year on Djibouti (Marks Reference Marks1974). As Bollée (Reference Bollée2003) points out, France's absolute assistance level is not particularly high, but because the population is so small, Djibouti is the biggest per capita recipient of French foreign spending. France now spends $160m annually on its military base, plus an additional $35m on local military and aid.
While aid from France shaped much of Djibouti's current political economy, the country has seen additional fortune arise out of others' troubles in the past decade. First, conflict between Ethiopia and Eritrea forced Ethiopia to use Djibouti as its primary trade route to the sea. Trade with Ethiopia nearly doubled from 1997 to 1998, rising from 1·7mt to 3·1mt (Bollée Reference Bollée2003: 482). Second, Djibouti's revenues increased dramatically following the US terrorist attacks on 11 September 2001, when Djibouti allowed the US to open a military base near its capital for a reported $30m annually. Establishing this base has been a major accomplishment for the US, which had been looking to open a base in Africa ‘for over thirty years’, according to a US military informant. Located mere kilometres from Yemen, the site of the 2000 bombing of the USS Cole, US Camp Lemonier is considered a great success in the ‘war on terror’.
In exchange for a five-year lease, the US promised Djibouti considerable financial and development assistance. Djibouti now receives more aid from the US than any other country in Africa, up from only $7m per year before the base was opened (EIU 2005), and it is seen by the US as one of two African countries of ‘high strategic importance’ (USAID 2006).Footnote 5 The arrangement has been favourable to the tiny state, which not only renewed its five-year lease with the US in 2006, but expanded it dramatically. The base grew from about 90 acres to nearly 500, at a rental rate that remains undisclosed (Giordono Reference Giordono2006).
Thus, ‘the “war on terror” has proved a bonanza for the mini-state on the Red Sea’ (Bollée Reference Bollée2003: 484). The Djiboutian government used the high-priced American lease to leverage more money from the French, who increased their annual payment from $20m to $34m (ibid.). Other donors and NGOs reported to the author that heavy American support has made working with the GoRD more difficult: they sense that the GoRD knows it has leeway to make demands or govern imprudently, counting on the US for backup or to turn a blind eye (Brass Reference Brass2008). This impact is not intended by the US, which, for its part, reports working to improve donor coordination and reduce duplication of efforts – but in general coordination is weak.
THE RESOURCE CURSE IN DJIBOUTI: EXTENDING THE THEORY
The resource curse concept is straightforward: countries that possess abundant extractable natural resources become ‘cursed’ by these resources and suffer a number of negative economic, social and political consequences, especially if the resources are discovered while the country's economy is undeveloped.
The mechanisms behind this ‘curse’ are more complicated, and a variety of explanations have been offered by both political scientists and economists as to why and how externalities occur. Ross (Reference Ross1999: 301–7) reviews the accounts offered by economists, citing four main explanations. First, resource exporters often suffer from declining terms of trade, which widen the gap between themselves and rich industrialised states. Second, international commodities are often subject to sharp price fluctuations, which transfer to domestic economies and make investment risky and government revenues unreliable. Third, resources are often extracted by foreign investors, who are sometimes able to repatriate their profits such that the rents never reach the country from which they are derived. Finally, the most common and cogent theory: ‘Dutch disease’, in which large natural resource exports create major inflows of foreign exchange, which cause the local currency to appreciate. At the same time, the booming resource sector draws capital, and sometimes labour, away from existing industries and export-based agriculture – especially when the economy is otherwise underdeveloped. These potentially profitable sectors are further discouraged by the appreciated currency, which makes imports cheap and locally produced goods too expensive for sale on the world market.
As several scholars point out, however, these explanations suggest that negative results are inevitable by-products of certain resources, and that governments have no agency (Ross Reference Ross1999; Watts Reference Watts2004). This is, of course, not true: ‘oil [or other natural resources] is not a conjurer of kleptocrats or warlords’ (Reed Reference Reed2006: 22). While windfall inflows may make it difficult for policy makers to focus on sound decision-making, they still have considerable choice: leaders could make policy decisions that mitigate rather than enhance the effects described above. As Ross (Reference Ross1999: 307) puts it, ‘the failure of states to take measures that could change resource abundance from a liability to an asset has become the most puzzling part of the resource curse’.
Political explanations that describe why or under what conditions leaders make these decisions help unravel this puzzle. Several scholars propose mechanisms by which bureaucratic inefficiencies and outright looting occur. Some say that where resource revenues are extremely high relative to the rest of the economy, policy makers fall into patterns of irrational short-sightedness, even ‘wealth induced stupors’ (Ross Reference Ross1999: 310) that explain poor decisions. Others favour rationalist explanations: Collier & Hoeffler (Reference Collier and Hoeffler2005), for example, hypothesise that leaders who know they have access to resources for a limited period of time – in the case of instable regimes or electoral term limits – loot and distribute more commonly than they follow sound fiscal management and investment practices. Moreover, without a need to tax citizens, many states never have a reason to develop sound bureaucratic institutions, or those they have can atrophy (Chaudhry Reference Chaudhry1997). Alternatively, states can develop capable institutions only in areas directly relevant to the leading export (Shafer 1994, cited in Ross Reference Ross1999), or governments can become overextended from trying to do too much at once with windfall rents (Karl Reference Karl1997).
While some of these explanations are valid for Djibouti, I argue that the country's economic policy decisions derive directly from leaders' power imperative.Footnote 6 It is the desire to maintain control of the state – and specifically the resources attached to it – that drives Djiboutian decision-making. Regardless of their intentions, which are nearly impossible to gauge, President Guelleh and his predecessor's actions have been to invest in sectors that bring revenue directly to the state, while neglecting investment in education or economic diversification.Footnote 7 Skill-levels are so low that even manual labourers are imported. An undiversified economy and uneducated populace mean two things for the Djboutian government: without education, it is unlikely that an organised opposition will arise to pose a significant threat; and with control over nearly the entire economy, the government limits access to resources that could bankroll a political challenge.
However, as Djiboutian leaders know well, the undiversified economy lowers the opportunity costs for oppositions to instigate conflict (Collier & Hoeffler Reference Collier and Hoeffler2005). People in Djibouti have very few exit options, and their voices can, for the most part, be ignored (Hirschman Reference Hirschman1970). As in other resource-rich states, violent conflict becomes one of the only ways to be heard. Thus maintaining compliance (if not exactly loyalty) is a government priority. For this reason, Djibouti keeps tight control over civil society, the media and opposition. And it keeps the educated populace extremely small.
POLITICAL AND ADMINISTRATIVE REALITIES: IMPERATIVES TO MAINTAIN CONTROL
Djibouti's political economy reminds us that undiversified, resource-rich economies can make a leader's desire to maintain power extremely strong, since revenue-generating alternatives are limited. This is particularly true when resource extraction occurs through the state, and considerable investment is required in the first instance, as with Djibouti's bilateral agreements and port infrastructure.
Moreover, monopoly rental income allows the government to thrive regardless of public opinion. Those in control of the state apparatus are not dependent on citizens for revenue, which comes to most states via taxation (Tilly Reference Tilly1990). This means that the normal relationship between governed and governor, wherein citizens hold both rights and responsibilities to the state and vice versa, does not develop. In other words, there is very little state accountability in Djibouti.
As in many resource-rich states, the result is a government inclined towards authoritarian, highly consolidated power, repressive tactics and an administrative apparatus that is a far cry from Weber's idealised notion of bureaucracy. Despite elections, the country is not democratic, and it has exceedingly low levels of human rights and civil liberties. Finally, the tiny state has seen persistent armed conflict over control of state spoils. While many of these feature of Djiboutian governance existed before the recent flood of rents, they remain pronounced or have expanded at a time when most African states have been persuaded to liberalise politically or lose economic assistance.
Centralisation of power
Power is extremely centralised in Djibouti, around the office and the person of the president, who makes many of Djibouti's laws and policies through unilateral Presidential Decree. This political control extends to the economy, in which the president is said to have a close hand in all major rent-earners. The other branches of government have extremely little say: the legislature is in session only two months of the year, during which time it acts primarily as a rubber stamp for the president (IRI 2005). For example, in 2003, the legislature's presiding officer questioned the necessity of discussing the national budget, and the debate was cut short (CDD 2007). Nor does the legislature have power to approve or refuse presidential appointments, and it is completely controlled by the ruling coalition of parties, leaving accountability low.
Likewise, the judicial system is not independent of the president (FH 2006; US SD 2007). Rule of law is extremely weak, and the constitution is frequently disregarded: arrests are made without warrants, the presumption of innocence is neglected at trial, detainees and prisoners are beaten by security forces, freedom of speech and assembly are repressed, and intimidation tactics are common (FH 2006).
According to most informants, real decisions are made informally by a small coterie around the president. Although Djibouti's current president, Ismaïl Omar Guelleh, came to office in 1999, he served for twenty years as the Chief of Staff and head of state security forces for the first president, and his cabinet is nearly identical to that of his predecessor. Most of the members of his inner circle are also members of his ethnic clan, a sub-group of the majority Issa population. This group is disproportionately favoured in government, and has controlled the military, civil service, ruling party and security apparatus since independence (FH 2006). Resource rents allow the GoRD to maintain support to the centre through patronage, distributing resources to key constituencies, rather than developing the country as a whole. While this patrimonial basis of compliance is common to most African states (Bratton & van de Walle Reference Bratton and van de Walle1997; Jackson & Rosberg Reference Jackson and Rosberg1984), the degree of centralisation in Djibouti is particularly high; the GoRD derives more of its revenue from rents than the typical African state.Footnote 8
Nevertheless, since the 2001 peace accords that ended Djibouti's civil war (discussed below), and as a major component of Djibouti's Poverty Reduction Strategy Paper, the GoRD has promised to enact decentralisation as a means of distributing power to local areas, including predominantly opposition Afar territories. In principle, administrative powers have devolved to elected Regional Councils. In reality, however, no devolution has occurred. Despite holding the inaugural Regional Council elections in April 2006, numerous Regional Council members, civil servants and politicians relate that the councils ‘do nothing’, as there has been no transfer of funds from the central government. Instead, local power remains in the hands of a loyal presidential appointee, the District Commissioner. Promises that Regional Councils can develop direct relationships with donors also remain a mirage: a very senior Ministry of Foreign Affairs official reported that all Regional Councils can talk with donors, but funds will continue to be collected and distributed from the central ministry.
Democracy in name only
The resource curse posits a negative correlation between resource rents and democracy (Jensen & Wantchekon Reference Jensen and Wantchekon2004). In many poorly institutionalised states where elections do occur, however, votes are often cast based on the allocation of goods during the campaign and promises for supplies to continue after the election. Thus savvy incumbents often distribute resource rents to co-opt opposition (Ayittey Reference Ayittey1998). Rather than buying off the opposition, Djibouti's government has taken a different strategy: near complete exclusion and repression, under a thin veneer of democracy.
Djibouti holds presidential elections every six years and has officially allowed opposition parties since 1992. Yet the RPP (Rassemblement Populaire pour le Progrès) and its allies have been in control of the government since independence, winning not only every presidential election but also every seat in the 65-member National Assembly in every legislative election to date. The government blames this on lack of coordination among opposition parties. A closer examination of the situation, however, reveals a number of electoral laws, policies and practices that effectively preclude the possibility of an opposition win.
For example, parliamentary elections work on a winner-take-all system, meaning that the party winning a simple majority of votes in a constituency receives all the parliamentary seats. In Djibouti City constituency, an Issa and RPP stronghold, the winning party receives thirty-seven seats, giving that party an automatic majority in the National Assembly. Therefore, even when the opposition earned 38% of the national vote in 2003, it gained no voice in parliament. Moreover, only parties with existing representation in the National Assembly are eligible for state funds during elections – meaning only the RPP has ever received funds. To contest a presidential election, a candidate must raise 5m Djiboutian francs (approximately US $28,500) – an inordinate fee in such a poor country (CDD 2007). In the 2005 elections, the only challenger raised this sum, but still withdrew, arguing that he couldn't campaign with such tight governmental control of the media and harassment of the opposition (FH 2006). Similarly, the opposition boycotted the 2008 parliamentary elections, calling for proportional representation. According to opposition leader Ismael Guedi Hared, the government refuses to make electoral changes, ‘because it feels threatened’ (Goujon Reference Goujon2008). Indeed, in these elections voter turnout was as low as 0·3% in some polling stations (ibid.). Opposition parties have protested against these and other unfair practicesFootnote 9 – with little government response.
Politics of fear: repression of opposition, civil society and media
Harassment of opposition electoral candidates is symbolic of the GoRD's general attitude toward dissent; it maintains a very short leash on criticism, free speech and free assembly. At present, there is a single government-operated internet service provider, phone company and mobile operator, and each is said to be extensively monitored. Multiple informants reported that they are careful not to critique the government in emails or over the phone, and one US Government informant reported removing a wire-tapping device from her home telephone. This sort of tapping is done without a warrant (CDD 2007). The government has also been known to block critical websites, such as that of the Association for Respect of Human Rights in Djibouti, and to cut telephone or electricity to opponents (US SD 2007).
Moreover, with 100% RPP control of the Assemblée Nationale, there is no national venue for debate. The media cannot serve as a forum either, because it is state-controlled and tightly monitored; reporters are based in the Ministry of Information. In theory, each political party is allowed a print journal, but constant harassment, surveillance, seizure of equipment, removal of newspapers from newsstands and repeated unwarranted arrests of opposition journalists keep these papers quiet (FH 2006; RWF 2004, 2007a; US SD 2007). For the same reasons, journalists in general maintain strict self-censorship.
Like the media, civil society organisations (CSOs) maintain a neutral or positive view of the state. Since education and training opportunities are rare, these organisations tend to be exceedingly weak. Those with greater capacity are often led by community leaders who are also civil servants and therefore dependent on GoRD approval (IRI 2005). CSOs are not often consulted in the development of public policy (RoD 2004: 30). Organisations that have spoken out publicly have been subject to attacks, threats and surveillance. Technically, labour unions are legal, but the GoRD thwarts their operations by firing their leaders, preventing their meetings and even replacing leaders with pro-government sycophants (FH 2006; US SD 2007).
As with so many other resource-rich states, Djibouti's human rights record is appalling (US SD 2007). The US State Department (2007) and Freedom House (2006) cite the following violations: discrimination on the basis of ethnicity within the government; abuse of prisoners and detainees, including prolonged pre-trial detention; overcrowded, unsanitary, inadequately staffed prisons; unwarranted and arbitrary arrests; denial of constitutionally guaranteed free speech and assembly; use of force to disperse legal strikes or rallies; violence and discrimination against women; and official impunity.
The administration is not new to these intimidation tactics, since the current president ran the Djiboutian security service under the first president. International observers who have brought these issues to attention have been dealt with harshly. UN representatives sent to Djibouti to investigate labour rights in 2006 were expelled (CDD 2007). And President Guelleh accused French officials of wanting to destabilise the country when they investigated the involvement of two of Guelleh's close aids in the mysterious 1995 death of French judge Bernard Borrel, who was in Djibouti to probe Guelleh's possible role in arms trafficking and terrorism.Footnote 10
Bureaucratic nightmares: inefficiency defined
The semi-authoritarian, patrimonial and informal nature of politics in Djibouti combined with extremely low levels of human capacity means that the GoRD is beset with issues of poor governance and bureaucratic inefficiencies, as resource curse theory would predict. In Djibouti, ‘dysfunctional’ financial and budgetary management, extremely ineffective public spending patterns, low transparency and accountability, and high corruption (RoD 2004: 5) combine with a politics of control, such that little is or can be achieved by the bureaucracy.
Within the GoRD, this means that public enterprises and ministries often have no stated objectives, no long term planning, weak implementation abilities, and no information management systems (ibid.). One result is that reliable data and statistics are nearly non-existent, and published estimates of population, health, demographics, poverty, employment and budgeting are suspect (USAID 2006: v). An unpublished IMF report stated that ‘the lack of reliable day to day accounting in the operations of the Djiboutian state throws doubts on the reliability of the globally computed figures’ (Bollée Reference Bollée2003: 483).
According to three senior informants in different GoRD departments, administration always runs in ‘emergency’ or ‘crisis management’ mode, as a few highly competent civil servants attempt to prod their insufficiently capable and very slow-moving departments into action. Because of this, pre-planned monitoring and evaluation is uncommon, and programme management is rarely based on results (RoD 2004): departments simply move from one disaster to another. Even in this crisis mode, UN agencies report that the GoRD lacks the capacity to respond to the current drought. Indeed, the UN has felt it needed to ‘lead from behind’ in order to aid drought victims (UN 2006).
In the agricultural sector, we find detailed examples of bureaucratic negligence: the national livestock law has not been updated since before independence, the most recent animal census dates from 1978, and the Ministry of Agriculture, Livestock and the Sea's (MAEM) three-year plan, while well-intentioned, is more of a wish-list than an actionable plan. Several high-level Ministry officials admitted that its implementation depends nearly 100% on external financing. Even then, it is difficult for MAEM to get the GoRD to pay the share required by most development partners.
Similar red tape makes business investment and development difficult. Obtaining a business license is a multi-step problem involving two ministries and the Chamber of Commerce. For foreign investors, it is more difficult still, as the Finance Ministry will issue a licence only if an investor has an investor visa, but he can't obtain this visa without already having the licence (HF 2008). Generally speaking, extensive government regulations, political interference and corruption hinder transparency and opening, operating and closing a business.
Armed civil conflict for access to the spoils
Finally, Djibouti, like so many other resource-rich countries, has been subject to violent civil conflict over resource spoils (Collier & Hoeffler Reference Collier and Hoeffler2005; Sachs & Warner Reference Sachs and Warner1995). Since before independence, politics in Djibouti has been highly charged along ethnic lines, polarising the country's two ethnic groups, the Issa and the Afar. The Issa, often referred to simply as ‘Somali’ since they are a Somali clan, live in the southern third of the country and comprise most of the population – increasingly so, due to high emigration levels from Somalia and Ethiopia. The Afar, on the other hand, are of Ethiopian origin, live in the northern two thirds of Djibouti, and have been present in the territory for longer.
Considerable tensions lie below the surface and erupted into civil conflict during much of the 1990s. These tensions largely developed during colonialism, as the French significantly favoured the Afar. They were generally better educated and held more positions of power than the Issa, and were in favour of remaining a territory of France. After attempts to secede and create a ‘Greater Somalia’ uniting Somalis in Ethiopia, Somalia and present-day Djibouti were thwarted, Djiboutian Issas pushed strongly for independence (Schraeder Reference Schraeder1993). With a then-small majority of the population, they voted in a referendum for complete self-rule. Following independence, power shifted almost completely to the Issa majority, which has grown considerably due to immigration from Somalia.
After fourteen years of political and economic exclusion, an Afar rebellion, led by the Front for the Restoration of Unity and Democracy (FRUD), began in November 1991. Within a year of intense fighting, former President Hassan Gouled Aptidon recognised that at least nominal appeasement was necessary. His government produced a new constitution in 1992, promising decentralisation, press freedom, human rights and multi-party elections. Most conflict ended by 1994, though smaller FRUD factions continued to fight until 2001, when the final peace accord was signed. By 1999, FRUD supported the Guelleh RPP presidential bid, since Guelleh promised to enact many of the as-yet-unimplemented portions of the 1992 constitution. As many of these promises have still not been implemented years later, however, tensions have recently been on the rise. As recently as February 2006, there were reports of armed clashes between FRUD rebels and the GoRD in the Mabla mountains north of the capital (EIU 2006).
The economic impact of the conflict is also considerable. Djibouti's economy was severely weakened by the diversion of a significant portion of the GoRD budget from social services and economic development to military needs. The army grew to over 100,000 men, and 25% of the government budget went to military expenses (Bollée Reference Bollée2003) – another symptom of the resource curse. Ironically, with the economy weakened, access to the state and the rents it controls only becomes more important.
RESOURCES+POLITICS OF CONTROL=A CURSED ECONOMY
While many theorists imply that resources somehow magically lead to negative externalities, this paper asserts that it is specifically the way in which leaders respond to the existence of rent-producing resources that matters most. Djibouti's leaders, fearful of losing control, have helped create economic realities of ‘the curse’. This section illuminates how government policy decisions result in inequitable distribution, lack of diversification of the economy or a skilled workforce, and bureaucratic inefficiencies.
A dual economy: go-karts and paint ball on an average of $1.23 per day?
Economic growth in Djibouti has virtually stagnated: the 1–3% average rate of growth in the past decade barely compares to rapid population growth. Economic development has been hindered by a pervasive lack of economic freedom, including: ‘weak business freedom, trade freedom, freedom from government, property rights, and freedom from corruption’ (HF 2008). At the same time, however, total government expenditure in Djibouti is high, equal to 36·5% of GDP (ibid.). This figure, combined with soaring poverty rates, suggests that the GoRD's revenue streams do not reach the people. Instead, as in most rent-based economies, government revenue largely benefits the elite at the expense of the poor. As in most resource cursed states, the GoRD's economic policies have created a profound dual economy.
Djiboutian GDP per capita is $890 (RoD 2004), making Djibouti a middle-income country. Yet this figure is deceptive. When purchasing power and the extremely high cost of living are taken into consideration, real GDP per capita is only $450 ($1.23 per day), and Djibouti is one of the least developed countries on earth, ranking 149 out of 177 on the 2007 UN Human Development Index (UNDP 2007).
Living costs are elevated in Djibouti for several reasons. The Djiboutian franc has been pegged to the US dollar at artificially high rates since independence. This fixed exchange rate seriously damages the ability of poor producers to export, as products are overpriced by regional and global standard. The country exports (or really, re-exports) $90m worth of cargo, yet it imports a whopping $373m of goods, according to 2004 estimates (HF 2008). It is doubtful whether the GoRD will devalue its currency, as the short-term consequences on Djiboutian purchasing power would be debilitating – nearly everything available for purchase in Djibouti is imported, and devaluation would cause prices to rise even higher. Moreover, the level of foreign exchange in the rent-based economy would probably keep the rate very high.
Prices are also high in Djibouti because elites and foreigners from the large expatriate community are willing and able to pay vastly higher prices for goods and services than are locals, creating the ‘false’ or ‘dual’ economy so common in resource-rich states. Evidence for the dual economy abounds: it can be seen in housing rental costs, where even Spartan apartments average around $1,000 per month – about double the average yearly income of a Djiboutian. It can also be seen in the luxury goods readily available for the few who can afford them: giant gas-guzzling, two-tank Toyota Land Cruisers; French mega-chain supermarkets full of imported cheeses, meats, chocolates, wines and fruit; a bowling alley, go-kart track and paint-ball facilities; a dog-and-cat veterinarian for the pets of the rich (one of only three veterinarians in a country with over a million livestock); surprisingly tasty and well-decorated Italian, French, Seafood and Chinese restaurants with industrial-sized caged air conditioners and electricity generators designed to make their wealthy clientele chilly even when it is 43°C outside and the public power supply has been cut.
Limited investment outside the leading sector
As Dunning (Reference Dunning2005) demonstrates for Mobutu's Zaire, leaders in Djibouti seem to feel that investments in public goods like infrastructure and education pose a threat to their political power, and therefore their economic largesse.
Infrastructure
Save for the well-connected elite, Djiboutians lack access to the most basic infrastructure. About one third of the population has no reliable access to water (Doualeh Reference Doualeh2002); rural women often walk more than 10 km to collect it from almost-dry wells up to 20 metres deep (RoD 2004). Piped water is brackish where available, with salt content far above WHO-recommended levels and borehole-exit temperatures over 40°C (ibid.). Not surprisingly, those who can afford it consume only filtered water, which tankers can regularly be seen distributing, often at a rate four to six times higher than official tariffs (ibid.).
Similarly, only 10% of the country's roads are tarred, and these are either in the capital or link it to the country's few towns (EIU 2006). Even the long road to Djibouti's new export-oriented animal quarantine centre was unpaved – and tyre-burstingly bumpy – at the time of research. Rural dwellers, who comprise 20% of the population, are mostly semi-nomadic pastoralists. Many would like to settle and trade, but they lack even the basics for livestock trade, such as animal handling facilities, transit points, markets, sanitary controls, access to veterinary care, refrigeration facilities and land-use laws. Partially because of this, animals leaving Djibouti's quarantining facility are nearly all of Ethiopian or Somali origin.
Electricity is similarly unavailable, with only 15·9% of the extremely poor and 43% of the urban poor having access. All fossil fuel for electricity is imported and combusted in several loud and inefficient diesel turbines, reported to be ship engines. At a rate five times higher than in neighbouring Ethiopia (RoD 2004), electricity in Djibouti is more expensive than nearly anywhere else in the world. One expatriate reported that, during the hot season, air-conditioning the homes of its expatriate employees cost her employer an average of $500–$1,000 per home per month. While agreeing that this was exorbitant, she argued that it was necessary to get qualified workers to come to the hardship posting. Not surprisingly, energy costs are ‘considered to be the primary obstacle’ to development in Djibouti (ibid.: 58).
Similarly, elites use great sums of government-earned income to travel abroad for health care, leaving average Djiboutians with only one hospital in the country, eighteen doctors per 100,000 people, a life expectancy of forty-six years, and under-five child mortality rates a shocking 133/1,000 live births (UNDP 2007). It seems that government spending is directed only at the wealthy.
Education
Like infrastructure, education, skilled labour and capable organisation levels are abysmally low, arguably because a skilled populace threatens those guarding access to resources through the state (Hall et al. Reference Hall, Rodeghier and Useem1986). As colonial governments throughout Africa realised, ‘formal education soon revealed its potential as a powerful instrument for struggle in the hands of the colonized and oppressed’ (Assie-Lumumba Reference Assie-Lumumba2006: 30). Primary school enrolment is an appalling 43%, literacy rates among adults are only 27% (RoD 2004), and Djibouti now ranks 171 out of 177 countries on the education indicator of the UN Human Development Index (UNDP 2007). Only a tiny fraction of the population receives post-secondary training. Djibouti has no established university, so all highly educated Djiboutians are trained abroad, often funded by government scholarships. Foreign training raises the costs of education substantially and is therefore reserved for the extremely small elite. As in other resource-producing states, this means a disproportionate amount of the government budget for education is spent on these few privileged individuals (Reed Reference Reed2006). Payment of teachers' salaries is regularly in arrears, forcing many to leave the profession (US SD 2007).
Vocational training and skilled labour are also in severely short supply. Looking at the underdeveloped agricultural sector, for example, the lack of organisational capacity is pronounced: scores of collectives, associations and community organisations exist, but achieve next to nothing because they lack organisational capacity, lobbying skills, technical know-how and material resources, and there are almost no places in which to learn them. Market gardeners report that they learned what they know about coaxing vegetation from Djibouti's arid, infertile soils ‘through trial and error’. Similarly, while three competing animal handling companies exist at the port, port officials report that workers are so underskilled, they are not capable of using the basic animal handling technology that the port acquired at the request of traders. Even low-skill labourers have to be imported from around the world: the author encountered a number of construction workers from the Philippines, Uganda and Tanzania.
Even within the government, most ministries are short on adequately skilled personnel – both in numbers, which are highly concentrated in the capital, and in skill-level, as staff have limited ability to assess trends, to collect and process even baseline data, and to lead major development projects. MAEM, for example, has access to only two public veterinarians (one of whom is also the Presidential Advisor for Food Security and the other is the Director of MAEM's Livestock Division, meaning neither has much time to treat animals), three administrators, about ten veterinary technicians, fifteen animal nurses, and twenty other workers in the Livestock Division – this to serve approximately 500,000 people and 1·2m animals. To make matters worse, senior MAEM officials report that personnel are not replaced when they leave, and no new technicians or vets are being trained, meaning MAEM's skill-level will only decrease in the future.
Diversification of the economy: thwarted by a politics of economic control
Nearly all documents and informants consulted indicated that this extreme lack of human capacity is a major cause of economic hardship and underdevelopment in Djibouti. Without investment in the people, there is little chance for any investment in the rest of the economy to succeed. However, as in some other resource-cursed states, I argue that the Djiboutian government maintains these low levels deliberately. It is undeniably true that economic diversification has been minimal in the tiny ocean-side state, leaving the country with a ‘highly atypical economic structure’ (USAID 2006: v). As Dunning (Reference Dunning2005: 452) has shown for other countries, perhaps the Djiboutian leaders know that they ‘face an important trade-off: while they might like to promote the diversification of the economy, thereby reducing fiscal volatility and potentially improving aggregate economic performance, diversification may create societal bases of power outside of [their] control … [which can] facilitate future challenges to the political power of state incumbents, especially during the economic downturns and fiscal crises that typically characterize resource-reliant countries. Thus, while diversification may be economically rewarding, it can also be politically costly.’
It should come as no surprise, then, that industry is nearly non-existent in Djibouti, where the only factories are a Coca-Cola bottling and water desalinating plant, an ice-making factory, and a flavoured drink and popsicle plant. The private sector remains very small generally, and the GoRD is the country's largest employer (the US Government is the second). In terms of maintaining control, this works in the government's favour, because the many people reliant on the government for their livelihoods are unlikely to protest too loudly against it (IRI 2005). Political reach extends far beyond central government employees to peri-urban and rural areas, where a very large percentage of the population benefits from remittances sent from government-employed relatives.Footnote 11
Even where there are incentives to develop an industry, the GoRD drags its proverbial feet. For example, Djibouti's potential for growth in the fishing industry is vast: a sustainable 48,000 tons/year is estimated to be available in the waters off its 370 km of shoreline (EIU 2005). A catch anywhere near this amount could contribute not only to internal food security, but could also bring in significant export revenue and employment. At present, however, only 350 tons of fish (0·73% of the potential) are landed on average annually by about 90–140 boats – a figure unchanged in over forty years (Thompson & Adloff Reference Thompson and Adloff1968). Fishing now comprises less than 0·1% of GDP annually. The work is considered gruelling, and fishermen are among the poorest segments of the population (RoD 2004).
It does not have to be this way. Part of the trouble for fishermen is that they lack incentives to increase production, largely due to political interference in the industry. The Port de Pêche maintains a monopoly on the purchase of fish for resale, and all catch is required to pass through it for inspection purposes. This system not only reduces prices earned for fish, but also increases transaction costs to fishermen, who docked closer to the market before the Port opened. It also means that fishermen far from the only official port cannot afford to sell their catch legally, resulting in a black market trade to Yemen. Export of fish is hindered by out-of-date regulations, insufficient sanitary standards and the lack of an internationally-recognised ‘competent authority’ to make inspections. The very small amount of fish exported must first be sent to Dubai for inspection and certification.
The government has been given aid to kick-start the industry, including an $11m project by the African Development Bank (ADB) to build a modern fishing port and processing plant. The project took so long to complete, however, that by the time it opened in 1999 it no longer met physical regulations demanded by the European Union countries. ADB agreed to another $2m in order to upgrade the facility, yet this money had not appeared at the time of research, since the GoRD is behind in its required contribution to the ADB.
At least two fairly senior MAEM officials felt that the project, the fourth attempt to privatise and develop the industry, is simply a ‘white elephant’ from which the GORD can milk funds. The current forty-year port concession belongs to an Iranian who is said to be a close friend of President Guelleh. Other potential investors from Italy and Israel have made inquiries about developing the sector, but all have been discouraged by the high input costs and lack of government interest.
While not as potentially profitable as fishing, livestock production is another underexplored area of diversification. Animal husbandry of small ruminants and camels is the sole or main method of livelihood for one third to one half of the country's population and 90% of the rural populace (Doualeh Reference Doualeh2002). A vast majority of Djiboutian livestock producers are extremely poor nomads, engaging in subsistence pastoralism, who are not well integrated into the national economy (RoD 2004). Developing the livestock sector would not be exceedingly difficult, particularly in the peri-urban areas around Djibouti City, yet helping Djiboutians to raise and sell their livestock has not been a government priority. In fact, most animals for local consumption are imported from Ethiopia. According to several informants, the official slaughterhouse is in considerable decay, as hygiene and refrigeration conditions are sub-standard or not-always-functional.Footnote 12 Most animals are slaughtered clandestinely.
Similarly, demand for milk in Djibouti City is extremely high, given the pastoral tradition of its populace, yet only 10% of demand is met through informal peri-urban production. There is no formal method for marketing fresh milk, or regulations governing it. Instead of diversifying into this industry, approximately 1·5 million kg of milk, eggs and other animal products are imported each year.
Despite the official acceptance of private veterinary practice in Djibouti, the GoRD tends to undermine for-profit animal health services by providing veterinary health care, medicines and vaccines to livestock producers free of charge. Some members of MAEM readily acknowledge that this practice undercuts even their own recent attempt to develop a system of auxiliary para-professional veterinary workers in regional areas.
Finally, while the recently opened, USAID-funded, quarantine facility was intended to allow local traders to access to the vast Arabian Peninsula market by providing minimal-cost quarantining, the GoRD thwarted its use as a tool of diversification. In an unexpected Presidential Decree after USAID contributed $4m, Guelleh removed USAID from the project and granted the facility as a concession to Société Abu Yassar, run by a Yemeni–Saudi investor who manages sixty slaughterhouses in Saudi Arabia as well as quarantine facilities in other countries.Footnote 13 Under USAID, the quarantine would have charged only for its service, not interfering in animal selling. Instead, the facility now charges elevated fees for quarantine and buys all animals, thereby monopolising profits since this is the only quarantine approved by Saudi Arabia. Djiboutian and Ethiopian traders interviewed are extremely discouraged at this change; many have found their supplies too small to profitably utilise the facility.
Only in the very recent past has the government begun to address the agricultural sector seriously. This interest does not appear, however, to stem from an attempt to diversify the economy. Instead it hopes to slow the flow of migrants to Djibouti City, whose service provision capabilities have become overwhelmed by rapid urbanisation in the past five severe drought years. The GoRD seems to hope that improving rural conditions will encourage people to stay rural. Still, the UN World Food Programme (WFP) has had to step in to feed those affected by the drought. A top WFP official reported that it is feeding an estimated 100,000 people, or 100% of the rural population, but cannot do so indefinitely.
***
Djibouti is a tiny country with an ‘abnormal’ economy (USAID 2006). Thus, Djibouti had virtually no choice but to make the most of its limited resources and has actually done fairly well for itself given its starting point. Yet even this quick review of the country reveals that it is beset by nearly every symptom of the resource curse.
Djibouti's particular resource curse teaches us a number of important lessons. First, the resource curse can come from non-traditional resources. Not only minerals, oils, timber and drugs can cause curse effects. Military rents, bilateral aid, and indeed possibly other forms of non-labour-intensive, monopolisable rent production, like Djibouti's port, can have the same impact. The implication is that it is not necessarily the existence of specific, extractable natural resources that matters, but rarther dependence upon government-controlled, alienable assets or enclave economies that generate sizable rents. As with oil and natural gas – resources that require massive capital investment and cannot therefore be exploited by small-scale production – the dividends from Djibouti's non-tangible natural resource, strategic geographic location, accrue to the state.
Second, leaders' decisions determine the degree to which the ‘curse’ is present. The initial condition of possessing high-demand natural resources does present peculiar policy challenges that can cycle back to the economy in either positive or negative ways. At best, the GoRD neglects economic diversification. At worst, the government pursues an intentional policy of near-complete control over the economy and, by extension, the people. Control over the economy means that Djibouti's leaders can spread wealth to key constituencies through non-productive handouts, leaving people forever dependent on the government (Collier & Hoeffler Reference Collier and Hoeffler2005).
It is important to recognise, however, that even if Djibouti's leaders wanted to diversify the economy and develop the country, which at least some politicians do want, they are currently stuck in a Catch-22: diversifying the economy requires an educated citizenry, yet educating the masses before the economy is diversified will likely result in a successful opposition. Educated people are more likely to know how to ‘rally the troops’ and wrest power from an incumbent. Additional research on how states have simultaneously increased education and diversified successfully could prove extremely useful for Djibouti's leaders.
Third, Djibouti teaches us that spending patterns of advanced industrial countries in their less-developed country allies can have severe and unintended consequences. For example, the US government's spending on the ‘war on terror’ can unintentionally support repression and abuse. Policy makers need to remember that, as much as assistance is needed by the impoverished in Djibouti and other countries, foreign aid can be manipulated for political ends.
