Beating the resource curse: the case of Botswana, M Sarraf, M Jiwanji

Content: Public Disclosure Authorized
~~w X
Public Disclosure Authorized
Public Disclosure Authorized
Beating the Resource Curse The Case of Botswana Maria Sarraf Moortaza Jiwanji
Public Disclosure Authorized
October 2001 The World Bank
Beating the Resource Curse The Case of Botswana Maria Sarraf Moortaza Jiwanji October 2001 Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Copies are available from the Environment Department, The World Bank, Room MC-5-126.
Contents ABsTRAcr v ACKNOWLEDGMENTS Vii ABBREVIATIONS iX Chapter 1 Introduction 1 Chapter2 The Resource Curse 3 Economic Explanations of the Resource Curse 4 The effect on non-boom tradablesectors andthe Dutch Disease 4 Skill accumulationand the resourcecurse 5 The boom-sector: Low linkages with the rest of the economy 5 The Staple Trap Trajectory 6 Political Dimensions to the Resource Curse 6 Inappropriateeconomic management 6 Rent seeking 7 Policy Suggestions from the Literature 8 Chapter 3 The Case of Botswana 9 Development of the Mining Sector 9 Management of the Mineral Boom 10 Managementofgovernment's budget andaccumulationof internationalreserves : Managementof the exchange rateandeconomic diversification 13 Other Aspects of the Economy 14 Chapter 4 Conclusion 17 NoTEs 19 REFERENCES 21 Environmental Economics Series
Beating the Resource Curse - The Case of Botswana FIGuREs 1. Revenue, expenditure, and development expenditure as a percent of GDP 12 2. Reserves in months of imports of goods and services 12 3. Formal employment by sector, 1989 14 TABLES 1. Comparison between Botswana and Sub-Saharan Africa for selected indicators 9 2. Contribution of the mining sector to GDP, Government revenue, and export earnings in selected years 10 3. Government current revenue, expenditure, and surplus (million current pula) 11 4. Growth of output for selected groups of countries, 1970-1996 13 5. Gini index in Botswana 15
Environment Department Papers
The endowment of natural resources has often been associated with disappointing Economic Development. This phenomenon is referred to in the literature as the "resource curse," which hypothesizes that economies experiencing resource booms, either through price increases or new discoveries, will experience unsustainable growth rates. There are various mechanisms through which a resource-boom can negatively impact on an economy. For instance, it can lead to excessive government expenditure during the boom period and drastic cuts when the boom ends; detrimental impacts
on non-boom tradable sectors; inefficient investment beyond the absorptive capacity of the country; and rent seeking behavior. By exploring the case of the mineral boom in Botswana, this paper will demonstrate that the resource curse is not necessarily the fate of resource abundant countries. The adoption of sound economic policies and the good management of windfall gains have allowed Botswana to continuously manage growth and to become one of the great success stories of developing countries.
Environmnental Economics Series
For providing the original inspiration and for guidance throughout, we are deeply grateful to Kirk Hamilton. We would also like to thank Gary McMahon, Richard Auty, Felix Remy, and
Aziz Bouzaher for reviewing various versions of this paper and providing valuable feedback. Finally we acknowledge the assistance of Thor Sigvaldason in helping to edit the paper.
Environmental Econornics Series
Booming Tradable Sector gross domestic product Gross National Product Heavy and Capital-Intensive Industry Non-Boom Sector Non-Tradable Sector
Environmental Economics Series
1 Introduction
Let us start by asking a puzzling question: if you suddenly inherit a pot of gold, will you be better or worse off in the long run? The obvious reply is better off; few people would argue that more income is a bad thing. Standard economic theory asserts that one can never be made worse off by a positive wealth effect. The answer, however, is not always that straightforward. It depends on how you manage this unexpected gain. If you decide to quit your job, you might spend the money over a couple of years or even take up a loan based on your improved circumstances. Once the pot of gold runs out, you might not be able to get your old job back. If you still have to service a loan, you may then end up worse off. If you decide to invest the money wisely in assets that generate revenue each year for the rest of your life, you will probably be better off.' This simple example illustrates what can happen when countries experience an export boom. The discovery of new natural resources (or a sudden increase in the price of an exportable resource) may unexpectedly increases revenues. Unfortunately, many countries have been unable to properly manage these windfall gains, ending up spending too much, too quickly. This phenomenon has been referred to in the literature as the "natural resource curse." A key question is whether the under-performance of resource-abundant
countries that experience export booms is inherently linked with the fact that they are richly endowed in natural resources. Alternatively, might the symptoms of the curse be avoided through prudent economic management? In this paper, we will show that the natural resource curse is not necessarily the fate of resource abundant countries. By exploring the case of Botswana, we will demonstrate how sound economic policies and good management of windfall gains can lead to sustained economic growth. In the first section, we describe the main dimensions for the resource curse, based on a survey of the current literature. This discussion includes both pure economic explanations for the curse (such as effects on the non-boom tradable sectors, effects on skill accumulation, and the low linkages of the boom sector) and more political dimensions that manifest themselves via inappropriate economic management and rent seeing activities. The paper's second section describes in detail the case of Botswana. This case study begins with a description of the economic development of Botswana since its independence, followed by a closer examination of the development of the mining sector and its contribution to economic growth. Finally, it explores the means by which the Government has managed the mineral boom.
Environmental Economics Series
2/ The Resource Curse
"We are in part to blame, but this is the curse of being born with a copper spoon in our mouths."
consensus on the measurement of resource abundance. The various metrics include export
- Keneth Kuanda, former President of Zambia
dependence (Sachs and Warner 1995), per capita land area (Wood and Berge 1994), export orientation, and population size
Throughout the 1980s and early 1990s many rTerseoouurcgehoaabbuuntddaanntt ccoouunnttrriieess hhaavvee ssuuffffeerreedd ffrroomm
(Surquin and Chenery 1989). In fact Stijns (W20a0e1r)(9a5rageuneosrbthuatwt nthcehgfisnadeings by Sachs and
low growth. This phenomenon has been termed
.inm9thee areno natu resuce
tthhegu"eRsetshoauRrrecseouCrCceuurrssee""abbyyuAtyuatny ((119999,3), wwhomyaeithmasrofnaubaulenodracnce from trade-flows to reserves or
arue tha reouc abnac seealroduction
of natural resources. In contrast,
associated with disappointing rates of economic
Wooduand Berg1 arguesta thetbasic
development. It should not be interpreted as an
findings are not sensitive to these
iron law, but rather as a strong recurrent
tendency (Auty 1994a). Furthermore, it poses an interesting conceptual question as to why a "resource boom," either through an improvement in terms-of-trade or a resource discovery, does not lead to sustained economic growth. The paradox of the resource curse is that a resource boom provides valuable foreign exchange, attracts foreign investment, and provides raw material for production.
The theory side of the resource curse literature seeks to explain the relatively weak performance of resource-abundant economies. Several explanations have been advanced, which offer a range of economic factors both internal and external to the economy. These include effects on production and investment structures within the economy and the price
Some studies, however, find that the resource curse phenomenon is not widespread (Graham 1995 in Mikesell 1997). On balance, however, there is substantial evidence that supports the resource curse hypothesis, ranging from mineral rich to oil-exporting countries. Sachs and Warner (1995), for example, provide a substantive empirical investigation of 96 countries, and find that economies with a high ratio of natural resource exports to GDP in
volatility of international primary goods markets (Auty 1998, Mikesell 1997). On the other hand, some authors argue that the root of the problem is political in that incorrect policy choice leads to the same outcomes and that resource booms inspire rent-seeking activities (Auty 1998, McMahon 1997, Ross 1999). The fundamental question that arises from this literature is whether or not resource abundance is inherently a precondition for poor economic performance (as the above
1971 (the base year) tended to have low growth rates during subsequent years (1971 to
quote by President Kuanda suggests), or whether the curse could have been avoided
1989). Unfortunately, there is no clear
with an appropriate set of policies?
Environmental Economics Series
Beating the Resource Curse - The Case of Botswana
Economic Explanations of the Resource Curse Export booms can cause major distortions in the economies of resource-abundant countries. These distortions tend to affect the structure of production and investment, and can persist long after the boom subsides. They increase domestic income, affect savings and investment, government expenditures, and relative prices in different sectors of the economy. An export boom will typically involve a surge in foreign exchange via increased exports of the affected products (and not other tradable sectors). This leads to an appreciation in the real exchange rate. Changes in the real exchange rate reduce the relative prices of tradable manufactured products relative to non-tradable products, such as construction and services (Mikesell 1997). For instance, in Bolivia the real exchange rate appreciated by 17 percent from 1973 to 1974, and then doubled between 1979 and 1983. As a result non-mining activity was relatively uncompetitive and its share of exports slumped to 5.2 percent of total exports in 1985 (Auty and Evia 2001). However, there is no single reason for the resource curse and not all the suggestions in the literature apply to each resource-abundant country'. The effect on non-boom tradablesectors and the Dutch Disease Many studies seeking to explain the resource curse (such as Usui, 1997; Auty, 1998) place major emphasis on the failure of resourceabundant economies to promote a competitive manufacturing sector, often termed the "Dutch Disease."2 The case of Peru, where GDP growth rates declined from 3.5 percent in 1970-1980 to negative 0.5 percent in 1980-1993, provides an adequate illustration of the damaging consequences for the manufacturing sector. During the 1980s, Peru experienced a resourceboom through dramatic increases in the price of copper and other minerals. The resulting appreciation of the real exchange rate 4
increased the relative export prices of nonboom tradable goods and subsequently damaged Peru's manufacturing and a ged Peru's Mikesellg and agricultural sectors (Mikesel 1997). One way to explain the apparent difficulties experienced in the manufacturing sector is to consider how governments reallocatewindfall gains from the booming tradable (BT) sectors to nontradable (NT) sectors (for example, construction), and non-boom tradable (NBT) sectors (such as manufacturing). A typical rreevspennuuess iintotao rthe oduoem-oesistic econtoomayor((ttbhhaet is, the NT sector), as was the case in Bolivia after the mineral-boom of the 1970s (Auty and Evia 2001). This "spending effect," as it is termed in the Dutch Disease literature (for example, Usui 1997), leads to movements of labor and capital towards the NT sectors (McMahon 1997). One of the reasons for the movement of labor and capital away from manufacturing was that NBT goods became more expensive relative to NT goods through the appreciation in real exchange rates, as was the case in Bolivia during the early 1970s (Auty and Evia 2001). As a result it was quite common for mineral rents to be used for the protection of the NBT sectors through subsidies and protectionist strategies. However, the inadequate performance of the weakened NBT sectors during post-boom downswings required levels of subsidy from the mining tradable sectors that were unsustainable. This was the case in South America (Bolivia, Guyana, pre-Pinochet Chile) and sub-Saharan Africa (Zambia, Zaire and Nigeria) according to Auty (1994b). Another reason for deficiencies in manufacturing sectors is that governments tend to "leap-frog" laborintensive manufacturing industries in favor of heavy and capital-intensive (HCI) industries, as was the case in Mexico and Brazil (Auty 1994a). There is also some evidence to suggest that countries which leap-frogged the laborintensive manufacturing sector displayed Environment Department Papers
The Resource Curse
higher inequalities of income distribution
Sachs (1996) models the incentives to invest in
(Auty and Kiiski, 2001 referring to the experience of southeast Asian countries).
education, by analyzing labor movements between the BT and NT sectors. Changes in resource-boom sectors, as explained above,
In summary, it has been proposed in the
push up the wages in the (unskilled) NT
literature that the greater the natural resource endowment, the higher the demand for NT goods, and, consequently, the smaller the
sectors. Therefore, younger workers will work in NT and they will be better off if they make no investment in education. Birdsall and others
allocation of labor and capital to NBT sectors.
h xett hc
relative losses in NBT sectors can explain low
(1997) found evidence for the differences in the incentives to invest in education between resource-abundant and resource-deficient couries.bHoever theresorif ical
rates of overall economic growth. In terms of the Dutthceh~DD~il~es~se*a~fse~et~Defufecct, thhe evvdidneenhcsebehas been
evidencestoHsuggest that natrl resouce epvroiddeunccteiotno hsausggaenset gtahtaivt enaitmupraalctreosnoutrhcee
inconclusive. In a survey of several empical investigations, McMahon (1997) finds no
manufacturing sector, learning-by-doing, and ultimately economic growth (Stijns 2001).
substantive evidence of the Dutch Disease
effect. In contrast, Auty and Evia (2001) argue
The boom-sector:Low linkages with the rest of
that the Bolivian economy in the early 1970s
the economy
showed clear signs of Dutch Disease through low diversificlaotwiond,verswficaiiotnh,withaa ssmmlalellrethratnhParnimenarlvywihfwbcaradfoad goods are generally produced in an
eexxppeecfatecdtduaaggrriiccuullttuurasaelctsoerctor aanunddmprpotetecicttveeddenlvwihfwbcalriandkfoagades to the rest of the economy. They are
typically highly capital intensive, with a small
international outputs.
(albeit well paid) workforce, and inputs are
Skill accumulationand the resourcecurse The manufacturing sector is often regarded as the principal source of technological progress, and as a consequence there are educational externalities associated with it. This externality effect comes through the knowledge and skills that are generated through the manufacturing sector. As a result, Sachs and Warner (1995) hypothesize that a shift in labor away from manufacturing will depress growth in labor productivity. Furthermore, a deterioration in the manufacturing sector could lead to a reduction in the demand for education and learning-by-doing and a commensurate fall in the long-term growth potential of an economy (Matsuyama 1992). In Matsyama's model, forces that push the economy away from manufacturing and towards agriculture (the boom sector) will lower the growth rate of the economy (cited in Sachs and Warner 1995).
generally imported (Auty and Kiiski 2001). Hirschman (1958), Seers (1964) and Baldwin (1966) established these arguments more formally, by encouraging the view that linkages between the primary sector and the rest of the economy are minimal (in Sachs and Warner 1995). Another implcation of weak linkages of the primary goods sector is that the only way general benefits can be derived from a resource-boom is through the taxation of resource rents. Therefore, in the face of widely fluctuating export revenues, governments may find it diffict to promote economic growth. This disadvantage of low linkages was evident in Bolivia (Auty and Evia 2001). Furthermore, Mikesell (1997) found that governments experienced widely fluctuating export revenues (Mikesell 1997), leading to fluctuating levels in overall government revenues.
Environnental Economics Series
Beating the Resource Curse - The Case of Botswana
The Staple Trap Trajectory
Political Dimensions to the Resource Curse
Auty provides a more formal analysis of the
The economic after-effects of resource-booms
resource curse by specifying a model that
have created hardship for resource-exporting
describes the behavior and incentive structure
nations. These hardships led to persistently
of resource-abundant economies experiencing
slow growth. Furthermore, there appear to have
resource booms. He attributes the symptoms of
bbeeen acctiivee ppoflciceies puurssueed byyggovvenrnumeetnts
the resource curse to a pattern of behavreiosoruhrece boom.Hthat exacerbated the effects of the resource curse
calls a "staple trap" trajectory. Countries on
istuish bete policiesthtwr
this trajectory tend to experience strong Dutch Disease effects, closed economies, few
inianpgpprporopprriaitteeeoor m miisgguidideedd andththosaet twhaertewere related to rent seeking. In either case, the root
incentives for the development of capital, and
cause was substantial increases in government
high dependency on the boom sectors for
revenues caused by the boom.
foreign exchange and revenues. When the
boom subsides, groups with vested interests
Inappropriateeconomic management
block the required adjustments to real
Natural resource booms have often reinforced
exchange and wage rates.
or created inappropriate economic policies in
resource-abundant countries (McMahon 1997).
An example consistent with the staple trap
Though there are a large number of studies on
trajectory is that of Bolivia (Auty and Evia
the implications of resource-abundant
2001). Bolivia was especially vulnerable to the
economies, there is little consensus on which
staple trap. It exhibited Dutch Disease effects,
countries have under-performed as a result of
which had already weakened the non-mining
inadequate policy responses to a resource boom
(Usui 1997). Nevertheless, there are some
econoy. eIvfiecsitmncnyt ws lo andstriking
examples of inappropriate economic
the economy was not diversified. Most of the
mineral windfall was not translated into
investment and went into higher consumption.
McMahon (1997) argues that one of the most
Perhaps most significantly, the resource boom
significant factors behind the negative impact of
created incentives to relax market discipline.
resource booms is the irreversibilityofgovernment
In terms of macroeconomic policy, the
expenditure. When the revenue streams from the
government failed to sterilize the additional
boom subsided, it was very difficult to adjust
foreign exchange inflows and it used its
expenditures down to levels on par with these
hydrocarbon reserves as collateral for foreign
smaller revenues. For instance, in Trinidad and
borrowing, which increased to 78 percent of
Tobago there was public pressure to share the
GDP from 1975 to 1979 (Auty and Evia 2001).
benefits of the boom, which led to large subsidies for food, fuel, utilites, and loss-
Tthheenrreessoouurrccee ccuurrssee ttuhche,esxeisstisoiiss mernely aasstlrtirgoornnowg th tenencrardaslickh,,excpto More significantly, it has been suggested that
mmakkinng reanntteeeroppfriGiseDe.sP.BwByas199a8c11tu, awllhhynennteeatghnaeatiavnen, uthaelre was considerable political difficulty in making cutbacks (McMahon 1997). Furthermore, these
prudent policy can avoid pitfalls (Auty 1994a).
increases in government expenditure
The next section, therefore, illustrates how
sometimes went towards the civil service by
incorrect policy can perpetuate the resource
way of an increase in the number of jobs and
pay levels. In Cote d'Ivoire, for example, the
Environment Department Papers
The Resource Curse
coffee and cocoa booms of 1976 to 1981 led to a 50 percent increase in expenditures on the civil service (McMahon 1997). Similarly, there are many examples of governments investing windfall gains inefficiently. Unproductive investment booms were evident in many countries. Lal and Myint, 1996 (cited in Auty and Kiiski 2001) find that the efficiency of investment in resourceabundant countries collapsed during the 1970s. Furthermore, McMahon finds evidence in several countries of political pressure on government to spread investments towards failing industries. Governments tended to invest the windfalls in the NT sectors (for example, construction), or in projects with low rates of return. Options for investment were not in abundance, given the undeveloped and constrained nature of financial sectors and restriction on holding foreign assets (McMahon 1997). However, the key mistake was that recurrent cost and capacity issues (such as skill requirements) were not taken into account. In fact, the case of Botswana is exemplary of how to avoid these investment pitfalls (see below).
They find evidence that increasing dependency on primary products is positively correlated with closed trade policy (cited in Auty and Kiiski, p14). In order to finance expansionary fiscal policies and/or irreversible government expenditures, governments have been known to borrow on the strength of their booms. Examples of boombased borrowing include Jamaica and Nigeria (Cuddington 1989; cited in McMahon 1997). Mexico implemented a highly expansionary fiscal policy aimed at rapid development, and spent its oil revenues in an imprudent way. Together with the onslaught of capital flight, these led to a severe current account deficit, leading Mexico to an accumulation of shortterm debt (Usui 1997). The example of Mexico stands in stark contrast to that of Indonesia where a conservative stance to foreign borrowing was adopted. Rent seeking Inappropriate economic management can also be influenced by the effort of rent-seekers that are both within and outside of the public
Auty (1993) provides an interesting argument,
sector. For instance, in Brazil, rent-seeking
groups bocked r ratwld
which attributes the low performance of
rov ed reconmo the unoptive
resource-abundant economies to their ability to
rban-idutrpa reas Au 19) incfaptit
postpone refoprmosstdpdouueesrseiimmfoppmllyysttoo tthhe elalragergeurban-industrial areas (Auty 1995). In fact, it
revenues from the primary sector. In particular, these large revenues reduced the incentives to
has been noted that resource-abundant economies are often more susceptible to rent-
develop competitive manufacturing sectors.
seeking behavior, due to the concentration of
Rents are used to support long-standing import
wealth either in the public sector or in the
substitution strategies,3 as part of the
hands of a small number of companies
protectionism of non-booming tradable sectors
(McMahon 1997). Auty (1998) argues that the
(as already mentioned above), long after they
existence of large resource rents distracts
have been of benefit (Auty 1998). This prolonged protection of non-boom tradable
attenton away from long-term economic development goals and towards rent seeking
industries eventually reduces the
competitiveness of manufacturing sectors
(Mikesell 1997). Sachs and Warner (1995b)
Bates (1994 cited in McMahon 1997) argues
hypothesize that governments seek to protect
that the traditional functions of the state, in
the manufacturing sectors (in fear of the Dutch light of a resource-boom, give way to the
Disease) through protectionist trade policy.
redistribution of revenues. As such, socio-
Environmental Econornics Series
Beating the Resource Curse - The Case of Botswana
economic development goals are pushed to the side by rent seeking and patronage. Lane and Tornell (1995 cited in McMahon 1997) contend that a resource boom can lead to a "feeding frenzy" in which rent seekers fight for the natural resource rents. Policy Suggestions from the Literature The resource curse literature, as a result of identifying the nature of the problem as well as analyzing success stories, is furnished with policy suggestions on how to avoid or mitigate the detrimental impact of the resource curse. One caveat to these policy prescriptions is that they cannot be applied across all resourceabundant economies experiencing resource booms. Nevertheless, there are some general principles that arise out of the literature, which stand out in relation to the Botswana experience: * Investment strategy: Do not invest beyond the absorptive capacity; consider all the recurrent costs associated with new investment programs; only invest when the expected rate of return is considerably above that which can be earned in risk-less foreign assets; in fact some argue that it is better to leave windfall gains in the hands of the private sector (Collier and Gunning 1996 in McMahon, p38); pursue investments with high social rates of return, particularly those in human capital and infrastructure (Seymour 2000) * Economic diversification: Improves economic performance by increasing the flexibility with which an economy can respond to external or intemal shocks (Daniel 1992 in Auty 1994b). Failure to
diversify is one important reason why many mineral economies experience such disappointing rates of economic growth (Auty 1994b) * Macroeconomic related policies: Avoid largescale debt; accumulate budget surpluses (for example, Indonesia, Usui 1997); follow a prudent exchange rate management policy by controlling the appreciation of the exchange rate (Mikesell 1997); create a stabilization fund to guard against commodity-price volatility (Seymour 2000). Other more specific policy suggestions include the promotion of autonomous fiscal and monetary authorities in response to the onslaught of rent seekers and special interest groups (McMahon 1997), and the adoption of Environmental and Natural Resource Accounting (EARA) policies which provide a rationale for the effective management of mineral windfalls to secure sustained rapid economic growth (Auty and Evia 2001). The above literature rview highlights several channels of influence through which resource booms have lead to negative effects in many resource-abundant economies. However, it is argued in this paper that being endowed with natural resource wealth is not an inescapable pre-condition for unsustainable economic growth. From these theoretical and empirical investigations of the resource curse, a number of policy suggestions have been put forward. The following case study of Botswana demonstrates how the adoption of such policy suggestions can lead to outcomes in contrast to the resource curse thesis.
Enviroranent Department Papers
3 The Case of Botswana
"...we intend to conserve our resourceswisely and not destroy them. Those of us who happen to live in Botswana in the 20th century are no more important than our descendants in centuries to come." - Hon. Sir QKI Masire, former President of Botswana Since its independence from Britain in 1966, Botswana has become one of the great success stories of developing countries. Between 1966 and 1989, it was the world's fastest growing economy. Botswana's Gross Domestic Product (GDP) grew at an average of 13.9 percent per annum between 1965-80, at an average of 11.3 percent between 1980-89, and at an average of 4.75 percent between 1990-98 (The World Bank 1991, 1998, and 2001). At independence, Botswana was among the twenty-five poorest countries in the world. Since then, the mineral sector has grown significantly and is now a dominant part of Botswana's economy. By 1989 the country was ranked as a lower-middle income economy and in 1998, income per capita had reached $3,460 (in constant 1995 USD) and Botswana was considered an upper-middle
income economy. Compared with other SubSaharan African countries, the economic growth of Botswana appears even more striking (Table 1). This rapid economic growth is not totally surprising given the discovery of large mineral deposits, mainly diamonds. What is remarkable about Botswana is the way in which the mineral boom was managed. By avoiding common pitfalls of mineral booms, such as the "Dutch Disease," the Government was able to continuously manage growth well. The purpose of this section is to explore the factors that led to such and extraordinary economic record, and to draw the lessons learned. Development of the Mining Sector The mining sector is largely dominated by the diamond industry and, to a lesser extent, by copper-nickel. The first diamond mine was discovered in 1967 by De Beers in the region of Orapa. Since then this mine has yielded more than 118 million carats of diamond. Today, diamond mining is dominated by the Debswana
Table 1. Comparison between Botswana and Sub-Saharan Africa for selected indicators
|Sub-Saharan Africa
1965 1989 199841 1965 1989 1998
GNP per capita (constant 1995 USD) Life expectancy at birth (years) Infant mortality rate (per 1,000 live births) Number of physician per 100,000 persons Pupil to teacher ratio inPrimary School Source: The World Bank 1991,1998, 2001.
379 2,84 3,460 529 581 539
50 57 465 43 51 50
108 55 58 149 101 94
4 19 20
40 32 25 42 42 41
Environmental Econornics Series
Beating the Resource Curse - The Case of Botswana
company which is jointly owned by the government (more than 50 percent) and by DeBeers. In 1994, the three main diamond mines in Botswana-Orapa, Letlhakane and Jwaneng- produced 15.5 million carats of diamond.
Table 2. Contribution of the mining sector to GDP, government
revenue, and export earnings in selected years
Mineral share of Mineral share of Mineral share of
total GDP
total revenues export earnings
Mineral development was the
major contributor to the growth in
Botswana's economy. The mining
sector reached its peak in 1989, when it accounted for more than 50 percent of GDP (Table 2). It was also a major generator of revenue for the government. The mineral share of government revenue grew from almost nothing at
Source: Central statistics office referred to inB.Gaolathe 1997, and in Harvey and Lewis
independence to about 60 percent in 1989. It has since been oscillating around 50 percent of total revenue. The contribution of diamonds and copper-nickel to total exports
followed by fast increase in government spending, resulting in an overall increase in domestic costs. Consequently, the competitiveness of other tradable goods6 are
varied between 75 percent and nearly 90 percent as of 1980 (except for 1981 when the diamond market was depressed). Despite its
hlkely to be reduced creating a slowdown in the growth of other sectors of the economy. Another common problem resulting from export booms
dominant role in economic production, government revenue, and exports, the mining sector is not a major employer. Due to its capital intensive structure, the sector only employed around 9 percent of the labor force in the early 1980s, and around 4 percent in 1989.
is that, once the boom is over, governments find it very difficult to cut back on spending, creating serious economic consequences and possibly a debt crisis. Botswana has managed to avoid most of the
economic problems associated with export booms by adopting appropriate macroeconomic
In the 1970s, many countries experienced important export sector booms. Mexico and Nigeria benefited from large increases in oil prices. Colombia, Cote d'Ivoire and Kenya benefited from an increase in coffee prices in the mid-1970s, as well as from technological improvements in coffee production. However, in the years that followed, many of these countries suffered from balance of payment problems and debt crises. As explained above, one of the reasons underlying this effect is that an increase in government revenues (resulting from a boom in the export sector), tends to be
policies. Two main objectives guided Botswana's economic policies: avoid external debt and stabilize growth on one hand, and encourage economic diversification on the other (Hill and Mokgethi 1989). The economic policies adopted by the government to achieve these objectives included the following. To avoid excessive increases in expenditure during boom periods, the government accumulated international reserves and ran budget surpluses earmarked for stability spending in leaner periods. This policy
Enviromnent Department Papers
The Case of Botswana
avoided having to drastically cut expenditures during bad years and reduced inflationary pressures. A second central policy was to manage the nominal exchange rate to avoid real appreciation of the local currency. This was achieved largely through the accumulation of intemational reserves. Preventing the local currency from appreciating allowed other tradable goods to maintain competitiveness on world markets, and hence encouraged economic diversification. Management of government's budget and accumulationof internationalreserves
This ability to maintain expenditures in line with long term growth had been a strong stabilizing force. Investment decisions: Given the high level of savings, the government had to decide how to allocate its surpluses between international reserves and domestic investment. A common problem in countries where windfall gains accrue principally to the government is that it will tend to invest in projects that have lower rates of return vis-a-vis the private sector. It is often argued, therefore, that a substantial part of the windfall gains should be kept in the hand of
As shown in Table 2, mineral revenues constituted a major source of goverment revenue. Consequently, any fluctuation in diamond revenues would directly affect government revenue and hence spending. In line with the above objectives, the government decided not to increase spending whenever revenue increased. Instead, it based expenditure levels on longer-term expectations of export earnings and government revenues. Despite considerable political pressure, the temptation to spend everything in the treasury was successfully avoided. Instead, any excess revenue was used to build up foreign exchange reserves at the Bank of Botswana. These reserves were thus available to be drawn on in years when revenues were low. Controlover expenditures:Government savings can be measured by the budget surplus. As can be seen in Table 3, the government saved a large fraction of revenues, avoided excessive increases in expenditure, and sustained high recurrent surpluses. Faced in 1981/82 and in 1994 with a fall in diamond export earnings, surpluses declined while expenditure levels were maintained. The government thus managed to avoid drastic cuts in expenditures when revenues decreased.
the private sector. However, the public allocations of the gains in the case of Botswana were done very wisely.
Table 3. Government current revenue, expenditure, and surplus (million currentpula)
revenue expenditure surplus
Source: Statistical Information Management and Analysis, The World Bank.
Envirornmental Econornics Series
Beating the Resource Curse - The Case of Botswana
Domestic investments, as measured by
revenues), they were intentionally kept well
development expenditure, were based on expected intermediate and long term revenue
below revenue levels. Avoiding over-investment also helped reign in inflationary pressures.
flow, taking into account the recurrent expenditure iAvolved in each new development project. Faced with a boom, countries often
Internationalinvestment: Between 1976 and 1996, foreign exchange reserves increased from
make investment decisions without proper account of the recurrent expenditures associated
US$75 million to US$5 billion. Once again, the rationale to save windfall gains from mineral
with each new investment. As a result, once the
revenues for use when export receipts declined.
boom is over countries are unable to cover the recurrent costs of all the new investments, and
The part of the recurrent surpluses (Table 3) not invested domestically was used to accumulate
projects often have to be abandoned before completion.
foreign reserves. By 1995, the government had accumulated reserves equal to 25 months of import cover (Figure 2). If these reserves were
In Botswana, however, the government astutely
not accumulated during booms, the government
avoided undertaking any new development
would have had to either borrow or reduce
projects if there was no provision to
Figure 1. Revenue, expenditure, and development
The government has consistentlyexndtrasapcntoGD
produced National development plans expenditure as a percent of GDP
that determined its spending. Once a
plan is voted into force by parliament,
Reurent revnue
it is illegal to implement any
X so
additional public projects without
going back to parliament. This system %
expdiure /
has proven to be effective in
controlling spending, because it
20 -
prevents the inception of a project for
" -" " ' -
which no provision was made to cover
the total costs over time.
0 AN ^Ap'^3 \*
^ o o O 'q N"qA3Nq
Moreover, decisions to invest domestically took into account the absorptive capacity of the economy. Since the availability of skilled manpower was a large constraint in
Source: Statistici Information Management and Analysis, The World Bank. Figure 2. Reserves in months of imports of goods and services 7 30
Botswana, the government felt that increasing development expenditure beyond the capacity of the country would result in a rate of return lower
2S t 20 '8 is5
that what could have been earned on alternative assets (international
reserves in particular). With reference
to Figure 1,although development spending could have been further increased (given the huge inflow of
0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 Source: International Financial Statistics, International Monetary Fund.
Environment Department Papers
The Case of Botswana
imports when diamond revenues declined. Instead of having to respond under pressure with drastic, fast acting policies, the government had considerable leeway for more thoughtful decision making. Another advantage of investing a part of the windfall gains in foreign exchange rather than domestically is that it prevents the exchange rate from excessively appreciating. When hit with the decline in diamond earnings in 1981/82, as mentioned above, the government used some of the store of savings to avoid any drastic decline in expenditures. However, faced with uncertainty conceming the duration of the diamond crisis, the government took the precautionary measure of assuming that it might be a long term crisis rather than a temporary one. As a result, it quickly adopted a package of adjustment policies, which included limiting increases in lending, increasing the prime lending rate, freezing wages and salaries and devaluing the pula.8 The rationale for reacting quickly with a package of policies was twofold. First, to spread a moderate impact on the whole economy rather than making one drastic change in a particular sector. Second, to take moderate measures before the situation became too serious and more extreme actions might be required. Management of the exchange rate and economic diversification The Government of Botswana paid close attention to the management of foreign
exchange rates. A common problem in countries experiencing export booms is an appreciation of the local currency leading to a reduction in the competitiveness of other sectors in the economy. By avoiding real appreciation of the pula, other traded goods could continue to compete successfully. Botswana's policy was, therefore, to use the exchange rate as a tool to promote economic diversification. The accumulation of foreign exchange reserves during boom periods was consistent with this objective. However, even if economic diversification was a goal, the government had also to worry about inflation. In times where the stability of domestic prices dominated the diversification goal, the pula was allowed to appreciate. The effort of the govemrnent to diversify the economy can be seen in Table 4. Although the share of manufacturing value added in GDP is very small (only 5 percent in 1995), its growth has been very dynamic, especially when compared to the average growth of the manufacturing sector in Sub-Saharan African countries or upper-middle income economies. The growth of the service sector was also very important, far outweighing the growth in other similar countries. The agricultural sector, as dsuiscccueessde. dItbgerloeww,awt aasvtehrey sselocwto,r othccaatsdioidnanlloyt negative rate, and its contribution to GDP fell from 13 percent in 1980 to less than 4 percent in 1995.
Table 4. Growth of output for selected groups of countries, 1970-1996
(average annual growth rate, percentages)
Man acturin
70-80 80-90 90-96 70-80 80-90 90-96 70-80 80-90 90-96 70-80 80-90 90-96 1 70-80 80-90 90-96
Botswana 14.5 10.3 4.1 8.3 2.2 -1.2 17.6 11.1 1.8 22.9 8.8 2.6 14.8 11 7.1
Uppermiddle income
5.9 1.4 2.9 3.2 2.5 1.7 6.1 0.7 2.9 6.6 1.3 3.4 6.3 2.0 3.7
Sub-Saharan 3.8 1.7 2.0 1.7 1.8 2.1 3.8 1.1 0.9 4.3 1.3 0.8 4.9 2.2 2.0 Africa Source: The World Bank 1995 and 1998.
Environmental Economics Series
Beating the Resource Curse - The Case of Botswana
Given that the employment potential of the mining sector is limited due to the capital intensive nature of its operation, the creation of jobs in other sectors of the economy was a major concern for the governnent. Even if the government's overall strategy was to minimize intervention, it still adopted a few incentives to create jobs in the manufacturing and services sectors.10 Despite the small size of the manufacturing sector, it had great potential to create employment, especially in the construction sector (Figure 3). Employment in manufacturing and services has grown tremendously during the past decades and in 1989 they represented 25 and 32 percent of formal employment respectively. Although employment in agriculture only represents 4 percent of formal sector employment, it pacrccoeuunnttssffoorrtthheemmoafajojorriittyyemofpsiieinnncfftooeorrrmmnaaalllizsaetcitoonr employment. It is important to note that in Botswana formal sector employment only represents 30 percent of the total labor force. The remaining 70 percent constitute the informal sector employment.11
Moreover, the legal framework to control environmental impacts of mining operations is well established.1 Examples of actions taken to protect the environment during mining operations include: the incorporation of a sulfur reduction plant into the Selebi-Phikwe coppernickel smelter; the installation of power lines underground to save the flamingo population near the Sua Pan soda ash company; and the suppression of coal dust by the wetting system in the coal mine of Moruplue Colliery (B. Gaolathe 1997). Although diamond mines do not pose serious environmental hazards, various measures were still taken to ensure that diamond mining operations would be as environmentally friendly as possible. A major aspect underlying Botswana's environmental protection from mining activities is the prvt oprtr of eneveirreopnnmilental cosots, whereby 'im et e ws measue Other Aspects of the Economy In addition to the protection of the environment from mineral activities, Botswana is among the
Not only has Botswana's government successfully managed revenues from the mining sector, it has also paid considerable attention to environmental protection in implementing mineral projects. Commitments to environmental protection were part of the concession agreements with mining companies.
top ten countries in terms of preserving its natural environment; about 19 percent of its total land area is estimated to be nationally protected (The World Bank 1998). This area comprises national parks and game reserves, some of them having gained interational fame for their abundant wildlife (for example, the Central Kalahari Game reserve and the
Figure 3. Formal employment by sector, 1989 12 Agriculture
Gemsbok National Park). Despite these accomplishments, it is fair to say that Botswana has not solved all of its environimental
problems. Large shortcomings remain and need
Manufacturing 25%
further efforts, such as deforestation from overgrazing, soil erosion and water pollution.
t | l l lill
Through many years of sustained growth, there
have also been two main shortcomings which
i 1 38 X 1 Mining contrast with the overall picture of economic
4% R B
disuccess: the agricultural sector, and income
32% Source: Labor Statistics 1989, Gaborone, Republic of Botswana cited inPerrings (1996).
The agriculturalsector:The poor performance of the agricultural sector can be seen in Table 4.
Environment Department Papers
The Case of Botswana
Between 1980-90, while the industrial,
inequality has neither improved nor worsened
manufacturing and services sectors grew at an average rate of 11 percent, 23 percent and 11
in Urban areas and has barely improved in rural areas. Table 5 presents the Gini14 index
percent respectively, the agricultural sector
calculated in each of the 3 surveys undertaken
grew at only 2.2 percent. More recently, between
in Botswana since independence. World Bank
1990-96, agricultural GDP contracted at an
estimates of income distribution also suggest
average of 1.2 percent. The decline in
that very little improvement (almost none) has
agricultural productivity is mainly the result of
been made since independence. In 1970-75, it
severe and prolonged periods of drought
was estimated that the share of income accruing
experienced by the country and the over-
to the lowest 20 percent of the population was 4
utilization of rural resources. The evidence for
percent, and the share of income accruing to the
increasing pressure on scarce rural resources is
top 20 percent of the population was 60 percent.
visible through the depletion of water reserves
By 1986 these numbers had changed to 3.6
available to villages, water pollution problems,
percent and 58.9 percent respectively.
overgrazing, rangeland degradation, and the
depletion of wood around large settlements.
Identifying the factors responsible for the lack of
While the dependence of agricultural growth on improvement in income distribution in
the sustainable use of agricultural resources has
Botswana is a complex task. Nevertheless, two
been recognized in Botswana, action to improve
issues seem relevant. First, many authors
the situation has been lacking. Given that
recognize that long periods of drought are
agriculture accounts for 70 percent of the labor
generally associated with an increase in
force, this is an issue of considerable concern.
inequality. This occurs because large and rich
As explained in paragraph 46, formal sector
farmers with large reserves of livestock can
employment constitutes only 30 percent of the
better endure drought than poor ones. Second, it
labor force, while the informal sector
is also accepted that rapid economic growth and
employment accounts for the remaining 70
resource-abundance usually results in
percent, most of which is in the agricultural
increasing inequality. In the context of the
sector. Robust employment in the agricultural
mineral development of Botswana, this view
sector would have a very positive affect on the overall rate of unemployment, which is
seems to be correct; the direct benefits were mainly restricted to a fortunate minority.
currently quite high.
Despite Botswana's bad performance in terms
of income distribution, it is still favorably
Income distribution:Unequal income
comparable with other resource-abundant
distribution in Botswana has always been
countries, such as Brazil and Peru (Auty,
recognized as a major problem. Many claim that
personal communication). Nevertheless, the
income inequality has been widening since
Government's recognition of this problem has
and that much of Table S. Gini index in Botswana
the population
Rural income household income and Household income and
did not benefit
distributionsurvey expenditure survey
expenditure survey
from the overall
growth in the
Per capita urban
economy. Income areas
Per household
surveys indicate
rural areas
that income
Source: Central Statistics Office, cited In Hudson and Wright 1997.
Environmental Economics Series
Beating the Resource Curse - The Case of Botswana
pushed it to invest in many social aspects of the economy, such as education and health. But the problem that seems to be facing a now better
educated population is one of serious unemployment (estimated at 21.2 percent in 1994).
Environment Department Papers
4 Conclusion
The "resource curse" literature shows, through various channels of influence, how resource booms can actually harm resource-abundant economies. A review of the literature offers both economic and political dimensions to the problems of resource-abundant economies. For instance, an increase in copper prices in Peru lead to an appreciation of the real exchange rate which subsequently damaged the manufacturing and agricultural sectors. In Bolivia, low linkages of the booming mineral sector to the rest of the economy did not promote sustainable economic development. In terms of political influences, Brazil's attempts to remove protection for uncompetitive sectors were blocked by rent seeking groups. In short, the curse of resource booms is very real. However, the question raised in this paper is whether or not resource-abundance and subsequent resource-booms lead to inevitably low economic performance. The case of Botswana illustrates how a natural resource curse is not necessarily the fate of all
resource abundant countries, and that prudent economic management can help avoid or mitigate the detrimental effects of the resource curse. The discovery of large diamond deposits allowed Botswana to witness an important export boom and the world's fastest growth in GDP. The country moved from being the 25th poorest country in 1966 to an upper-middle economy thirty years later. The most important factor in Botswana's long term sustained economic growth was its ability to avoid common problems associated with export booms and the adoption of sound economic policies. Its main objectives were to avoid extemal debt, stabilize growth and to encourage economic diversification. This paper has explored the various economic policies adopted by the Botswana government to achieve those objectives. Even if the agricultural sector and income distribution have had a less successful fate, various lessons could be drawn from Botswana's capacity to manage the revenue of the resource booms and to sustain long term economic growth.
Environmnental Economics Series
1. This example is drawn from a similar example cited in McMahon (1997). 2. The Dutch Disease model is named after the disappointing experience of the discovery of natural gas in the Netherlands. The term is popularly used to refer to all economic hardships associated with resource exports. Its more formal definition, however, describes two effects of a resource boom: an appreciation of the real exchange rate; the tendency to draw capital and labor away from non-boom tradable sectors, making them more un-competitive (Ross 1999). 3. Import substitution strategies were promoted by the United Nations Economic Commission, as part of the development campaign of the 1960s and 1970s (cited in Sachs and Wamer 1995). 4. 1998 or latest available year. 5. The decrease in life expectancy since the early nineties is mainly due to the high incidence of HIV infection in Botswana. 6. By other tradable goods, it is meant tradable goods other than the one experiencing a boom (such as minerals, in the case of Botswana). 7. Represents international reserves expressed in terms of the number of months of import of goods and services which could be paid for.
8. For additional information please refer to Hill and Mokgethi (1989). 9. Industry includes mining. 10. For more information on this issue, please refer to Siwawa-Nadi (1996). 11. The size of the labor force estimate is based on the projection of the population assumed to be economically active. Informal sector employment is the difference between the estimated labor force and formal employment (Perrings 1996). 12. Manufacturing includes: electricity, water, and construction. Services include commerce, transport, and financial. 13. The legislation related to the environmental impacts of mining operations include: The Atmospheric Pollution Prevention Act of 1971; the Mining, Quarries, Works, and Machinery Act of 1973; and the Mines and Mineral Act of 1976. 14. A Gini index measures the extent to which the distribution of income among individuals or households within an economy deviates from a perfectly equal distribution. A Gini index of 0 represents perfect equality while an index of 100 implies perfect inequality (The World Bank 1998).
Environmental Economics Series
Auty, R. 1993. Sustaining Development in Mineral Economies: The Resource Curse Thesis. Routledge, London. ___. 1994a. "Industrial Policy Reform in Six Large Newly Industrializing Countries: The Resource Curse Thesis." World Development, Vol 22, No 1 pp. 11-26. . 1994b. "The Resource Curse Thesis: Minerals in Bolivian Development, 197090." Singapore Journal of Tropical Geography, Vol. 15, No. 2 pp. 95-111. . 1995. "Industrial Policy, Sectoral Maturation, and Postwar Economic Growth in Brazil: The Resource Curse Thesis." Economic Geography vol. 71 No. 3. ____. 1998. "Resource Abundance and Economic Development: Improving the Performance of Resource-Rich Countries." The United Nations University World Institute for Development Economics Research, Helsinki. Auty, R., and J. L. Evia. 2001. "A Growth Collpsewith Point Resources: Bo-ivia." CCohlalapptseerr 111 iinn AAuuttyy,,RR..,, eedd.. RReessoouurrcceeInequality Abundance anAd~bEuncodna~onmceic D~ evel~opme~nt. ~ WIDER Studies in Development Economics, OxfordOUUxfnnoivdveerrssittyy PPreessss. Auty, R. and S. Kiiski. 2001. "Natural Resources, Capital Accumulation, Structural Change and Welfare." Chapter 2 in Auty, R., ed. Resource Abundance and Economic Development. WIDER Studies in Development Economics, Oxford University Press. Birdsall, N., D. Ross, and R. Sabot. 1997. "Education, growth, and inequality." In N.
Birdsall and F. Jasperson, eds. Pathways to Growth:ComparingEast Asia and Latin America. Washington DC: Inter-American Development Bank, 93-127. Gaolathe B. 1997. "Development of Botswana's Mineral Sector." Aspects of the Botswana Economy, Selected Papers. Gaolathe N. 1997. "Botswana's Boom and Recession Experience: A Discussion." Aspects of the Botswana Economy, Selected Papers. Harvey, C., and S. Lewis. 1990. Policy choice and development in Botswana. St Martin's Press, New York. Harvey, C. 1993. "Botswana: Is the economic miracle Over?" Journal of African Economies, Vol. 1, No. 3. Hill, C., and N. Mokgethi. 1989. "Botswana: Macroeconomics Management of Commodity Booms, 1975-86." Successful Development in Africa, Economic Development Institute of The World Bank. Hudson, D., and M. Wright. 1997. "Income * v in B,otswana-Trends Since ~EIcndnepoeyndenSceel.c"teAsPpeacptesrso.feethlpemBnottswana Economy, Selected Papers. caoC,uGrse: Myth9o7r RTeaeNlittyrl?R"sumcimeo, World Bank Institute. Mikesell, R. "Explaining the Resource Curse, with Special Reference to Mineral-Exporting Countries." Resources Policy, Vol. 23, No. 4, pp. 191-199, 1997 Mohohlo, L. 1997. "Central Banks as Protectors of National Wealth: Botswana's Case." Aspects of the Botswana Economy, Selected Papers.
Environmental Economnics Series
Beating the Resource Curse - The Case of Botswana
Perrings, C. 1996. SustainableDevelopment and PovertyAlleviation in Sub-SaharanAfrica: The Case of Botswana. Macmillan Press, London. Ross, M. 1999. "The Political Economy of the Resource Curse." World Politics, Vol. 51, No. 2, pp. 297-322. Sachs, J., and A. Warner. 1995. "Natural Resource Abundance and Economic Growth." Development Discussion Paper No. 517a, Harvard Institute for International Development. Seymour, J.M. 2000. "East Tmior's Resource Curse?" Global Policy Forum, UN Security Council. . Siwawa-Ndai, P. 1997. "Industrialization in Botswana." Aspects of the Botswana Economy, Selected Papers. Stijns, J. C. 2001. "Natural Resource Abundance and Economic Growth Revisited." (Unpublished Second Draft) University of California at Berkeley.
Syrquin, M., and H. B. Chenery. 1989. "Patterns of Development 1950 to 1983." World Bank Discussion Paper 41, Washington DC: World Bank. Usui, N. 1997. "Dutch Disease and Policy Adjustments to the Oil Boom: A Comparative Study of Indonesia and Mexico." Resources Policy, Vol. 23, No. 4, pp. 151-162. Wood, A., and K. Berge. 1994. "Exporting Manufactures: Human Resources, Natural Resources and Trade Policy." Journal of Development Studies, Vol. 34, pp. 35-59. The World Bank. 1998. World Development Indicators. . 2001. World Development Indicators. . 1991. World Development Report. . 1995. World Development Report. Wright, M. 1997. "The Use of Mineral Revenues in Botswana: Super Caution vs. Pragmatism." The Research Bulletin, Vol. 15 No.1, Bank of Botswana.
Environment Department Papers
Environment Department The World Barik 1818 H Street, N.W. Washington, D.C. 20433 Telephone 202-473-3641 Faxsimile 202-477-0565 0 Printed on 100% post-consumer recycled paper

M Sarraf, M Jiwanji

File: beating-the-resource-curse-the-case-of-botswana.pdf
Title: World Bank Document
Author: M Sarraf, M Jiwanji
Author: World Bank Group
Published: Tue Sep 24 15:03:50 2002
Pages: 36
File size: 1.32 Mb

, pages, 0 Mb

What is NaProTechnology, 3 pages, 0.65 Mb

, pages, 0 Mb

ENTITLEMENT TABLES, 24 pages, 1.37 Mb

Circle Game (The, 218 pages, 0.97 Mb
Copyright © 2018