Uppsala University Department of Economics

Diamonds and sustainable growth – The success story of Botswana

Joakim Hilldén and Johan Mesterton

Bachelor Thesis Minor Field Study

Fall 2005 Thesis advisor: Javad Amid

Abstract

Numerous studies have confirmed a statistically significant negative relationship between

natural resource abundance and economic growth. This has been labeled “The Resource

Curse”. In this paper we try to explain why Botswana, a country heavily dependent on its

diamond industry, has managed to generate sustainable growth. Economists have

advanced several explanations for the negative impact of natural resources on long-term

growth. This paper focuses on the following important problems: First, a boom in a

natural resource can pull resources away from other sectors of the economy, thus harming

their international competitiveness, a phenomenon called the Dutch disease. Second,

abundance in natural resources may lead to poor institutional quality in many countries.

Thanks to conservative fiscal policies and accumulation of foreign reserves the local

currency did not appreciate during the boom, and Botswana avoided the most severe

symptoms of the Dutch disease. Historical tradition of democratic procedures and sound

institutions at the time of diamond discovery has contributed to a high institutional

quality in Botswana.

Keywords: Botswana, Diamonds, Natural resource abundance, Resource curse, Dutch

disease, Institutional quality.

Acknowledgements: Research for this paper was carried out in Gaborone between April

and June 2005. First we would like to thank SIDA for giving us the opportunity to write

this paper. We also would like to thank Annika Jagander and Stefan Andersson at the

Swedish Embassy for their generosity and care, Keith Jefferis for valuable insights about

the Botswana economy and Anders Sandström and the Nilsson family at Sanitas for their

hospitality.

1

1. INTRODUCTION........................................................................................................... 3

2. THE THEORY OF THE CURSE OF NATURAL RESOURCES .................................. 5

2.1 THE DUTCH DISEASE .................................................................................................. 6

2.1.1 The symptoms of the Dutch disease ................................................................... 6

2.1.2 Avoiding the Dutch disease................................................................................ 8

2.2 INSTITUTIONAL QUALITY, CORRUPTION AND NEGLECT OF EDUCATION .................... 9

3. HOW BOTSWANA BEAT THE RESOURCE CURSE ............................................... 13

3.1 THE DUTCH DISEASE IN BOTSWANA ........................................................................ 15

3.1.1 To what extent did Botswana suffer from the Dutch disease? ......................... 15

3.1.2 What did Botswana do to prevent Dutch disease effects? ............................... 18

3.2 INSTITUTIONAL QUALITY, CORRUPTION AND NEGLECT OF EDUCATION .................. 23

4. EXPLAINING THE SUCCESS OF BOTSWANA........................................................ 29

4.1 AVOIDING THE DUTCH DISEASE............................................................................... 29

4.2 BOTSWANA’S HIGH INSTITUTIONAL QUALITY ......................................................... 31

5. FINAL DISCUSSION.................................................................................................... 34

REFERENCES................................................................................................................ 36

APPENDIX...................................................................................................................... 38

2

1. Introduction Many economic studies describe Botswana as one of Africa’s star performers. The former

British colony achieved independence in 1966 and was at that point one of the poorest

countries in the world. Because of almost total neglect from 1885 to 1966, the

independent government of Botswana inherited an economy that was very

underdeveloped even with African standards: Botswana had a railway through the eastern

part of the country but there were almost no tarred roads, and secondary school had only

graduated 100 students until 1966 (Harvey and Lewis 1990:15-23). Being landlocked,

lacking in infrastructure and lacking in skilled labour and educated people, not many

things pointed to a successful economic development for the country. However, during

the post-independence period Botswana has transformed into a fast growing economy.

The country has been characterized by an impressive macroeconomic performance.

During the three decades after independence, Botswana was the fastest growing economy

in the world with real per capita income growth averaging 8.2 percent per annum during

1966-96 (Hope 2002:2). Over the ten-year period from 1993 to 2003 the per capita

income growth was more modest, averaging approximately three percent per annum

(BIDPA 2004). Still, Botswana is today one of the few African economies to be classified

as upper middle income by the United Nations and the World Bank.

Many observers have claimed that this economic development has been realised thanks to

the large findings of diamond deposits in the 70’s and 80’s. However, considering the

large bulk of economic research that has shown how often natural resources lead to poor

economic stability and growth, one should rather state that Botswana achieved enviable

long-term macroeconomic development despite the big growth in the diamond mining

industry. The difficulty in generating sustained growth from a natural resource could be

expressed in the words of former Zambian president Kenneth Kaunda, who explained his

country’s poor economic performance by stating that: “We are in part to blame, but this is

the curse of being born with a copper spoon in our mouth” (Saleshando 2005).

3

Intuitively, one would expect an abundance of natural resources to be a blessing for an

economy. However, in almost all cross-country regressions, natural resources are found

to have a statistically significant negative impact on a country’s growth (Sachs and

Warner 1995:2; Sachs and Warner 2001:828-31). This phenomenon has been labelled the

resource curse, a name implying that countries with an abundance of natural resources are

cursed rather than blessed. Botswana, one of the world’s biggest exporters of diamonds,

provides an intriguing example in this regard since it has successfully managed to avoid

the biggest difficulties in dealing with the revenues from its natural resources and

obtained world record growth rates during decades after independence. While many other

resource-dependent economies stagnated, Botswana’s economy blossomed and

seemingly managed to beat the infamous resource curse.

Economists have advanced several explanations for the negative impact of natural

resources on long-term growth. This paper focuses on several important problems that

have appeared in resource-abundant countries. First, a boom in a natural resource can pull

resources away from other sectors of the economy, thus harming their international

competitiveness, a phenomenon called the Dutch disease. Second, abundance in natural

resources has lead to poor institutional quality and led to rent seeking behaviour in many

countries. Also, resource-abundant countries appear to pay less attention to education,

which is essential to development. These reflections lead us to the following research

question: Why has Botswana not suffered from the resource curse?

This paper will begin with a presentation of the resource curse theory. We will present

empirical evidence that supports the existence of a curse of natural resources, and present

the different explanations for the curse. In chapter 3 we discuss the impact of the

diamond boom on the country’s macroeconomic performance. Here we assess the

question of how Botswana avoided the resource curse. In the fourth chapter we try to

explain the most important factors to Botswana’s good economic performance. Finally, in

our final discussion, we discuss Botswana’s present situation and also comment on some

problems the country is facing.

4

2. The Theory of the Curse of Natural Resources Numerous studies have confirmed a statistically significant negative relationship between

natural resource abundance and economic growth. The pioneering work on the resource

curse theory was done by Jeffrey Sachs and Andrew Warner who have thoroughly

analyzed, both theoretically and empirically, the effect of natural resources on economic

growth. In an often cited article, Natural Resource Abundance and Economic Growth,

they show that, controlling for initial GDP per capita level, an increase of one standard

deviation in the share of primary exports is associated with a decrease of 0.93 percent in

annual GDP per capita growth (Sachs and Warner 1995:8). This intuitively surprising

empirical finding is now widely accepted and has even been said to be “...one of the most

robust findings in the empirical growth literature...” (Bulte et al. 2004:1). Examples of

countries seemingly cursed by their natural resources are not hard to find. In Africa, the

so called "blood diamonds" are widely believed to play an important factor in explaining

the civil wars, and consequently poor growth, in countries such as Sierra Leone and

Angola. In Latin America, both Bolivia and Venezuela actually found themselves having

lower GDP levels after their respective oil booms (Sachs and Warner 1999:51).

Many scholars have tried to explain the existence of this resource curse. One simple

explanation has been that an abundance of natural resources could lead to what a scholar

has labelled “overconfidence” (Gylfason 2001:847). The intuition behind this explanation

could be formulated in the words of the French sixteenth century philosopher Jean Bodin:

“Men of a fat and fertile soil, are most commonly effeminate and cowards; whereas

contrariwise a barren country make men temperate by necessity, and by consequence

careful, vigilant, and industrious” (quoted from Sachs and Warner 1995:4). The idea is

that a society with an abundance of natural resources and easily gained wealth tends to

become lazy and so confident of its wealth that it does not find motivation for hard work

and economic reforming – growth is simply taken for granted.

Obviously, other explanations have also been advanced to explain theoretically the

empirical findings of the resource curse. In this paper, we will focus on two different

5

channels of the resource curse. We distinguish both purely economic explanations as well

as more political theories of the transmission mechanism between natural resource

abundance and poor economic growth: (1) A natural resource boom can have negative

macroeconomic effects on an economy, pulling resources away from other sectors of the

economy, a phenomenon called the Dutch disease. (2) Abundance in natural resources

has been shown to induce poor institutional quality, high corruption and lead to a neglect

of education, all of which have a negative impact on growth.

2.1 The Dutch disease

The phenomenon Dutch disease has gotten its name from the effects that North Sea gas

production had on the Dutch economy. Large revenues from the gas production forced

the Dutch guilder to appreciate against other currencies which weakened the

competitiveness of the tradables (e.g. manufacturing) sector and caused unemployment.

The traditional Dutch disease model, which is used for a medium run analysis, describes

an economy with three sectors: the booming sector, the lagging sector and the non-

tradable sector. The booming sector could be the oil industry or any other major

exporting industry. In Botswana’s case the booming sector is the mining industry. The

lagging sector includes other tradables in manufacturing and agriculture. The non-

tradable sector includes services, utilities, transportation and so forth (Roemer 1985:237).

The booming and lagging sector face given world prices and there is only one mobile

factor, labour (Norberg and Blomström 1993:163). In the case of a boom in a natural

resource, resources might be drawn from sectors that are more conductive to long-term

growth to the booming sector. If so, the total effects on the economy’s possibilities to

grow in the long run could be negative.

2 . 1 . 1 T h e s y m p t o m s o f t h e D u t c h d i s e a s e

One of the assumptions of the Dutch disease model is full employment, and according to

the theory the booming sector attracts labour out of the other sectors, forcing wages to

increase in the whole economy. Increasing wages lead to increasing prices for domestic

6

goods and the lagging sector experiences lower returns and output, caused by a loss of

competitiveness. There are two effects in the Dutch Disease process, the resource

movement effect and the spending effect. The first effect occurs as a result of increasing

demand of labour in the booming sector, which draws labour from the other sectors. The

spending of the extra income generated by the booming sector (possibly through

government spending based on taxes levied on the booming sector) increases the demand

for domestic resources. The non-tradables sector, which is not facing international

competition and a given world price, will raise prices to maintain profitability. The

tradables sector, however, cannot raise prices, as they are operating on international

markets, and profitability declines in spite of higher factor costs, which leads to

movement of resources from the tradables to the non-tradables sector. The real exchange

rate appreciates, as a consequence of the increased price ratio between non-tradables and

tradables. This mechanism is called the Balassa-Samuelsson effect. Therefore, the

tradables sector, which becomes less competitive, contracts. This has been labeled de-

industrialization in economic literature. It is important, however, to be aware that neither

the resource movement effect nor the spending effect need occur if a country has

extensive underemployment (Lewis Jr, 1989:1562-63).

Changing the structure of the economy, the Dutch disease itself does not per definition

have adverse long-term effects on a country’s possibility to grow economically, since

growth in the booming sector can more than offset the stagnation in the lagging sector.

The long-term impact of a resource boom on an economy depends on the nature of the

booming sector and the lagging sector (Roemer 1985:245). The Dutch disease can be

harmful to long-term growth if the lagging sector is more conductive to growth than is the

booming sector. In the Dutch disease model, the lagging sector is usually the

manufacturing sector, a sector generally considered to be growth-generating, with the

positive externalities and increasing returns to scale associated to it and its linkages with

the rest of the economy. This is why economists often refer to the manufacturing industry

as the main “engine of growth” (Bulte et al. 2004:5). Mineral industries, such as the one

in Botswana, generally have weak linkages to the rest of the economy, thus limiting the

spill-over effects from that sector to the rest of the economy: “Natural resource abundant

7

economies benefit less from the technology spill-overs that are typical in manufacturing

industries...” (Papyrakis and Gerlagh 2004:182)

Even more problematic is the case where the lagging sector is the one that creates job

opportunities. According to the traditional Dutch disease model, presented above, one

key assumption is full employment. In that case, the more labour intensive the booming

sector is, the bigger are the resource movement and spending effects, and the bigger is the

threat of negative Dutch disease symptoms. However, this assumption of full

employment is rarely fulfilled, especially not in developing countries, and the threat of an

over-heated labour market is smaller. In the case of large unemployment, it is when the

booming sector is not labour-intensive and has the above mentioned weak linkages with

the rest of the economy, hence not creating employment opportunities, that the disease

could be detrimental to long-term growth (Roemer 1985:245, Bulte et al. 2004:5).

2 . 1 . 2 A v o i d i n g t h e D u t c h d i s e a s e

There are a number of different options to prevent symptoms of the Dutch disease to arise

and to protect the lagging industry. Reducing Dutch disease effects is a question of long-

term view and resisting the temptation of overspending, which has been described as

“swallowing bitter medicine” (Roemer 1985). One policy option is to protect the real

exchange rate from appreciating, in order not to harm the competitiveness of exporting

industries. To avoid or reduce a real appreciation is called exchange rate protection.

Central banks in natural resource booming countries therefore play an important role in

carrying out monetary policies that keep the exchange rate from appreciating too much.

One instrument for exchange rate protection is to try to neutralize the large revenues from

the booming sector by keeping the revenues out of the country. In other words, the

government should accumulate reserves of foreign currency and prevent the reserves

from becoming monetized in the domestic economy (Roemer 1985:247). Another

measure to ensure long-term macroeconomic stability is for governments dependent on

resource revenues to run budget surpluses. This reduces inflationary pressures in the

8

economy and provides a “cushion” if world market price for the natural resource would

drop.

One obvious problem with revenue sterilization and conservative fiscal policies is that it

can be tempting for a government to increase expenditures and win votes in elections

when export revenues are practically flowing into the country, a problem known as the

“time inconsistency problem”. When a country, as does Botswana, has humanitarian

problems like poverty, unevenly distributed incomes and high prevalence rates of

HIV/AIDS it becomes even more difficult not to increase expenditures. Governments

with politicians that are trying to accumulate reserves of foreign currency, instead of

dealing with humanitarian problems, could also be risking their jobs and the political

stability in the country (Roemer 1985:248).

2.2 Institutional Quality, Corruption and Neglect of Education

Another transmission mechanism by which natural resources have been found to affect

economic growth negatively is by inducing poor governmental policies. In cross-country

studies, an abundance of natural resource is typically associated with poor institutional

quality, with relatively high corruption and with poor educational performance.

Before discussing the impact of natural resources on institutional quality it is important to

note that a distinction between two types of resources has emerged in recent literature on

the resource curse: point resources and diffuse resources. Point resources are resources

with high geographical concentration that can be protected and controlled at a relatively

modest cost, for example oil fields or diamond mines. This type of resources has been

found to be easily captured by elites, to lead to rent seeking and corruption and

consequently to induce bad institutional quality. This is widely believed to ruin the

possibilities of economic growth, since “…malfunctioning government institutions

severely harm economic performance through a reduction in both incentives and

opportunities to invest and innovate” (Leite and Weidman 1999:3). Diffuse resources, on

the other hand, are more spread out geographically and the ownership of them tends to be

9

shared by many. This type of natural resources therefore results in better institutions and

generally also economic policies that promote economic growth (Bulte et al. 2004:14).

Related to the question of institutional quality are the ones of rent seeking and corruption.

As described above, there is a risk that political elites capture the control over natural

resources in an economy with an abundance of point resources. Once these rich resources

are in the hand of an influential group of people there is imminent risk of rent seeking

behaviour1. When people start to invest time and money in trying to capture parts of the

“exogenous” wealth created by natural resources instead of taking active part in

economic production, the economy is unlikely to grow (Isham et al. 2003:6). Tightly

linked to rent-seeking is the problem of corruption. Leite and Weidman (1999:23) show

that an abundance of point resources significantly increases the level of corruption in a

country. Their study also confirms what many scholars have found, namely that

corruption is detrimental to economic growth. They therefore conclude by claiming that

data supports the “…hypothesis of the corruption channel being an important explanation

for the slow growth of resource-rich economies” (Leite and Weidman 1999:31).

Another problem associated with an abundance of natural resource that has been

identified is that a country which gets heavily dependent on a resource can experience

problems with crowding out of human capital because they are confident that their natural

resource is the most important asset. Natural resource abundance may therefore lead to

insufficient attention to education. Gylfason (2001:851-4) argues that it is not the natural

resource itself that seems to be the problem, but the failure of public authorities to deal

with problems that come with the findings of a natural resource, namely by neglecting the

importance of education. In 1997, for example, OPEC countries spent less than four

percent of their GDP on education, compared to almost five percent for the rest of the

world. Gylfason shows that public expenditure on education, expected years of schooling

for males and females and secondary-school enrollment for both genders all show a

statistically significant negative relationship with natural resource abundance.

1 For a theoretical discussion of the relationship between natural resource abundance and rent-seeking, refer to Torvik 2002.

10

It is well established that education stimulates economic growth through several channels

of transmission. Education increases the efficiency of the labour force, it creates better

conditions for health care, it promotes good governance and institutional stability, all of

which have a positive effect on growth (Gylfason 2001:851). Barro (2001:14) analyzes

the impact of education on growth rates, and shows that one extra year of male schooling

at secondary or higher level increases the yearly growth rate of GDP by 0.44 percent.

Female education also raises the real growth rate of GDP by lowering the fertility rate.

His results also show that both the quality and the quantity of education significantly

increase a country’s growth rate, with the quality of education having the largest effect of

the two.

It should be pointed out that it is difficult to find a sound theoretical framework that

explains why some countries invest more in education than others. However, one

proposed explanation of the negative relationship between educational spending and

natural resource abundance can be derived from the Dutch Disease theory. According to

this a natural resource boom often leads to a contraction of the manufacturing sector.

Since human capital is an important production factor in the manufacturing sector,

Gylfasson (2001:856) argues that the need for high-quality education declines when the

sector contracts. Additionally, the manufacturing sector is often considered to be an

important source of technological progress, and as a consequence there are educational

externalities associated with it. If the manufacturing sector contracts it is most likely to

have a negative impact on externalities like knowledge and skills. If the contracting

sector is not the manufacturing sector, but for example the agricultural sector, the impact

on human capital investment will not be the same. Since the demand of well educated

labour is low in the agricultural sector, a contraction of this sector does not necessarily

lead to decreased spending on education. However, also the booming sector must be

considered when analyzing the overall impact of a resource boom on human capital

investment. If a strongly expanding capital intensive mining sector does not need a lot of

educated labour, then that could still decrease incentives to invest in education. This can

make it more difficult for future expansions in sectors abundant in skilled labour, which

11

in turn can lead to slow technological progress (Papyrakis and Gerlagh 2004:189; Sarraf

and Jiwanji 2001:5).

Papyrakis and Gerlagh (2004:189) empirically investigate the relative importance of

different proposed resource curse transmission mechanisms and find that the neglect of

education is an important one. They estimate, controlling for other possible resource

curse mechanisms' effects on growth, that the relative importance of educational neglect

of the total resource curse effect is about ten percent. They claim the importance of

neglect of education to be twice as important as the so called "corruption channel".

12

3. How Botswana Beat the Resource Curse Botswana’s development level at independence was poor by any standards. Besides cattle

production, which was Botswana’s traditional and largest export sector during the time of

independence, the country had a very shallow industrial base (Harvey and Lewis

1990:30). During decades after independence, Botswana experienced very high growth

rates. The far most important reason for that is the finding and successful exploitation of

diamonds in the beginning of the 1970s. The increase in production and exports of

diamonds have given Botswana unique economic conditions and largely contributed to

the major boost to government revenues. In the mid 1990s, between 75 to 80 percent of

government revenues stemmed from the diamond industry in a combination of royalty

payments, profits tax and dividends from the 50 percent partnership in the mining

company Debswana. The other 50 percent of the company is owned by the South African

company De Beer (Hope 1996:56).

Botswana’s economy has gone through a big transformation since independence. This is

well shown by the fact that the mining industry’s share of GDP increased from two

percent to 40 percent from 1966 to 1991. During the same period agriculture’s share of

GDP decreased from 41 percent to only five percent (Hope 1996:54). In 2002,

approximately 30 percent of the world’s supply of diamonds originated from Botswana,

and although the mining industry only employs four percent of the labour force, the

sector accounted for about 85 percent of total exports, 35 percent of GDP and 50 percent

of government revenues (Bank of Botswana 2002, CSO 1998). Considering the resource

curse theory introduced above, this great dependency upon a natural resource could have

been harmful for the country’s economy, but chart 1 tells a different story: Botswana

managed to do what many countries have not succeeded in; turning natural resources into

sustained economic growth rates.

13

Chart 1. GDP & GDP per capita in constant 1993/94 prices

0

5000

10000

15000

20000

25000

19741976

19781980

19821984

19861988

19901992

19941996

19982000

2002

Pula

/ M

illio

n Pu

la

BNP tot. Million Pula GDP per capita in Pula

Source: World Development Indicators 2005

As shown by the GDP evolution since the diamond boom in chart 1, Botswana’s mining

industry must be considered to have been a blessing for the economy. From a very low

level in the 1970s Botswana has 30 years later reached a relatively high level with a total

GDP of almost 20.000 million Pula and a GDP per capita of more than 10.000 Pula. For

the sake of international comparison, Botswana had a PPP-adjusted GDP per capita of

8714 US dollar in 2003 (World Development Indicators 2005).

Observing the impressive evolution of growth rates we can therefore conclude that the

country managed to generate sustainable growth and beat the resource curse and we will

now assess the important question of how this was done. To evaluate the importance of

the three resource curse mechanisms mentioned earlier, we will now look at the effect of

each one of those on Botswana’s economy. To evaluate the eventual Dutch disease

symptoms experienced by the country, we will look at growth in different sectors of the

economy, unemployment rates, inflation, fiscal status from year to year, the real

exchange rate and the foreign exchange reserves. In order to examine to what extent the

14

diamond boom affected the institutional quality in Botswana, we turn to look at

corruption and democracy indices and finally, to determine what the educational

performance of Botswana has been like, we will analyse public spending on education

and enrollment rates at different levels.

3.1 The Dutch Disease in Botswana

To assess the question whether Botswana suffered from the Dutch disease is quite

difficult. Totally eliminating Dutch disease effects associated with an export boom of the

magnitude Botswana has experienced seems impossible. Hence, the question of Dutch

disease symptoms in Botswana is not a yes or no question, but rather a question of the

severity of the disease. Botswana’s first diamond mines opened in the early 1970’s but it

was not until the beginning of 1980’s that they experienced a big boom. As the diamond

boom in Botswana is approaching a long term perspective it is important to keep in mind

that the following analysis is based upon medium run effects of the Dutch disease. In the

long run the three sector analysis can give quite different results. However, the line

between medium run and long run is far from clear cut in economics, and we argue that a

medium run analysis is still appropriate for the period studied in this paper.

3 . 1 . 1 T o w h a t e x t e n t d i d B o t s w a n a s u f f e r f r o m t h e D u t c h d i s e a s e ?

According to the classical Dutch disease theory the movement of labour force to the

booming sector from the other sectors raises the general level of wages in the country. In

the case of Botswana, it seems like there has not been any extensive labour movement.

Much of the need for labour force movements between the sectors appears to have been

eliminated since the mining industry is very capital intense and has only employed about

four to five percent of the labour force. The demand for skilled labour in the mining

industry has also contributed to the reduced labour force movement. Inherent

unemployment and demand for skilled labour has made it possible only for a small part of

the labour force to reach high wages, while the bad negotiation situation for unskilled

workers in the agricultural sector remains. The fact that the government workforce, which

15

accounts for a big part of the labour force, has been under regulated income policies has

also played an important role in preventing substantial salary increases (Jefferis 2005).

The combination of these factors may account for the modest overall wage increases

which consequently have alleviated the resource movement and spending effect during

the booming period. The spending effect might also have been mitigated because of the

fact that a large part of the consumption stems from import, and therefore reduces the

increase in prices in the non-tradable sector (Norberg and Blomström 1993:175-6).

However, Mogotsi (2002:145) claims that there has been an increase in both household

and government spending from the mid 1980s to the big recession in the beginning of the

1990s, while Jefferis (2005) argues that the spending effect was delayed and did not arise

until the late 1990s.

In countries suffering from the Dutch disease, the manufacturing sector is usually the

contracting sector, but in the case of Botswana it would be more correct to talk about a

de-agriculturalisation, since the agricultural sector’s share of the economy has declined

far more than the manufacturing sector’s. Also, employment actually rose in the

manufacturing sector, whereas it has decreased in the agricultural sector. Table 1 shows

growth in different sectors in absolute value and as share of total GDP2.

2 The table is based upon statistics made available by the Bank of Botswana and collected by Keith Jefferis. The UN World Development Indicators have similar statistics, but since they do not isolate the diamond insustry, which is important in this case, we use Keith Jefferis’s data. The data can be found in Appendix A.

16

Table 1. Sectoral growth (Million Pula & percent)

Year

Agric

value

Agric

share

Mining

value

Mining

share

Manuf

value

Manuf

share

Sevice

sector

Value

Sevice

sector

share

1974/75 423 24.1 223 12.7 120 6.8 495 28.1

1976/77 451 20.8 383 17.6 169 7.8 644 29.7

1978/79 421 14.8 707 24.8 209 7.3 794 27.8

1980/81 382 10.6 1,319 36.8 164 4.6 800 22.3

1982/83 340 7.6 2,291 51.0 206 4.6 898 20.0

1984/85 286 5.4 2,686 50.7 172 3.2 1,215 22.9

1986/87 296 4.8 3,025 48.8 251 4.1 1,563 25.2

1988/89 451 5.1 3,779 43.0 440 5.0 2,448 27.8

1990/91 478 4.8 3,943 39.4 469 4.7 2,818 28.2

1992/93 489 4.6 3,767 35.5 500 4.7 3,214 30.3

1994/95 459 4.0 3,899 34.2 532 4.7 4,080 35.8

1996/97 453 3.6 4,311 33.9 594 4.7 4,736 37.3

1998/99 443 3.1 4,588 32.1 661 4.6 5,471 38.3

2000/01 445 2.7 6,046 36.6 681 4.1 6,135 37.2

2001/02 433 2.6 5,865 34.7 682 4.0 6,623 39.2

2002/03 441 2.4 6,472 35.5 703 3.9 6,839 37.5

2003/04 446 2.3 6,918 35.9 700 3.6 7,130 37.1

Source: Jefferis 2005

Note: Due to lack of space we have decided not to include small and for our purpose less

relevant sectors, and the share of total GDP will therefore not sum up to 100 percent.

Botswana’s traditional export sector, the agricultural sector, has more or less stayed on

the same GDP level in absolute numbers as 30 years ago, but its share of total GDP has

contracted from 24.1 to 2.3 percent even though numerous government programmes have

been implemented to stimulate productivity and to boost incomes (BIDPA 2004).

Another verification of the de-agriculturalisation is that the agricultural sector’s

percentage of the total labour force has declined from 70 percent in 1980 to 42 percent in

1996 (African Development Report 2004:327). The manufacturing sector has had a large

relative contraction from 6.8 to 3.6 percent of total GDP 1974 -2004, but its output has

increased in absolute numbers. However, the employment in this sector grew even more,

which implies that the productivity in it has decreased (Norberg and Blomström

17

1993:171). The interpretation of the absolute values for the manufacturing sector isolated

would suggest that Botswana has not suffered from the Dutch disease, while an

interpretation of the relative values clearly suggests that they have. If one looks at the

relative figures for the agricultural sector there has been a large contraction. Whether this

is a Dutch disease symptom or not is hard to say, and will be discussed later.

Both the manufacturing and the service sector experienced a sharp decline in relative

terms in the beginning of the 1980s, and while the service sector did recover remarkably

it seems like the manufacturing sector never managed to pick up. Even though the mining

sector has grown substantially over a long period of time, the growth in the industry has

been more variable the last decade and is now showing signs of slowing down. The

service sector has shown significant increase and has been the fastest growing sector in

Botswana since the booming period. The service sector includes the growing government

sector, which of course to a large extent has been financed with diamond money.

To sum up, one could say that Botswana has not experienced the most severe symptoms

of the Dutch disease, but nevertheless, the growth in the diamond industry may have

caused other sectors to lag. However, the decline in relative terms of the manufacturing

sector, which is usually the explanation to the harmful effects of a resource boom, was

modest, whereas the low-productive and low-competitive agricultural sector has

experienced a significant contraction in relative terms. Hence, one could argue that the

de-agriculturalisation in Botswana shows that a “mild version” of the disease did exist.

3 . 1 . 2 W h a t d i d B o t s w a n a d o t o p r e v e n t D u t c h d i s e a s e e f f e c t s ?

One important symptom of the Dutch disease is the appreciation of the real exchange

rate, which causes non-booming sectors producing tradable goods to lag. Thus, real

exchange rate protection is essential in a country experiencing a large boom in exports. In

Botswana, the central bank has constantly strived to keep an appropriate exchange rate.

Its monetary policy seems to have been guided by the trade off between increasing the

value of the Pula, in order to keep the inflationary pressure down, and decreasing its

18

value in attempts to boost the export industry. During the 1980s, for example, when the

pula was pegged to a trade-weighted basket of currencies, the central bank tried very

actively to navigate towards these objectives. Between 1982 and 1985 the pula was

devaluated three times to boost exports, in 1985 and 1989 it was revaluated to decrease

the inflationary pressure and in 1990 it was devaluated by ten percent in order to offset a

decline in export earnings (Norberg and Blomström 1993:167). During our field study in

Botswana for this paper, in June 2005, the Bank of Botswana devaluated the Pula by 12

percent in an attempt to boost exports.

Since the early 1970s, the evolution of inflation has been much like a roller coaster ride,

with annual average rates peaking at 18.1 percent in 1974, to more modest figures below

ten percent during the last couple of years. As can be seen in Chart 2, the annual average

rate of inflation has never been under six percent since 1972. However in a developing

country perspective Botswana’s inflation rates have not been extremely high. With a

fixed exchange rate, a high inflation leads to a real appreciation of the domestic currency.

In the case of Botswana, the inflation rates do not seem to have caused a real appreciation

of the Pula during the booming period, as we will show in Chart 4.

Chart 2. Annual rate of inflation

0

2

4

6

8

10

12

14

16

18

20

19711973

19751977

19791981

19831985

19871989

19911993

19951997

19992001

2003

%

Source: Statistical Bulletin June 1999 & Financial Statistics Jan 2005, Bank of Botswana

19

As described earlier, protecting the real exchange rate from appreciating is essential to

avoiding Dutch disease symptoms. With active monetary policies, the central bank can

help protecting the real exchange rate by keeping the large export revenues out of the

country, thereby accumulating large foreign exchange reserves.

To examine whether Botswana has managed to keep the large diamond revenues out of

the country, the development of the foreign exchange reserves over a longer period of

time is of interest. A large foreign exchange reserve could be interpreted as if a lot of the

revenues have not been monetized domestically, whereas a small reserve could be

interpreted as if large parts of the revenues have been monetized in the country. It is also

of interest to see how the size of the reserve has varied over time. Large variations can be

due to an inconsistent policy of exchange rate protection.

Even though the findings of diamonds in Botswana occurred in the early 1970s, it was

not until the 1980s that Botswana experienced the major boom. Thanks to the great

increase in diamond exports, the trade balance of Botswana has shown a surplus since the

middle of the 1980s, which has made it possible to build up an impressive foreign

exchange reserve. Chart 3 shows the evolution of the foreign exchange reserve from

1982-20033.

3 Bank of Botswana has only published the evolution of the foreign exchange reserves since 1982.

20

Chart 3. Foreign exchange reserves

0

1000

2000

3000

4000

5000

6000

7000

19821983

19841985

19861987

19881989

19901991

19921993

19941995

19961997

19981999

20002001

20022003

US

Mill

ion

Dol

lar

Source: Bank of Botswana, Annual Reports

Chart 3 clearly shows that the long-run growth of the reserves has been tremendous. The

increase over just one year (1986-87) was almost 70 percent. After that the growth

successively decreased, but the annual percentage growth from the 1990s until 2000 has

steadily averaged between five and ten percent. In 2001, the foreign exchange reserves

were as large as 6317 billion US dollar, which was sufficient to cover the imports of

Botswana for 36 months (Bank of Botswana 2001).

The question of whether Botswana’s attempt to protect the real exchange rate from

appreciating succeeded has no definite answer. While the term “real exchange rate” is

most commonly used in literature, the accurate way of evaluating exchange rate changes

is in terms of the real effective exchange rate which takes into account the relative

21

importance of different trading partners4. Due to changing exchange rate regimes and

trade weights, it is impossible to obtain good data on the real effective exchange rate

from official sources. However, economist Keith Jefferis gave us access to his personally

calculated real effective exchange rate and it is his calculations we base Chart 4 upon5.

Chart 4. Real Effective Exchange Rate 1976-2003

60

70

80

90

100

110

120

19761978

19801982

19841986

19881990

19921994

19961998

20002002

Inde

x, 1

976=

100

Source: Jefferis 2005. Note that the graph shows the price of foreign currencies in terms of Pula. Hence, a decrease in

the index describes a depreciation of the Pula.

As can be seen in chart 4, Botswana managed well to protect the real effective exchange

rate from appreciating. The REER was more or less hovering around the same level until

1984 when it depreciated sharply for two years. Thereafter it did not show a lot of

variance until 1998 when the pula instead started to appreciate, and has done so since.

Given that the major boom took place in the 1980s, an appreciation rather than

4 The Nominal Effective Exchange Rate (NEER) is a weighted average of the most important bilateral nominal exchange rates, with weights based on the trade shares taking into account the relative importance of each currency in the effective exchange rate basket. The Real Effective Exchange Rate (REER) is obtained by adjusting the NEER for inflation differentials with the countries whose currencies are included in the basket. As the inflation rate in each country is assumed to broadly indicate the trends in domestic costs of production, the REER could be seen as a reflection of the foreign competitiveness of domestic products. 5 The file containing all data on which the REER is based is very large and has therefore not been included in an appendix. However, upon request the data can be obtained from the authors.

22

depreciation is expected during this time. Thus, the evolution of the REER has not been

according to what the theory predicts. Instead the appreciation occurs much later than

what is expected. Although Norberg and Blomström (1993:176-7) hold the exchange rate

protection to have been more successful than Mogotsi (2002:153-4) does, everyone

seems to agree the country managed to avoid the worst exchange rate symptoms of the

Dutch disease. This was made possible by government regulated incomes, through active

monetary policies by the central bank and by resisting the temptation of spending

diamond revenues domestically.

Another proof of Botswana’s determination to ensure long-term economic stability is

seen in the government budgets during the years of the boom. From the fiscal year6 1983-

84 until 1997-98 the government budget was in surplus every year (Bank of Botswana,

various issues). This clearly shows that the government did not let its expenditures get out

of hand despite constant large revenues from the mining industry. However, during more

recent years, in accordance with what was also seen in the case of the foreign exchange

reserves, some government budgets have shown quite substantial deficits. The 2002-03

budget very well shows that Botswana’s government still is heavily dependent on

diamond revenues: The budgeted deficit for the year was 1619 million Pula but the

revised budget showed a deficit of 2216 million Pula, which was entirely due to the fact

that weakness in the world diamond market and the Pula appreciation against the dollar

caused a 17 percent drop in mineral revenues during the year (Bank of Botswana

2002:51).

3.2 Institutional Quality, Corruption and Neglect of Education

If there is one transmission mechanism of the resource curse that can easily be rejected in

Botswana’s case, it is the one by which natural resource abundance leads to poor

institutional quality. When it comes to democracy, good governance and low corruption,

Botswana is being used as a role model for African developing countries.

6 In Botswana, the fiscal year runs from April to March

23

Since independence in 1966, Botswana has without interruption been governed by BDP

(Botswana Democratic Party). The party has pursued free market economic policies, with

some state interventions. Despite the fact that only one party has ruled in the country for

almost 40 years, Botswana is considered a well-functioning democracy with a

“…professional bureaucracy that has conducted and implemented policy-making

efficiently” (Taylor 2004:154).

As we described earlier, research has shown a relationship between abundance of point

natural resources and poor institutions, which in turns has been shown to have a negative

impact on growth. In the case of Botswana, the abundant natural resource is clearly a so

called point resource. The diamond deposits are highly concentrated geographically and

relatively easily controlled. Thus, the type of resource that Botswana is dependent upon is

of the kind that has had tendency to result in poor institutional quality and consequently

poor economic performance. However, the country seems to have escaped the pitfall of

developing a corrupt economic system despite the abundance of natural resources.

The most commonly used measure of corruption is Transparency International’s

corruption perceptions index (CPI). The surveys used for the index are based on data

from the last three years, and measures the degree of corruption in public and political

sectors. The scores relate to perceptions of the degree of corruption as seen by business

people and country analysts and ranges between 10 (highly clean) and 0 (highly corrupt).

Transparency International’s corruption perceptions index gives Botswana a rank of 31 in

the world, with a score of 6.0 (Transparency International 2005). This gives Botswana a

position among the best of all developing countries and also well above many richer

countries. For example, the country ranks higher than a couple of members of the

European Union, such as Italy and Greece. We therefore conclude that corruption is not a

severe problem in Botswana, despite the abundance of natural resources.

To assess the question of the institutional quality in Botswana, and to get a general

picture of institutions in the country, we will use six different indicators constructed by

the World Bank. Aware that some question the reliability and validity of such indices, we

24

have chosen those that seem appropriate for our analysis. These six different indicators of

institutional quality are Voice and accountability, Political stability, Government

effectiveness, Regulatory quality, Rule of law, Control of corruption and capture different

dimensions of institutional quality. Voice and Accountability is an indicator of the level

of political, civil, and human rights in the country. Political Stability measures the

likelihood of violent threats to, or violent changes in, government. Government

Effectiveness is an indicator of the quality of the bureaucracy and the delivery of public

services. Regulatory Quality measures the incidence of market-unfriendly policies. Rule

of Law measures the quality of contract enforcement, the police, and the courts and is

also an indicator of the incidence of crime and violence. Finally, Control of Corruption is

a measure of how well different forms of corruption, such as for example exercise of

power for private gain, are controlled (Kaufmann et al. 2005:4).

Table 2. Institutional quality

Indicator Botswana’s Percentile

Rank

Regional Average Income Category

Average

Voice and

accountability 68.9 32.7 64.6

Political stability 69.4 32.8 67.6

Government

effectiveness 76.9 27.6 62.2

Regulatory quality 79.8 29.5 63.0

Rule of law 70.5 27.6 64.4

Control of corruption 80.8 30.1 63.4

Source: Kaufman et al., 2005

Table 2 shows the level of Botswana’s institutional quality according to the six above

described indicators. The percentile rank shows the percentage of countries in the world

that have a lower rating than the country indicated. For all the indicators a higher value

implies a better governance rating. Table 2 also shows the average for countries in

Botswana’s region (Sub-Saharan Africa) and for countries in the same income category

25

as Botswana (Upper Middle Income). Hence, the rating of Voice and Accountability

should be interpreted as follows: 68.9 percent of all countries have a worse rating of

Voice and Accountability than does Botswana. The average of all Sub Saharan countries,

however, rates better than only 32.7 percent of all countries, whereas the average of

countries in the upper middle income category has a better rating than 64.6 of all

countries.

Table 2 as a whole can only be interpreted as a confirmation of the high institutional

quality in Botswana. Botswana rates way better than the Sub-Saharan African average in

all six governance indicators. Sub-Saharan Africa is a region that since long has been

raged by war and poor development levels, which is why low levels of democracy and

poor economic and political institutions comes as no surprise. More impressive is the fact

that Botswana in all indicators is rated to have a better institutional quality than the

average of all countries categorized as “upper middle income”.

As described earlier, countries abundant in natural resources also tend to neglect the

importance of education to long-term growth. In the case of Botswana, though, no traces

seem to found of such a development. Soon after independence, before the diamond

findings, the country invested heavily in education. The government first stressed the

importance of secondary education, but later also invested in primary education.

Inequalities in access to education were successively reduced by building schools in more

remote areas of the country. By the beginning of the 1980s, when the country was starting

to enjoy the virtues of the diamond industry, the successful increase in supply of primary

education further increased the demand for secondary education, which was responded to

by the government (Harvey and Lewis 1990:285-287). The focus on education has

continued since and has resulted in impressive improvements in the educational record of

Botswana.

The most common way to evaluate how much a government values education is to look

at the percentage of GDP spent on education. While the amount of public spending on

education does not say really anything about the outcome of the spent money, the

26

measure is the commonly used in economic literature and does give an idea of how much

priority is given by the government to education. As table 3 shows, for most of the period

1960-1999, Botswana has spent an increasing share of GDP on education. As mentioned

earlier, the world as a whole spends on average around five percent of GDP on education.

Compared to the world average of five percent, the 9.21 percent of GDP spent on

education in Botswana in 1996 is definitely very high.

Table 3. Public spending on education in Botswana, 1965-1999 (% of GDP)

1965 1970 1975 1980 1981 1982 1983 1984 1985 1986 1987 4,38 4,11 6,06 5,68 5,58 5,60 5,60 5,69 5,22 6,17 6,19

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1999

5,52 5,78 6,20 6,94 6,98 7,58 7,40 8,09 9,21 7,93 8,60

Source: World Development Indicators 2005

Concerning the level of education in countries, one has to distinguish between quantity of

education and quality of education. Measuring the quality of the education in Botswana

would be very difficult, since we do not know either of any international comparisons of

the level of Botswana’s educational system, or of any internationally standardized tests

used in schools in Botswana. However, looking at enrollment rates at different levels of

education we can get a good idea of the quantity of education in Botswana during the last

decades. As mentioned earlier, Barro (2001:16) has shown that the quantity of education

has a significantly positive effect on growth, even though the effect of educational quality

is even more important.

Important measures of educational participation are the enrollment rates on primary,

secondary and tertiary level. As could be expected, given the status of education in

Botswana at independence and the big attention given to improving it by the government,

table 4 shows that enrollment rates in Botswana have risen significantly in both primary,

secondary and tertiary education.

27

Table 4. Enrollment rates over the last decades

Year 1970 1975 1980 1985 1990 1995 2000

School enrollment, primary (% net) 46,0 57,2 75,6 89,4 93,3 81,3 79,6

School enrollment, secondary (% net) .. 10,8 14,4 22,6 33,5 44,5 54,6

School enrollment, tertiary (% gross) .. 0,7 1,2 1,8 3,2 5,3 4,6

Source: World Development Indicators 2004

28

4. Explaining the success of Botswana Our empirical findings give a rather straight-forward answer to our research question why

Botswana did not suffer from the resource curse. The country pretty much avoided all the

important transmission mechanisms that have been shown to lead, from an abundance of

natural resources, to poor economic growth. The case is not clear-cut on the question of

the extent of Dutch disease symptoms in Botswana. However, the two important

transmission mechanisms of poor institutional quality and neglect of education cannot be

identified at all in the case of Botswana.

4.1 Avoiding the Dutch Disease

One of the harmful effects is when growth generating industries, such as manufacturing,

are crowded out. Botswana has successfully managed to avoid de-industrialisation

because of thorough strategies for the economic policy, income regulations and since

supplementary labour from the labour pool easily could fill vacancies, hence limiting the

resource movement effect and the rise in overall wages. Also, the agricultural sector,

which has clearly taken the big hit in the transformation of the economy, was never a

very growth-generating industry. One can argue that symptoms of Dutch disease did

appear, but in a dissimilar form, a de-agriculturalisation. The contraction of the

agricultural sector might have been an effect of labour movements to an expanding

service sector that likely has been stimulated through the growth of and the increased

spending in the government sector. However, poor soils and difficult droughts are other

possible explanations to the contraction of the sector and even if Botswana would not

have suffered from any Dutch disease effects at all, the agricultural sector may still have

contracted due to the low productivity in the sector.

The fact that the manufacturing industry has performed quite well since the diamond

boom is not compatible with the Dutch disease model. This suggests that the industry

may have been fairly competitive internationally even though it, as a share of total GDP,

contracted domestically. The decline in productivity in the manufacturing sector can be

29

due to skilled labour being absorbed by a growing service sector. Since the

manufacturing industry has been rather small, both in absolute and relative figures, one

could argue it is unlikely that this sector has played a major part in the strong overall

growth in Botswana. However, the growth in absolute numbers of the manufacturing

sector might have had some implications for the economic development. For many years

the government in Botswana supported the manufacturing industry heavily. For example,

attempts at diversifying the economy away from diamond dependency were made by

trying to put a large and competitive clothing industry in place. These attempts have

stopped in recent years since the industry did not show signs of being able to compete

internationally (Jefferis, 2005). Still, it is possible that the support of the manufacturing

industry had other positive effects on the economy. As described earlier, the

manufacturing sector is generally considered as an engine of growth and a sector with

plenty of linkages to the rest of the economy. Hence, in supporting the manufacturing

industry, Botswana might have weakened the de-industrialization that can be so

detrimental to long-term growth.

A growing service sector is compatible with the Dutch disease theory. The difficulty in

importing services (compared to consumer goods) might be a part of the reason for the

impressive growth in the domestic service sector. However, a growing service sector is

often observed in growing economies. It is therefore difficult to determine whether the

growth in the service sector in Botswana is due to Dutch disease effects or simply an

effect of the modernization of the economy.

It is not easy to explain why Botswana suffered less from the Dutch disease than did

many other mineral exporting countries. The fact that the ruling party, the BDP, has had a

big support and been in office for almost 40 straight years has likely helped the

government to implement and maintain persistent economic policies, like the exchange

rate protection and income regulations. Income policies might have given the sensitive

tradable sectors some protection in order to adjust more smoothly to the rapid change of

the economy. The foreign exchange reserves have had a steady growth during a long

period of time, and it seems as though Botswana has resisted the temptation of

30

monetizing a lot of the mining revenues in the country. This shows the country’s

commitment to exchange rate protection and also to economic stability in order to avoid

de-industrialization.

It is justified to ask the question whether Dutch disease really is a disease, or if the

symptoms are part of a normal economic adjustment process that is unavoidable, like a

shift to a new equilibrium? Whereas it is clear that Botswana has not had any severe

symptoms of the Dutch disease, the transformation of the economy did nevertheless cause

other sectors to lag. However, the decline in relative terms of the manufacturing sector,

which is usually the explanation to the harmful effects of a resource boom, could have

been worse, and instead the low-productive and low-competitive agricultural sector’s

share of GDP decreased significantly. While the de-agriculturalisation in Botswana

indicates that a bout of the disease actually did exist, this form of the disease was

obviously not very serious.

4.2 Botswana’s High Institutional Quality

Different theories have been advanced to explain the good democratic institutions in

place in Botswana. According to Harvey and Lewis (1990:10), local tribes in Botswana

also have a long tradition of democratic procedures in decision-making. Before the

creation of modern institutions in Botswana, local tribes played an important role in

governing society. These tribes, kgotla, were democratically governed and the leaders of

the tribes listened to advice and were responsive to the tribe members’ opinions. This is

likely to have helped in creating a democratic environment at independence. Also,

Botswana has a long history of pragmatic negotiation with its powerful neighbors. This

might have given Botswana a habit of negotiating and compromising, which is valuable

in democratic societies.

Another reason could be that Botswana’s first governments largely consisted of former

cattlemen. Many of the post-colonial problems in Sub-Saharan Africa have been

associated with neglect of rural interests, due to lack of rural representation in central

31

politics. This was never the case in Botswana (Harvey and Lewis 1990:9-10). The fact

that Botswana leaders had a background in cattle farming, which was the main productive

sector at independence, created a situation where improved infrastructure and economic

institutions was not only in the country’s economic interest, but also in the self-interest of

the country’s leaders (Taylor 2004:155). This situation contrasts strongly to that of many

African countries during the last decades. Many of these have a long history of leaders

that have acted purely in self-interest, most often at the cost of the country and its

population. In Botswana, however, the self-interest of the leaders induced policies that

were also beneficial to the country as a whole.

In the introductory chapter we described how the British colonialists’ neglect of

Botswana contributed to the poor economic heritage of the country. The other side of the

coin is that the British’ neglect of Botswana paradoxically resulted in better political

institutions, which in turn has helped the economic development of the country. Since the

British Empire only had a limited interest in Botswana the colonial rule there was limited.

As a result, the impressive pre-colonial institutions essentially survived the colonial

period. Hence, in contrast to many other colonized countries, Botswana found itself with

decent political institutions at the time of independence (Acemoglu et al. 2001:20).

The former deputy chief of Bank of Botswana, Keith Jefferis (2005), points at the

importance of the order in which the institutions and diamonds were found by stating:

“Since the institutions arrived before the diamonds the institutions reflects poverty and

not richness, and when diamonds were found the surplus was saved rather than spend”.

Another way of putting it is that Botswana was “lucky” to discover diamonds after the

institutions were formed, instead of forming institutions after the diamonds were found.

Had the diamonds arrived before solid institutions were in place, there is imminent risk

that poor institutions had been put in place. This view is shared by Acemoglu et al.

(2001:24), who claim that “Botswana got off onto the right track at independence and by

the time the diamonds came on stream, the country had already started to build a

relatively democratic polity and efficient institutions. The surge of wealth likely

reinforced this.”

32

Another important feature of the timing of the diamond findings was that they came well

after the decolonization of Botswana. One can only speculate about what would have

happened would the diamonds have been discovered during the British rule that lasted

until 1966. It is easy to believe that diamond findings before 1966 would have increased

the British’ interest in Bechuanaland. First, a lot of the wealth emanating from diamond

revenues would definitely have been captured by the British Empire and therefore would

not have benefited to the Batswana. Second, in accordance with the discussion above, an

increased British interest and involvement in Bechuanaland would probably have had a

negative impact on the well-developed pre-colonial institutions in place in Botswana.

To sum up, we have shown that Botswana has used their diamond resources into long-

term economic growth. This development contrasts with the experience of many other

natural resource dependent countries, which we partly attribute to the country’s ability to

avoid severe Dutch disease effects. However, equally important factors to Botswana’s

impressive economic performance have been the well-functioning and stable institutions,

the relatively low corruption and the big educational improvements.

One issue we have not approached in this paper is the possible, and probable, interplay

between the transmissions channels we have analyzed. We mainly treat them separately,

although they are likely to affect each other in different ways. The high institutional

quality of Botswana, for example might have affected the way the country prevented

some of the Dutch disease symptoms to emerge. For instance, we mentioned the

democratic stability as a determining factor for the government’s long-term engagement

to macroeconomic stability. Nevertheless, we mainly discuss the mechanisms separately.

A more statistical approach to the issue, using a set of cross-country data over time,

would allow for further conclusions to be drawn on how these different resource curse

mechanisms interact.

33

5. Final discussion If there is one big problem to the future economic growth, development and poverty

reduction in Botswana, it is obviously the raging HIV/AIDS epidemic. According to

UNAIDS (2004:2), the prevalence rate of the population between the age of 15 and 49 is

37.3 percent, which is among the very highest rates in the world. While the HIV/AIDS

epidemic is unanimously acknowledged as very serious, there remain disagreements on

the impact of it on life expectancy in Botswana. Whereas the Botswana Central Statistics

Office’s 2001 population census estimated the life expectancy to around 55 years,

international estimates claim it to be below 40 years, compared to 60 years around a

decade ago.7

Another issue that puts in doubt the policy success of Botswana is the persistence of high

economic inequality and alarming poverty rates. According to the Central Statistics

Office’s Household Income and Expenditure Survey (CSO 2005), the already high GINI-

coefficient of 0.537 in 1993-94 had increased to 0.573 by 2002-03, which indicates that

Botswana is one of the most economically unequal countries in the world. This provides

strong evidence that the whole population has not benefited equally from the strong

economic growth. Somewhat more encouraging is the evolution of poverty, which has

decreased substantially during the past decades: In 1985/86, 57 percent of the population

lived in poverty. In 1993-94 the figure had gone down to 47 percent, and according to the

latest survey the poverty ratio was down to 30 percent in 2002/03 (CSO 2005). However,

the critical observer will claim that a poverty ratio of 30 percent in a country that had the

highest growth rates in the world for decades is still far from acceptable. One reason for

the high inequality and poverty in Botswana are the high unemployment rates which have

plagued the country for decades. According to the most recent estimates from Botswana,

the unemployment rate was at 23.8 percent in 2002/03 (CSO 2005).

7 For readers interested in the HIV/AIDS situation in Botswana, and its effects on the economy, we recommend The Macroeconomic Impact of the HIV/AIDS Epidemic in Botswana by Robert Greener, Keith Jefferis and Happy Siphambe in The Botswana Journal of Economics, March 2004, 49-65.

34

As is the case with many natural resources, diamonds cannot enrich Botswana

perpetually. According to a confidential report, quoted in the Sunday Standard (2005),

Botswana’s mining company Debswana will close down all its mines in 20 years time.

By then, the revenues from the diamond mines will have dropped from the current 12

billion pula to less than four billion, at which point mining will no longer be profitable.

The article concludes by stating that “…if the fears expressed by Debswana engineers

come true, Botswana will have to come up with world class industries in the next few

years to take over from diamonds and fill the gap” (Sunday Standard 2005). This paper

has shown that Botswana succeeded well with the task of adapting to a life with the

diamonds. Knowing, however, that diamonds do not last forever, the story has not yet

ended positively. Only the future will show if the country can also adapt to do without the

diamonds.

35

References African Development Report, 2004. African Development Bank, New York: Oxford

University Press Inc..

Acemoglu, Daron, Johnson, Simon and Robinson, James A., 2001. “An African Success

Story: Botswana”, CEPR Discussion Paper 3219.

Bank of Botswana, 2005. Financial Statistics, Jan 2005.

Bank of Botswana. Annual Report, various issues.

Barro, Robert J., 2001. “Human Capital and Growth”, The American Economic Review 91

(2), 12-17.

BIDPA 2004, Botswana Institute for Development Policy Analysis, Briefing, 1st quarter

2004.

Boschini, Anne D., Pettersson, Jan, and Roine, Jesper, 2003. ”Resource Curse or Not: A

Question of Appropriability”. Working paper 2003:12, Stockholm School of Economics.

Bulte, Erwin H., Damania, Richard and Deacon, Robert, 2004. “Resource Abundance,

Poverty and Development”. Working Paper 21’04, Department of Economics, University of

California, Santa Barbara.

Central Statistics Office, 2005. Household Income and Expenditure Survey 2002/03,

Gaborone: Government Printer.

Gylfason, Thorvaldur, 2001. “Natural resources, education, and economic development”.

European Economic Review 45, 847-859.

Harvey, C. and Lewis, S. 1990, Policy Choice and Development Performance in Botswana.

New York: St. Martin’s Press.

36

Hope Sr, Kempe Ronald, 2002. From crisis to renewal: Development policy and

management in Africa, Boston: Brill.

Hope Sr, Kempe Ronald, 1996, “Growth, Unemployment & Poverty in Botswana”, Journal

of Contemporary African Studies, 14 (1), 53-67.

Isham, Jonathan, Woolcock, Michael, Pritchett, Lant and Busby, Gwen, 2003. “The Varieties

of Resource Experience: How Natural Resource Export Structures Affect the Political

Economy of Economic Growth”, Middlebury College Economics Discussion Paper No. 03-

08

Jefferis, Keith, 2005. Personal Interview with Keith Jefferis, former deputy Governor of the

Bank of Botswana.

Kaufmann, Daniel, Kraay, Aart and Mastruzzi, Massimo, 2005. “Governance Matters IV:

Governance Indicators for 1996-2004”, World Bank. Available at:

http://info.worldbank.org/governance/kkz2004/

Leite, Carlos and Weidman, Jens, 1999. ”Does Mother Nature Corrupt? Natural Resources,

Corruption, and Economic Growth”. IMF Working Paper 99/85.

Lewis Jr, Stephen R, 1989, “Primary Exporting Countries”, 1541-1600 in Chenery, Hollis,

Srinivasan, T. N., (Eds.) Handbook of Development Economics, Vol 2, Amsterdam: Elsevier

Science Publishers B.V.

Mogotsi, Imogen, 2002, “Botswana’s Diamonds Boom: Was There a Dutch Disease?”. The

South African Journal of Economics 70 (1), 128-155.

Norberg, H. and Blomström, M., 1993. “Dutch Disease and Windfall Gains in Botswana”,

162-178, in Blomström, M. and Lundahl, M. (Eds.) Economic Crisis in Africa. Perspectives

on Policy Response. London: Routledge.

37

38

Papyrakis, Elissaios, Gerlagh, Reyer, 2004. “The Resource Curse Hypothesis and its

Transmission Mechanisms”, Journal of Comparative Economics 32, 181-193.

Roemer, Michael, 1985. “Dutch disease in developing countries: Swallowing bitter

medicine”, 234-252 in Lundahl, Mats (Ed.) The Primary Sector in Economic Development.

London: Croom Helm.

Sachs, J. D. and Warner, A. M., 1995. “Natural Resource Abundance and Economic

Growth”. NBER Working Paper 5398.

Sachs, J. D. and Warner, A. M., 1999. “The Big Push, Natural Resource Booms and

Growth”. Journal of Development Economics 59, 43-76.

Saleshando, Dumelang, 2005. “We Are Not Victims Of Resource Curse”, The Monitor

15/5/05

Sarraf, Maria and Jiwanji, Morrtaza, “Beating the Resource Curse - The Case of Botswana”,

Environmental Economic Series Paper No. 83, October 2001.

Sunday Standard, 2005. “Debswana mines face closure”, 15/5/05

Taylor, Ian, 2004. “The HIV/AIDS Pandemic in Botswana: Implications for the ‘African

Miracle’”, 151-163, Poku, N. K. and Whiteside, A., (Eds.) The Political Economy of AIDS in

Africa. Aldershot: Ashgate.

UNAIDS 2004. “Botswana: 2004 Update”. Epidemiological Fact Sheets on HIV/AIDS and

Sexually Transmitted Infections. Avaialable at: www.unaids.org (quoted 17/6/05)

World Development Indicators, 2005. Data set compiled by the World Bank.

39

Appendix A. Value Added by Type of Economic Activity (Constant 1993/94 Prices) - Pula Million Economic Activity 1974/75 1975/76 1976/77 1977/78 1978/79 1979/80 1980/81 1981/82 1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1. Agriculture 423,4 431,1 451,1 445,5 421,4 422,0 381,5 391,4 339,6 293,1 285,7 318,9 296,4 492,0 450,8 2. Mining 222,9 365,3 382,7 704,2 707,2 975,5 1 318,9 1 585,5 2 291,2 2 617,3 2 685,6 2 790,8 3 024,6 3 146,8 3 778,7 3. Manufacturing 120,1 159,2 168,7 159,7 208,8 141,2 163,9 206,1 205,6 195,1 171,6 224,9 251,1 349,3 440,4 4. Water and Electricity 31,9 48,4 40,5 44,0 56,8 54,0 55,9 57,5 56,4 69,9 84,1 113,1 125,4 144,2 155,5 5. Construction 263,0 267,1 239,7 311,1 288,1 318,1 316,0 228,0 184,9 283,8 292,2 260,7 301,8 360,5 582,6 6. Trade, Hotels & Restaurants 159,2 179,0 188,3 182,2 248,3 196,6 157,6 190,7 132,6 102,9 244,4 361,6 332,8 409,0 565,7 7. Transport 26,1 23,5 29,6 32,8 32,0 56,0 57,1 66,0 74,5 84,2 110,6 141,5 132,0 211,2 262,1 8. Banks, Insurance & Bus Services 88,3 97,5 109,9 107,0 149,1 218,4 181,3 210,9 221,1 236,4 297,4 367,4 375,7 423,4 656,9 9. General Government 247,1 305,0 346,0 348,3 396,3 404,1 461,4 499,6 544,2 599,9 672,8 730,5 854,7 1 046,5 1 225,2 10. Social and Personal Services 43,1 57,8 55,9 57,4 61,7 62,1 76,4 87,8 104,7 104,1 127,8 145,4 160,8 214,7 338,4 P e r c e n t a g e o f T o t a l 1. Agriculture 24,1 20,7 20,8 17,2 14,8 13,0 10,6 10,1 7,6 5,9 5,4 5,6 4,8 6,9 5,1 2. Mining 12,7 17,5 17,6 27,1 24,8 30,0 36,8 40,9 51,0 52,6 50,7 48,9 48,8 44,2 43,0 3. Manufacturing 6,8 7,6 7,8 6,2 7,3 4,3 4,6 5,3 4,6 3,9 3,2 3,9 4,1 4,9 5,0 4. Water and Electricity 1,8 2,3 1,9 1,7 2,0 1,7 1,6 1,5 1,3 1,4 1,6 2,0 2,0 2,0 1,8 5. Construction 14,9 12,8 11,1 12,0 10,1 9,8 8,8 5,9 4,1 5,7 5,5 4,6 4,9 5,1 6,6 6. Trade, Hotels & Restaurants 9,0 8,6 8,7 7,0 8,7 6,0 4,4 4,9 3,0 2,1 4,6 6,3 5,4 5,7 6,4 7. Transport 1,5 1,1 1,4 1,3 1,1 1,7 1,6 1,7 1,7 1,7 2,1 2,5 2,1 3,0 3,0 8. Banks, Insurance & Bus. Services 5,0 4,7 5,1 4,1 5,2 6,7 5,1 5,4 4,9 4,8 5,6 6,4 6,1 5,9 7,5 9. General Government 14,0 14,6 16,0 13,4 13,9 12,4 12,9 12,9 12,1 12,1 12,7 12,8 13,8 14,7 13,9 10. Social and Personal Services 2,4 2,8 2,6 2,2 2,2 1,9 2,1 2,3 2,3 2,1 2,4 2,5 2,6 3,0 3,8 Economic Activity 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 1. Agriculture 465,4 478,4 491,2 488,6 467,2 459,4 489,9 453,1 479,9 443,4 404,6 444,5 433,2 441,2 446,1 2. Mining 3 603,0 3 943,3 3 935,2 3 766,9 3 956,2 3 899,4 4 076,3 4 310,7 4 721,8 4 588,5 5 142,3 6 045,9 5 864,9 6 471,7 6 918,5 3. Manufacturing 442,0 469,2 517,7 499,6 430,5 531,5 572,8 593,7 625,8 661,4 684,3 681,3 682,5 703,3 699,8 4. Water and Electricity 196,3 167,7 178,6 209,5 240,3 256,4 256,9 268,8 295,4 333,5 371,1 391,3 405,7 444,2 461,7 5. Construction 706,8 760,8 789,2 666,8 710,1 722,5 746,5 787,9 822,1 916,9 939,4 954,8 999,7 1 005,5 1 054,5 6. Trade, Hotels & Restaurants 667,3 588,5 535,2 541,6 882,3 1 085,8 1 192,7 1 359,0 1 422,7 1 501,9 1 595,6 1 700,0 1 839,7 1 901,3 1 956,5 7. Transport 281,4 322,0 364,5 390,3 406,5 435,9 437,7 456,4 497,8 578,7 594,0 605,5 625,4 631,2 639,1 8. Banks, Insurance & Bus. Services 804,8 880,7 917,9 1 050,2 1 144,4 1 231,6 1 351,5 1 367,9 1 500,8 1 636,3 1 707,3 1 794,7 1 922,2 1 972,5 2 071,4 9. General Government 1 249,8 1 348,8 1 552,0 1 622,4 1 706,7 1 762,4 1 854,9 2 009,4 2 195,7 2 333,3 2 474,2 2 640,6 2 861,0 2 965,5 3 102,4 10. Social and Personal Services 391,6 419,2 441,6 456,1 470,1 504,1 531,4 558,1 574,6 617,7 645,2 663,2 704,6 724,5 769,1 P e r c e n t a g e o f T o t a l 1. Agriculture 5,1 4,8 4,6 4,6 4,2 4,0 4,1 3,6 3,5 3,1 2,7 2,7 2,6 2,4 2,3 2. Mining 39,2 39,4 37,0 35,5 35,8 34,2 33,9 33,9 34,4 32,1 33,7 36,6 34,7 35,5 35,9 3. Manufacturing 4,8 4,7 4,9 4,7 3,9 4,7 4,8 4,7 4,6 4,6 4,5 4,1 4,0 3,9 3,6 4. Water and Electricity 2,1 1,7 1,7 2,0 2,2 2,2 2,1 2,1 2,2 2,3 2,4 2,4 2,4 2,4 2,4 5. Construction 7,7 7,6 7,4 6,3 6,4 6,3 6,2 6,2 6,0 6,4 6,2 5,8 5,9 5,5 5,5 6. Trade, Hotels & Restaurants 7,3 5,9 5,0 5,1 8,0 9,5 9,9 10,7 10,4 10,5 10,5 10,3 10,9 10,4 10,2 7. Transport 3,1 3,2 3,4 3,7 3,7 3,8 3,6 3,6 3,6 4,0 3,9 3,7 3,7 3,5 3,3 8. Banks, Insurance & Bus. Services 8,7 8,8 8,6 9,9 10,4 10,8 11,2 10,8 10,9 11,4 11,2 10,9 11,4 10,8 10,8 9. General Government 13,6 13,5 14,6 15,3 15,5 15,5 15,4 15,8 16,0 16,3 16,2 16,0 16,9 16,3 16,1 10. Social and Personal Services 4,3 4,2 4,2 4,3 4,3 4,4 4,4 4,4 4,2 4,3 4,2 4,0 4,2 4,0 4,0

40