The Resource Curse in Africa

The Resource Curse in Africa Chapter 1: Introduction The resource curse is the theory that countries with an abundance of natural resources, such as oil and minerals, achieve less economic growth than countries that are not endowed with natural resources. There are authors that argue this point (Auty 1990, Gelb 1988, Sachs and Warner 1995, 1997, 1999) and there are those that believe the resource curse is less to do with resources and more to do with political management (Brunschweiler 2008a, 2008b, Ross 1999, 2001). This theory appears to be contrary to the immediate instinct felt by many that natural resources will provide an opportunity for countries to develop by using increased revenues associated with a discovery of resources or an increase in world prices of such resources. The first section of this paper discusses a few of the most important mechanisms in which the resource curse can manifest itself. These are through institutions, corruption, conflict, Dutch disease and human capital. The second section studies the case of Botswana by providing an understanding of Botswanas economic and development path I will investigate how Botswana avoided some of the traps resource abundant countries usually fall into and how Botswana managed diamond mining in order to benefit the country. I will also perform a small comparison between Botswana and Sierra Leone to highlight the different channels in which resource abundance can affect economies. In my final chapter I will analyse whether Botswana has been successful in fully escaping the resource curse symptoms by discussing any problems the diamond mining may have caused. This paper, whilst small, highlights the importance of investigating phenomena such as the resource curse. Many countries have fallen into poverty since the emergence of extractable resources in their economies, yet Botswana has managed to avoid such problems. It is important to analyse how Botswana did this, so that lesson may be learned and used to help other countries avoid such problems in the future. The Resource Curse Literature Review Explanation of resource curse There are examples of resource poor countries outperforming resource rich countries throughout history. In recent times the Asian Tigers have achieved fast industrialization and economic growth despite having few natural resources, where as diamond rich countries such as Sierra Leone still remain low on world economic and social indicators (World Bank Development Indicators 2009). But do natural resources always lead to poor economic development? Or are there other variables in the context of which the natural resources are placed that determine economic development? There has been much discussion on the resource curse topic. Prominent among them are studies by Sachs and Warner (1995, 1997, 1999). Although in the conclusion to the paper Sachs et al admit their findings are far from definitive the general findings are that there is evidence for a negative relation between natural resource intensity and subsequent growth (Sachs and Warner 1995:p27). They analysed data from 95 developing countries by looking at annual growth rate between 1970-1990 and resource based exports in 1970. Sachs and Warner classified high abundance of natural resources as exports of agriculture, mineral and fuel as a percentage of GDP. From this simple analysis they discovered the existence of the so-called resource curse and they then tested the theory by controlling a number of other variables that could explain the relationship between resources and slow economic growth. The following discussion will provide a summary of the key mechanisms identified in the literature. Critique However it is important to note that not all academics support the literature on the existence of the resource curse, Brunnschweiler (2008a) is the most prominent. The main critiques of work by Sachs and Warner are the variables used to measure resource wealth. Brunschweiler for example believes per capita mineral wealth is more appropriate. The question is also raised as to whether Sachs and Warner were right to include agriculture in their regressions. Although agriculture is indeed a primary natural product, as it is the outcome of utilising the resource of land, it can be said that agriculture takes a different path in the economy than minerals or fuel. It may be better to classify agriculture separately when considering the resource curse, especially in the context of the third world where many economies are agricultural based. Further in this paper I will analyse whether agricultural resources cause the same effects on the economy as mineral resources. Arguments against the Resource Curse Big push theory Whilst there is no denial that the resource curse has effected countries like Venezuela, Nigeria and the Congo (Wenar 2008) there have been cases such as Botswana and Norway that have given strength to the opposing big push theory. In the big push model, developing economies are stuck in a trap. In order to develop their economies they need to industrialize so they are able to create wealth of their own. However there are large fixed costs associated with industrialization that developing economies cannot afford. Thus, the name big push comes from the idea that developing economies require a large injection of capital in order to develop. This injection of capital can be used to invest in economic infrastructure and will allow a more rapid accumulation of human capital which further allows social and economic development. (Murphy, Shleifer and Vishny 1989, Birdsall et al 2000) Ironically, this injection of capital could arise from large resource revenues, which suggests that the resource curse may be avoided if the capital is handled in a productive manner, for example a long-term and sustainable plan. An important point made by Walker and Jourdan (2003) is that as access to resources and minerals is becoming easier due to decreased transportation costs, countries are able to sustain industry without having a large natural resource base. This is could mean that countries who are not resource abundant might be better off, as they experience less adverse side-effects that I will discuss in this paper than resource rich economies. Dutch Disease The first effect, cited by Sachs (1995) and many others (Norberg 1993, Gelb 1988) is the Dutch Disease. Although it is often linked to the discovery of a natural resource, Dutch disease can occur when there is any positive income shock. For example a significant rise in primary product world prices can create sudden increased revenues for primary product exporters. The earliest Dutch Disease model I can trace was first created by Corden (1984) and the model has been constantly remodelled and analysed by other academics since. There are two strands of the Dutch disease model, the resource movement effect and the currency appreciation effect. Currency Appreciation Effect Dutch disease is often known as de-industrialisation because as one sector of the economy booms, (in the case of this paper we are talking about the tradable natural resource sector) other tradable sectors of the economy become less competitive. This is because a sudden increase in exportation of a natural resource of any kind can cause currency appreciation (Dutch disease). Whilst this is good for the country as it makes imports cheaper, it makes all the exports from the country (apart from the natural resource) less competitive in the world market as it costs other countries more money to from that particular country. The same path is also true for investment in this sector. This is why the manufacturing sector of resource abundant countries often shrinks. Resource Movement Effect The resource movement effect is the relocation of production factors away from the manufacturing sectors towards to booming (natural resource) sector. Davis and Tilton (2005:238) believe the Dutch disease actually allows a country to benefit from its new found mineral wealth by encouraging resources to flow from other sectors of the economy to the booming sector: However, this resource movement is also a cause of the shrinking of the manufacturing sector noted above. An important point argued in Sachs and Warner (1997) is that the shrinkage of a manufacturing sector itself is not a problem. The problem arises when the shrinkage causes slow economic growth, such a case may occur when an economy becomes more dependent on their natural resources. The advisability of this is not good (Jefferis 1998) as the economy becomes more vulnerable to world price changes in the natural resource. In turn, these often volatile price changes make it hard for governments to make mid or long term economic plans and policies. This is often said to be what happened in the oil rich Middle East in the 1970s (Auty 1990) governments were over optimistic about the earning power of their resources and then the oil prices fell dramatically. However the shrinkage of the manufacturing sector can have a negative impact on the economy because productivity grows faster in the manufacturing sector than in the resource sector (REF) and a decline in this sector means the economy is losing out on this productivity. A similar argument is made by Gylfason (2001) about learning-by-doing and technological advances. The shrinking of a tradable manufacturing sector also creates job losses; usually this could be compensated for as the primary product resource sector expands. But most minerals and oil sectors are capital intensive and not labour intensive (Sarraf 2001), so they are not able to absorb the unemployment. Institutions Another strand of the resource curse theory is the analysis of the relationship between resource abundance and institutions. Defining institutions is a difficult job as it can involve many different aspects of a countries history, culture and government. The main reason why institution analysis is vital to discovering the relationship between resource abundance and economic growth is that institutions affect policy structure and policy structure sets the arena in which an economy and resources are managed. Colonisation Firstly, the history of a nation can go a long way to explaining the current situation a nation faces today. Acemoglu et al (2001) notes the importance of colonisation is the determination of institutions. In his view there are two types of colonisation. There is settlement colonisation whereby the colonisers decided to settle in the region, perhaps due to a low incidence of deadly infectious diseases, as occurred by the British in North America. The settlement colonies are mostly made up of the new world, such as North America and Australasia. The second type of colony are extractive colonies, whereby the colonisers extracted resources that they found valuable, be it people as slaves or minerals. Naturally, these two different paths have caused quite different outcomes in institutions. Intuitive thinking would lead us to believe that traits of a colonisers such as property rights and rules of law would be embedded in there colonies. Indeed, Murshed (2001) and Acemoglu et al (2001) publish papers along the same lines. This theory leads to the thinking that colonies with exploitative colonisers tended to not establish foreign good institutions to the same strength as settled areas. It could be argued, as by Murshed, that patterns of exploitative behaviour with regard to resources were learned and inevitably repeated by colonies. On the other hand, settled colonies tended to retain institutions of law and property rights that European colonisers may have brought over. Acemoglu has raised the point that different colonising nations have left very different institutions behind. For example, he argues that the British colonies inherited better institutions with regard to respect for the rule of law and democracy (Acemoglu et al 2001:p12). In either case it is evident that institutions brought in from Europe have remained in some form. However, we must not forget that the colonies of Africa, Asia and Latin America existed long before they were discovered by the Europeans. Prior to colonisation these countries had their own functioning political and social institutions and it could be that European invaders only adapted these institutions to fit their needs and left many existing ways intact. A small but important point to note is that colonisation could also have had an impact on the ethno linguistic and ethno fractionalisation of a country because artificial country borders were placed upon areas of land with no regards to considering the existing, and it could be argued natural, borders. These artificial borders were emplaced for the ease of the colonisers and very little consideration was given to existing social borders, for example between tribes or geographical boundaries. This enforced ethno fractionalisation can be the cause of conflict within a country, even if natural resources are not in the equation, a prime example of this is in Rwanda. Leite and Weidmann (2002) are of the opinion that resource wealth does not directly affect economic growth, but that it resources affect the likelihood of corruption, which therefore influences economic growth. Bulte, Damania and Deacon (2005) further this argument by pointing out it may not be the existence of institutions that matter but the quality. In Bulte et al (2005) analysis they differentiate between two types of resources. Point resources which are geographically based and therefore an abundance of these resources are typically associated with inequality in terms of power and the division of the surplus, and often are accompanied by vertical relationships between agents (shareholders, managers, labourers). (Bulte et al 2005:p1031). Whereas diffuse resources, such as agricultural land, are more geographically spread and are therefore more equally distributed and less able to be protected by an elite. It is the belief of Bulte et al (p1034) that point resources attract worse quality institutions than diffuse resources with regard to corruption and government performance. Corruption An abundance of natural resources provides substantial revenue for an economy but unfortunately in many developing countries where there are weak infrastructures and poor people, the temptation of this revenue can cause corruption especially in the political sphere. When a government experiences large flows of finance, especially if these flows are relatively sudden, for example a discovery of minerals or oil, or new technology that helps extract resources, it can be hard for a government to manage such flows (Dietz 2005). They may not have had experience in dealing with large sums of money. These sudden windfalls increase the opportunity for corruption as it is hard to keep track of the money and therefore it is easier and to steal and waste. Corruption also comes in the form of laziness. The political elite may chose to ensure they remain in power by buying political favour using the resource revenues. This undermines democracy, but as politicians are able to obtain large sums of money from resources it is easier to buy political favour than to develop good policies and there is little incentive to build infrastructure in other areas of the economy, as resources are the main source of income. One would assume that if areas of the economy were to start failing or not being developed and maintained to a satisfactory standard of the citizens then the citizens would demand action from the government. However, in circumstances where an abundance of resources are in the country, the government often tries to buy favour from the people by not taxing the citizens, instead they use the resource revenues to provide basic infrastructure, such circumstances could be classed as the rentier effect cited by Mehlum (2006) and Brunschweiler (2008a). On the other hand, the government could decide to use the resource revenues to aid them in an effort to block the formation of social groups. The government might try to do this because they fear groups independent of the government may demand more from a government that is unwilling to give more. As Ross (2001: p335) argues Scholars examining the cases of Algeria, Libya, Tunisia, and Iran have all observed oil-rich states blocking the formation of independent social groups; all argue that the state is thereby blocking a necessary precondition of democracy. This is one of the many ways resources appear to affect politics. Whilst the basics are still provided and the people have more money in their hands, the situation can cause problems as the government is not longer held accountable as it is not using the peoples money. Therefore the relationship between government and citizens breaks down. This leads to a less democratic society and one that Karl (1997) believes would be one more vulnerable to civil war. Conflict As previously discussed weak institutions and corruption can both lead to conditions that breed conflict as they diminish the governments ability to function properly. Although a lot of studies are unable to show a strong link between resource extraction and civil conflict (Ross 2004), in recent history there have been many examples of the internal conflict within countries that are abundant in diamonds, such examples are Sierra Leone and Angola. Collier and Hoeffler (2001) cite that war emerges as either a product of grievance or greed. In the case of natural resources it appears that greed is most likely due to the enormity of the revenues compared to other forms of government revenue. The conflict often takes the form of civil war within a country as fractions of society jockey for control over the resource wealth (although the conflict can also take the form of hidden conflict within governments). Collier (2004) suggests that high social and economic inequality, lack of political rights and religious or ethnic divisions in society cause civil wars. The presence of natural resources can act as a catalyst by highlighting these faults and at the same time creating a financial incentive for war. If conflict does occur then (Brunschweiler 2008b) believes it could be the case the conflict makes countries dependent on resource extraction which is the default response when other economic sectors are not performing well. This indicates that once a country enters into conflict due to resources, they may have entered a vicious cycle that will be hard to stop. Fractionalisation within resource abundance societies has been greatly studied (Easterly and Levine 1997, Brunschweiler and Erwin 2009) as it is believed that societies that are fractioned by class struggles, ethnicity or religion have weaker institutions (Hoedler 2006). In turn weaker institutions lead to a diminished ability of the government to control situations and therefore fighting is more likely to erupt (Arezki et al 2007). Whilst the fragmentation in society is the basis for the fighting, the presence of resources can be seen as the trigger or catalyst for conflict. Fighting is bad for economic growth as it decreases productive activities, which lowers productions and lowers incomes (Hoedler 2006). Therefore in general the consensus is that the more homogeneous the society, the less likely the risk of conflict. Having high revenues from resources can also lead to rentier state symptoms as discussed previously. In this case it may be that those in control of resource revenues are constantly on the lookout for opposing groups trying wrest control from them. Unfortunately as they are the ones controlling the revenues, they have money to repress citizens by not only repressing social groups but by employing armed conflict if they require. This is why (Brunschweiler 2008b) believes that as governments are able to fund themselves they are more likely to be authoritarian. Although Rosss (2001) paper primarily discusses the Middle Eastern oil states, he admits that his findings can relate to other mineral economies outside the Middle East. This repression can also distort the economy by squashing entrepreneurial talent (Alayli 2005) However there are some scholars who believe that resource abundance can actually help avoid conflict, for example, Brunnschweiler and Bulte (2008b) say resource wealth raises income, and higher incomes, in turn, reduce the risk of conflict. However, they admit it is a small reduction in risk and it could be that the large prize (resource revenue) that people are able to fight over is a stronger incentive than higher incomes. The link between ethnic fragmentation and the resource curse has been investigated in a paper by Roland Hodler (2006). The aim of his paper is to explain why resources can be a blessing for some countries and a curse for other. For him, there are two effects of natural resources. Firstly, income of a country rises if the country chooses to use the resource for its own industrial benefit or exports them to other countries; this is a direct positive effect. Secondly, an indirect negative effect is natural resource wealth increases conflict, but only (according to Hodler) if there are multiple groups opposing each other. Hodler focuses on rivalling ethnic groups, but other groups that could affect the equilibrium are class groups and political groups. In Hodlers model the resource abundance is a blessing to a country if the direct positive effect is greater than the indirect negative effect, but a curse if the negative outweighs the positive and thus a relatively homogenous society is l ess likely to experience a resource curse as there are less opposing groups challenging the equilibrium. Linked with this argument is that of Bannon and Collier (2003:3) that ethnic dominance alongside resource richness breeds conflict. Ethnic dominance especially in government or institutions has an important advantage because that race then have the power in moderating and equalising ethnic relations, or neglecting and perhaps exacerbating them (Good 2005:p31) The magnitude of the negative effect in Hodlers paper is determined by the number of opposing groups. The higher the number of groups the weaker property rights. I take my definition of property rights from Acemoglu et al (2001) whereby people have secure property rights (rights against expropriation and that those with productive opportunities expect to receive returns on their investment and that a broad cross-section of society have the opportunity to invest. Furthermore Acemoglu et al (2001) make the point that relative political stability is needed in order to maintain these property rights. Human Capital As we can see that the resource curse is interlinked with many aspects of economics. It is also linked with human capital: education and health. Education is important in economic development as it raises labour efficiency, provides a more participative society and a better quality of life (Barro 1997), but is educational development being affected by the resource curse? Evidence by Gylfason (et al 2001: p850) shows that school enrolment at all levels tends to be inversely related to natural resource abundance, as measured by the share of the labour force engaged in primary production, across countries. Questions could be raised about the validity behind using such measures and whether other measures are more appropriate but there is plenty of evidence from other authors such as (Birdsall et al 1997) which come to the same conclusion. There are two prominent arguments about the effect the resource curse has on education and vice versa. Firstly, the vast revenues created by an abundant resource can be used by forward thinking governments to fund education (Sachs and Warner 1997). On the other hand, it has been argued by Gylfason that some resource dependent economies choose not to invest in education infrastructure as they see little immediate need for it because high skill labour and high quality capital are less common in primary production then elsewhere (Gylfason 2001: p10). However focusing on resources (and neglecting education) hinders the learning-by-doing process. This process is more likely to develop, along with gains in technological advances, in the manufacturing sector (Sachs and Warner 1995). Thereby depending on resources and neglecting education can slow economic growth as a whole as there is no incentive to increase the earning power (both at individual and national level) that can be achieved thr ough education. It is also worthy to note that education is strongly linked with a higher rate of absorbing new technologies from other countries (Birdsall 1997). The Case of Botswana Although in the previous section I discussed ways in which an abundance of natural resources could lead to slow economic growth, there have been countries that are resource rich and have had good economic development: for example Norway and Botswana. Norway became one of the top scoring countries on both economic and social indicators in the world (Larson 2003) since the extraction of oil in the early 1970s. Graphs showing growth of Botswana compared to other African nations here In this section I aim to discuss the experience of Botswana through the same key mechanisms I used in the previous chapter. These mechanisms are Dutch disease, institutions, conflict and human capital. By using the same key mechanisms I hope to show how Botswana has avoided the problems that cause the resource curse. Botswana has developed relatively rapidly considering that Botswana was the third poorest country in the world before independence (Beaulier 2003:p233) As Acemoglu et al (2001) points out there were only 22 graduate Batswana, who studied outside the country and only 12km of paved road. It seems that Botswana was in the same position as the majority of Sub-Saharan Africa. But since the average growth in Africa has been negative since 1965 (Acemoglu et al 2001) how has Botswana managed to achieve the highest rate of per capita growth of any country in the world in the last 35 years (Acemoglu et al 2001)? Especially considering that as a resource rich country it could be expected that Botswana would have slower growth than those without resources. Dutch Disease There are contradicting views as to whether Botswana experienced Dutch disease and whether this was due to the presence of diamonds. Mogotsi (2002:129) thinks that a mild Dutch disease occurred in Botswana as there was no large pre-existing manufacturing sector, so when mining occurred, the skilled labour from the small manufacturing sector moved to mining. Less skilled agricultural workers filled the place of the old manufacturing workers. As they are less skilled there is some loss of productivity and efficiency in the manufacturing sector. However Pegg (2009:p2) believes that there is little evidence that agriculture or manufacturing in Botswana has suffered from Dutch disease effects like the Dutch Disease model predicts when there is a large tradable mining sector. This is because there is very little resource movement as the diamond industry in Botswana as diamond mining is capital intensive and site specific (Jefferis 1998). This lack of movement means that few positive externalities are present in Botswanas mining industry. This is evident in the employment rates. Whilst Botswana has many good economic and social indicators, unfortunately a high unemployment rate is not one of them. While mining production contributed 40% to GDP, it absorbed only 4% total employment (Iimi 2006a:p7). This has large implications for income distribution and inequality in Botswana. As wages are higher in the diamond industry (REF) it distorts wealth in the economy. It has been said that only around half of the population have benefitted from the increased revenues, outside of gains in education, healthcare and infrastructure. This is reflected in around 50% of the population still living below the poverty line despite GDP per capita being around $1000 as there is a small workforce for diamonds and a high unemployment rate in general. (http://www.thuto.org/ubh/bw/bhp5.htm) However, in the resource curse theorem if Dutch disease were to occur then imports would be cheaper. As Botswana is 80% Kalahari Desert (Beaulier 2003) agriculture is not a major industry and as such Botswana imports most of its needs. 75% of imports come from neighbouring South Africa (Iimi 2006b:p18) there are very little visible effects of the negative sides of Dutch disease. Currency appreciation is the most obvious side-effect of resource related Dutch disease. But large diamond revenues have not caused Botswanas currency, the Pula, to be consistently overvalued. (Pegg 2009:p4) Although Botswana faces a difficult situation with regards to exchange rates. Botswana must managed the exchange rates carefully as it imports 75% of its goods from South Africa (REF) but Botswanas exports are valued in US dollars. Therefore Botswana must try to keep the Pula stable against both the South African rand and US dollar at the same time to avoid increased prices of food or decreased earnings due to falling dollar prices. So far Botswana has managed this well. Botswana has also been forward thinking by accumulating large foreign exchange reserves (Jefferis 1998) which are important and useful to have because it gives them the ability to manipulate exchange rates to aid the domestic currency should it need it. The government also created the Public Debt Service Fund (PDSF) in 1972. It recognises that the diamond revenues may be beyond the governments absorptive capacity and so the PDSF allows the government to save money rather than overheat the economy by spending it. (Pegg 2009:p3).The Revenue Stabilization Fund (RSF) is especially useful in times of economic downturn like the current financial crisis, as they government are able to finance normal spending by using the savings rather than borrowing. Of course, although good governance has caused what is seen to be a success with regards to revenue management (Samatar 1999; Leith 2005), it has also been said (Pegg 2009:p2) that stability of rent streams also helped Botswana control the massive flows and not fall into resource related Dutch disease. This has also led Botswana to move upper middle income status in the World Bank classification. This is impressive as before independence Botswana was classified as a low income country. (World Bank Income Classification). Institutions Several authors have put forward the argument that inclusive pre-colonial institutions are responsible for Botswanas economic development as institutions are a reason why food policies are chosen and also enable good policy choices to stick. Beaulier 2003) Before colonisation it seems that Botswana society was generally inclusive. An important institution of traditional Botswana society is the role of kgotlas which are an assembly of adult males in which issues of public interest were discussed (Acemoglu et al 2001) Botswana society allowed open dissent of the King and chiefs in kgotlas which provided a fair and accountable society. A further point argued by Acemoglu et al (2001) and also by Englebert (2000) and Iimi (2006a) is that the relatively unintrusive nature of British colonialism left a lot of traditional and functional institutions intact. During the scramble for Africa in the 1800s Britain agreed to granted Botswana protectorate status requested by Batswana chiefs in 1885 (Beaulier 2003). The chiefs wanted protection from the South African Boers who were moving towards Botswana. However, Britain apart from protecting fr