McMahon_Natural Resource Curse-Myth or Reality

Final Draft

THE NATURAL RESOURCE CURSE:

MYTH OR REALITY?

Gary McMahonConsultantEconomic Development InstituteWorld Bank, Washington, D.C.October 21, 1997.

The Natural Resource Curse: Myth or Reality?

I. Introduction

The natural resource curse has long been a favorite topic of discussion and analysis among economists. As early as 1576, Jean Bodin of France said, "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."1 While the arguments have become somewhat more subtle, the debate continues to this day and shows little sign either of slowing down or decreasing in importance. In this paper, we will outline, discuss and critically analyze the main theoretical arguments, illustrating the different hypotheses with numerous country experiences. The main conclusion of the paper is that the curse (or "disease") is not a result of fate or bad luck, but rather is induced by poor economic policies in general or an erroneous or inadequate policy response to a shock to the resource sector. The prevalence of import substitution strategies in developing countries in the 1960s and 1970s was of particular importance. Nevertheless, a country which has both natural resources and a dominant rent-seeking political culture will most likely find it difficult to shake off the curse or cure the disease.

While the emphasis of the paper will be on problems related to resource shocks, either due to price changes or increased production, we will begin in section II with a general look at the problems of natural resource abundance in the absence of booms or busts. As we will see, some of the most important results of theories related to natural resource shocks are equally valid if resource production and revenues are following a level path.

The next two sections will examine the problems of boom and bust affecting countries with a natural resource abundance. Section III will analyze the issues from a "pure" economic viewpoint, while section IV will focus on the more political or political economy perspectives of the problem.

In section V we present a brief survey of cross-country analysis of the effects of resource abundance. Section VI contains four short case studies in order to illustrate the points in the previous three sections. The countries discussed include two that had positive experiences with resource booms, Botswana (diamonds) and Colombia (coffee), and two which had very negative experiences, Nigeria (oil) and Trinidad and Tobago (oil). The last section contains conclusions and policy recommendations.

II. The Evils of Abundance

While the emphasis of the debate over the last twenty years has been on the surprising frequency that a positive shock has had a negative effect on a country with a large amount of natural resources, the potential of natural resource abundance to reduce economic growth has long been recognized, as the quote in the first section testifies. Even in the absence of shocks, there are a number of channels through which an abundance of natural resources can lead to slower or negative

1 As quoted in Sachs and Warner (1995: 1).1

growth. While we would dismiss the argument that it is likely to lead to sloth—mining and farming are not soft jobs—except, perhaps, in cases of a superabundance of natural resources, other arguments deserve more consideration. These include declining terms of trade of natural resources, the problems of price instability, susceptibility to rent-seeking activities, and weak backward and forward linkages.

The decline of the terms of trade of natural resources has long been associated with the names of Raúl Prebisch and Hans Singer.2 We will not address the issue of whether or not the terms of trade have actually declined once productivity increases and quality changes of industrial goods are taken into account. However, if declining terms of trade for natural resources has actually happened, it should lead to factor allocations out of natural resources, helping to eliminate or mitigate many of the problems to be discussed in this paper. It would also cause the exchange rate to depreciate over time, making the industrial sector more competitive internationally.

In the 1960s and 1970s, under the leadership of UNCTAD, there was a great emphasis upon the economic problems associated with price instability of natural resources.3 This debate was of quite a different nature than the more recent discussions of booms and busts. The emphasis was on the welfare losses due to unstable incomes, the difficulty in investment planning for products whose prices were subject to cyclical, but unpredictable, variations, and the dependence on foreign exchange for crucial imports of intermediate and capital goods.

Nevertheless, it is not clear that global resource price instability leads to slower growth and deserve the title of "curse". Individual countries could (and did) set up producer stabilization funds in order to smooth incomes. If producer prices are stabilized, it will also be easier to make investment decisions, a topic of particular importance for tree crop producers, given the long gestation periods. Moreover, unstable incomes theoretically should lead to higher savings rates (if persons are risk averse) and higher investment. Most studies undertaken at the time did not find a relationship between export instability and economic growth.4 A more recent study by Knudsen and Nash (1990) found that in general domestic price stabilization schemes have been quite successful.

Although not emphasized in the literature of the time, price instability may be more important if it leads to large fluctuations in government fiscal revenues, it is difficult for governments to smooth revenues, and downward adjustment is difficult and costly. In particular, government expenditure which rises in the upswing is difficult to reduce when prices fall. However, this is more of a poor policy response than a curse. In fact, we shall argue below that precisely this kind of policy induced self-affliction with respect to government expenditure is one of the most important causes of what is often erroneously called Dutch disease. In section IV, the asymmetric response of government expenditure to price changes will be explored further in the context of booms and busts.2 For reviews of some of their arguments, see Prebisch (1984) and Singer (1984).3 Laszlo (1978) contains a comprehensive collection of United Nations documents on UNCTAD’s “Integrated Programme for Commodities”.4 Two of the most important studies on export instability and growth were MacBean (1968) and Knudsen and Parnes (1975). The former found no relationship while the latter found that export instability had a weak positive effect on growth.

2

Finally, price instability can also lead to large fluctuations in export revenues. For economies dependent on imports of intermediate and capital goods, this could have important effects on growth (as in the foreign exchange gap theories). However, if the exchange rate is free to clear the market, the most competitive industries and profitable investments should be able to obtain the foreign exchange that they need. Moreover, one could even argue that foreign exchange shortages should be an incentive to manufacture capital and intermediate goods domestically.

The greatest treatise on economic stagnation caused by rent-seeking in primary economies may be Joseph Conrad’s novel Nostromo, first published in 1904. In this novel the owner of a silver mine in a fictional South American country manages to both survive and wield enormous political power by yielding to and manipulating the aspirations of the local and national bureaucracies, who do little else but seek rents and conspire in general. More generally, mineral economies are often more prone to rent-seeking because of the enormous concentration of wealth in one geographical area or in a small number of companies. Agricultural marketing boards, although usually created on the grounds of providing services to the producers, can have a similar, if less dramatic, effect. Needless to say, price or quantity booms of the primary resource can greatly magnify these effects. We will return to rent-seeking in Section IV.

A strong argument could be made that the natural resource curse thesis and affiliated theories are nothing but variations on the linkage theories, proposed by Hirschman (1958) and others. In these theories primary goods, especially mineral resources, are produced in an enclave type of economy which has few if any backward and forward linkages to the rest of the economy. Therefore, it is necessary to give incentives to other parts of the economy, especially manufacturing, where strong linkages and/or externalities exist. In fact, if the linkage theories are correct, the government should subsidize or protect the manufacturing sector as the gains to the economy due to the linkage effects will be greater than the losses due to protection or subsidies. In essence, a resource abundant country does not need a boom to get Dutch disease; it is born with it. We will return to this point in the context of learning by doing in the next section.

In sum, many of the problems of a resource economy which have come to be associated with booms and busts are likely to exist in more normal states of nature, if they exist at all. In the next section we will look more closely at the problems associated with resource shocks and try to determine to what extent they are the “nature of the beast” and to what extent they are policy induced.

III. The “Pure” Economics of Resource Shocks

It would be of no surprise to anyone to learn that resource abundant countries suffer if there are negative price shocks or depletion of their resource bases. The fact that they can suffer even worse from positive price shocks or increased production (either via technology change or new discoveries) surely justifies the appellation, “the dismal science”, often given to economics. However, we hope to show in the next two sections that Dutch disease and similar afflictions are largely policy induced afflictions rather than “pure”, unavoidable economic ailments. After

3

presenting the fundamentals of the Dutch disease, we will discuss its economic foundations, leaving the political symptoms to the next section as much as possible. That is, political cycles, special interest groups, patronage, corrupt or inefficient bureaucracies and the like will be assumed not to exist for the moment. Then we will do the same for the investment (construction) boom thesis. Table 1 at the end of Section IV contains a summary of the main results in the four cases.

IIIa. The Dutch Disease

The Dutch disease is the common name given to resource abundant economies who appear to have been hurt (or not helped much) by a positive price or quantity shock. Its name arises from the effects on the Netherlands manufacturing sector caused by the discoveries of North Sea gas.5 A model illustrating how a positive resource shock could result in deindustrialization was outlined in the seminal paper by Corden and Neary (1982).

The Corden and Neary economy has three sectors: boom tradeables (BT), non-boom tradeables (NBT), and non-tradeables (NT). BT are the primary goods, NBT is considered the manufacturing or industrial sector, and NT is the services sector. When the boom hits, there are two main consequences, the resource movement effect and the spending effect, as illustrated in Figure 1 . The resource movement effect comes about as resources move from both NBT and NT to BT. The movement of resources out of NBT is called direct deindustrialization. The spending effect occurs as with the boom income the demand for goods in general rises. While the prices of NBT are fixed at international levels, the increased demand for NT results in an increase in their relative prices. Hence, more resources leave the NBT sector to go to the NT sector. This is called indirect deindustrialization and, in most cases, it is expected to be much stronger than direct deindustrialization. Similarly, the increased export revenues from BT results in a real appreciation of the foreign exchange rate. This can happen via a natural appreciation of a flexible exchange rate or an inflation rate higher than the rest of the world in the case of a fixed exchange rate. It is likely that in a mining rich country the spending effect would heavily dominate the resource movement effect, although for a country rich in agriculture, the resource movement effect can (and often has been) quite strong when a shock hits.

To this point, there is a natural, welfare-improving movement of resources to the most profitable sectors. There is neither a disease nor a curse. However, if there are dynamic learning effects from the NBT sector which do not exist (or are smaller) in the other two sectors, the boom begins to take on the appearance of a Trojan horse. Dynamic learning or learning by doing effects are important if there are externalities to society as a whole; ie they cannot be captured by the firm. Hence, production and innovation in one firm can lead to productivity gains in other firms. Thus, while a movement of resources out of NBT to BT and NT increases GDP in the short-run, it reduces it in the medium-run or long-run due to the reduced externalities.

In the absence of externalities, the end of the boom should move the resources in the reverse direction, bringing the economy back to where it was before the boom, if at a somewhat higher level

5 Corden (1984) notes that the term “Dutch disease” appears to have been coined in The Economist of November 26, 1977.

4

due to the greater investment resulting from the higher incomes during the boom period. However, if

5

Figure 1: Dutch Disease

6

Figure 2: Weak and Strong Dutch Disease: Non-Boom Tradables and GDP

DD - Dutch Disease

Source: Author’s elaboration.

7

there are strong learning by doing effects, not only might the economy be at a lower level than it would have been in the absence of the boom, it may be difficult to move resources back to the manufacturing sector. Due to the implicit productivity which was lost during the boom, large parts of the manufacturing sector may be uncompetitive as countries which did not benefit directly from the boom have moved far ahead in the technological battle. Hence, there is deindustrialization.

A large part of the confusion in the debate and the literature on Dutch disease resides in what we would consider to be an incorrect interpretation of the above. The movement of resources out of NBT does not constitute Dutch disease. It is a natural factor movement due to price changes. “Weak” Dutch disease only occurs if the boom is over and the economy has readjusted, national income is less than it otherwise would likely have been and this decline can be traced to slow or negative growth in NBT. “Strong” Dutch disease occurs if the former occurs and the country’s NBT sector has deteriorated so badly (or fallen so far behind its competitors) that it cannot recover. That is, there has truly been a deindustrialization and the country is worse off because of it.6 The dynamic paths of non-boom tradables and GDP during the boom and bust are illustrated in Figure 2.

To this point we have not distinguished between temporary and permanent booms. If the boom is permanent (or is likely to exist for a very long time), there will be no reason for resources to move back. There is still the possibility, however, that the economy would have been better off without the boom due to the learning by doing effects. If the boom is temporary, then the possibility of Dutch disease is more likely the longer the boom lasts as there will be a greater loss of dynamic learning effects. Note that in economic theory the windfall from a temporary boom should be mostly saved as it will have only a modest effect on permanent (long-term) income, while a permanent boom should be mostly spent, as there is no need to increase one’s saving rate to guard against the end of the boom. We will return to these points in the next sub-section on investment booms.

While we will argue below that examples of “economic” Dutch disease are difficult to find, when they do occur solutions are not difficult, at least from a technocratic point of view. Let us assume that externalities related to NBT production do exist and the boom is significantly affecting production in this sector in a negative manner. If a rise in the real exchange rate via the spending effect is the main culprit, there are two solutions. The first, and highly preferable solution, is to limit the appreciation by accumulating foreign assets and relaxing some import constraints (which will also mitigate inflationary pressures). In fact, accumulation of foreign assets is precisely what the most successful commodity “boomers” have done in the last 25 years, including Indonesia and Botswana. Foreign asset accumulation can be undertaken by the public sector, especially if the windfall is mostly accruing to the government, or the private sector. In the latter case, it may often be necessary to liberalize access to foreign asset markets.7

6 An economy could become deindustrialized and still be prospering if it has invested the windfall boom income in assets which generate a stream of income which keeps national income or welfare at a level higher than it would have been in the absence of the boom. The situation can be compared to two people who win the lottery. The first person gives up his $50,000 per year job, spends all the money in 3 years, and cannot get his job back. Another person invests the lottery winning in assets which generate $100,000 per year, upon which he lives for the rest of his life.7 Note that if the resource is owned by multinational companies, taxes or royalties are low, and they repatriate most of the profits, the problem solves itself. Paradoxically, for a country suffering tremendously from Dutch disease, the

8

If import controls are also relaxed or tariffs are reduced, one must be careful not to permit increased or cheaper imports of products competing with those being hurt by the appreciation of the exchange rate. There is also the possibility that there will be an import boom of consumer durables if it is widely believed that the policy will be reversed when the export boom is over. A large rise in imported consumer durables could greatly reduce the long-run benefits of the boom by reducing the amount of the windfall used for capital formation.8

The second policy to counteract the spending effect is to subsidize the NBT sector with part of the windfall. In a perfectly depoliticized world there is little to choose between this option and the first one. As discussed in the next section, however, in practice subsidies to the NBT sector can have very serious consequences and make the (political) Dutch disease much worse than it would have been otherwise.

If a significant part of the problem is caused by the resource movement effect, the solution is not as easy. Labour and/or capital is moving to the boom sector, regardless of the real exchange rate. Subsidies or other instruments to increase the profitability of the NBT seem the only solution, but, as noted above and explained in detail below, these are fraught with problems. Happily, the resource movement effect has usually only been important during beverage crop booms, with the extra resources being attracted from the agriculture sector. As beverage crop booms are almost always of very short duration and as agriculture is rarely the sector of important learning by doing externalities, it is not likely that a country is going to suffer a serious case of Dutch disease from the resource movement effect.

While there have been numerous variations on and embellishments of the original theory of the Dutch disease, we would like to end this sub-section with a brief discussion of two of the most important. The Corden-Neary paper and many of the following theoretical and quantitative studies assume full employment. If there is unemployment, at first glance the argument for the disease is weakened substantially. First, there is no need to pull labor away from the NBT sector, reducing or even eliminating the resource movement effect. Second, there is much less pressure to increase wages and, hence, prices in the NT sector, reducing the spending effect. However, if the unemployment is due to wage rigidities, the Dutch disease could be worsened. For example, wages in the NBT sector may need to fall to maintain resources and competitiveness, but if they are rigid downward, the negative effects of a rise in the relative price of non-tradeables will be enhanced. Harrison (1994: 9) argues that one of the prime causes for the little Dutch disease that the Netherlands suffered was due to economy wide wage negotiations, which kept the pre-boom structure of wages intact. Note, however, that we are beginning to move from our “pure” economic model to the political economy model of the next section.

The second important variation comes from the curious observation, noted by Fardmanesh (1991) among others, that there are many more cases of (non-boom) agriculture suffering from a windfall than manufacturing. This result often occurred as manufacturing was highly protected, usually as part of an import substitution development strategy, before the boom, so in essence was a

next best remedy to not exploiting the resource at all is to give the country’s patrimony to multinational companies!8 In such a situation, trade policy is said to have become endogenized. That is, the trade policies in place at a given time depend on the current prices for the most important exports.

9

non-tradeable, leaving agriculture as the only large tradable sector. In fact, the boom may result in even more protection to manufacturing to offset any real exchange rate appreciation. Thus, any resource movement effect and spending effect that occurred were at the expense of agriculture. Some countries, such as Trinidad and Tobago and Nigeria, used the oil booms to begin full-scale implementation of their import substitution strategies. Not surprisingly, the agricultural sector was devastated in these two countries.

Of course, unless there are important externalities to agriculture, the above results do not constitute Dutch disease. In fact, a priori, someone with a strong faith in the theory of Dutch disease would believe that the economy as a whole would gain from the (increased) emphasis on manufacturing and the neglect of agriculture.

To summarize this sub-section, Dutch disease is the result of the movement of resources out of the dynamic NBT sector to the boom sector and the NT sector, the latter as a result of a rise in their relative price and the appreciation of the real foreign exchange rate. However, if there are dynamic externalities to the NBT sector, they should be encouraged (subsidized) even in the absence of a boom. A boom only means that you may have to (or want to) subsidize them even further, but the windfall allows for this option. Moreover, real exchange rate movements can be substantially mitigated by putting a large part of the windfall in foreign assets. In sum, it is difficult to find either practical examples of pure Dutch disease or a convincing theoretical foundation.

IIIb. The Theory of Investment Booms9

In the theory of investment booms, the Dutch disease theory is extended to allow for a more explicit distinction between consumption and investment goods. In this framework the boom is recognized as temporary and, quite correctly, most of the windfall is saved as the recipients of the windfall want to receive a long-term benefit from this short-term gain.10 The problem arises from a shortage of attractive financial assets in which to invest. In essence, it is impossible to discuss investment boom theory in pure economic terms as we did with the Dutch disease in the last sub-section, as it is precipitated by questionable previous policy choices. However, in this sub-section we will try to keep the arguments as “clean” as possible.

Investment boom theory, illustrated in Figure 3, assumes rational recipients of the windfall who try to invest it wisely but there are few options. The small size of the available portfolio is largely due to the undeveloped or repressed nature of the financial sector and restrictions on the holding of foreign assets, including currencies. The domestic financial markets are highly controlled and interest rates are usually negative or low, so there is little incentive to put the windfall in domestic markets. As noted, foreign markets are off limits. Therefore, the recipient has little

9 The theory described in this sub-section is often referred to as the theory of construction booms, but as we believe that the effects are similar whatever the type of (relatively unproductive) investment, we have chosen to call it by the more generic name. Nevertheless, a relatively high portion of investment in construction is likely to magnify the effects described in this sub-section.10 In more formal economic terms, the windfall only has a small effect on permanent income, so consumption does not increase significantly. Consequently, most of the windfall is saved and invested to increase the permanent levels of income and consumption to higher levels.

10

choice but directly invest his/her windfall. In the theory of construction booms—see Bevan, Collier and Gunning (1990)—restrictions on imports of capital goods coupled with scarce management talent result in the windfall largely being invested in construction. This investment puts further pressure on the price of non-tradeables, magnifying their relative price increase, and further damaging the NBT sector. Moreover, there is little long-run productivity increase associated with investment in construction, so the long-term gains for the country from the boom are quite limited.

Even if there are no major problems in importing capital goods, similar problems can arise. The enormous increase in savings results in a similar increase in investment, but many of the projects have low rates of return. That is, too much investment is undertaken at the same time, with rational agents investing until the rate of return equals the returns on domestic financial assets, which are well below foreign financial assets. This result can happen if the public sector is the main beneficiary of the windfall and tries to give a big push to the economy by investing in a large number of projects simultaneously. It can also happen if the private sector is the main beneficiary of the windfall. Of course, the increase in the price of the domestic component of investment also reduces the rate of return. Given the very low rates of return on domestic assets, investment projects which would be marginal in an unrepressed financial system are now undertaken. Moreover, the savings which do go into the financial system have the curious effect of causing a quasi-financial liberalization. That is, maximum interest rates become non-binding due to enormous liquidity in the banking system. As consumer loan markets are usually undeveloped or very small, the increased supply of credit puts tremendous downward pressure on interest rates and marginal investment projects are able to find financing.

In essence the large rise in savings leads to a non-optimal investment path from society's point of view. Many authors, including Corden (1984), Gelb (1988), and Bevan et al (1993), argue that it is important to stretch out the investment boom well beyond the export boom. The principal solution is very similar to the remedy of pure Dutch disease if the government is the prime beneficiary of the boom. The windfall savings should be invested in foreign assets and decumulated as profitable investment opportunities offer themselves. The rate of return on these investments (adjusted for social considerations when appropriate) should be at least as high as on the foreign assets. The Government of Botswana, as discussed in more detail in the mini-case studies, has been one of the most successful in using windfall savings in this manner. Two critical rules which it follows are: (i) it does not invest in projects unless the necessary (skilled) manpower is available; and (ii) it takes into account the recurrent costs attached to the project, only investing if projections of future government revenues are able to meet these costs.

If the private sector receives most or a substantial part of the windfall, the remedy may be more complex. Simply allowing access to foreign assets will not resolve the problem if the windfall is distributed among thousands of smallholder coffee farmers. The liberalization of domestic financial markets or, at least, a great reduction in the amount of financial repression is likely to be necessary.

Finally, as was implicit in the Botswana example above, an investment boom may be worse than a great opportunity missed. If there are high recurrent costs in the future or, as is usually the

11

case, the investment is made over a number of years, there will be substantial costs incurred after the boom is over. If the investment is government financed, this will result in great pressures on the fiscal deficit. In Nigeria, also discussed in more detail in the mini-case studies, many megaprojects had to be abandoned in the 1980s due to the inability to either complete the projects or cover recurrent costs.

To summarize this sub-section, rational savings behaviour on the part of public or private agents may have negative consequences for the economy if they are not able to increase and adjust their portfolios adequately. The end use of this savings may be a construction boom or a large amount of investment in low-return projects. The investment boom can be particularly damaging if it results in large demands on the government budget once the boom is passed. Smoothing out the investment boom will require investing in foreign assets and, often, at least partial liberalization of the domestic financial system. Finally, it should be noted that investment booms, construction or otherwise, are not "pure" economic phenomena, but are the results of the policies in place at the time of the boom, the most important of which is financial repression.

12

Figure 3: Investment Boom

13

IV. The Political Economy of Abundance

In the last section we argued that the natural resource curse as depicted by either theories of Dutch disease or investment booms is not a "natural" economic development. In this section we will attempt to show that the curse is the result of inappropriate pre-boom policies or inadequate policy responses to the boom. Any afflictions suffered by a resource abundant country as a result of a large positive shock, or during more normal times for that matter, are largely policy induced. While the government is ultimately responsible for these policies, it must be emphasized that it often comes under severe pressure from sections of civil society, including the private business sector, to undertake incorrect or poorly conceived actions. As in the last section, we will begin with the Dutch disease and then move to investment booms.

IVa. The Dutch Disease

Dutch disease can be caused or exacerbated by policies in existence at the time of the boom or policy responses to the boom. In this subsection we will analyze both cases, taking care to distinguish between inappropriate or misguided policies and those more closely related to rent-seeking and corruption. In both situations, however, the root cause is the tremendous increase in government revenues caused by the boom.

The most dramatic cases occur when the government owns the resource or receives a very large royalty, as has usually been the case with oil in recent decades. Nigeria, for example, saw government oil revenues increase from 5 to 19 percent of GDP from 1973 to 1979! Often, governments will increase export taxes to capture a large part of the windfall. In Ethiopia, for example, during the coffee boom, the marginal coffee export tax (when the price was more than $1 per pound) was tripled to 66 percent.

However, even when the government does not own the resource and does little or nothing to the tax system, it is likely to see an enormous increase in revenues. Part of this is just due to the short-run expansionary effects of the boom. The bulk of such increase, however, comes from the fact that the tax base is heavily dependent on trade taxes, so the increase in exports and imports has a disproportionate increase on fiscal revenues. The government share even increases further in a financially repressed economy, where government treasury bills (or some similar instrument) are one of the few available financial assets. In Kenya during the coffee boom, Bevan et al (1992) estimate that the implicit tax rate on government treasury bills was 45%. In total, they estimate that the government received about 50 percent of the windfall, most of which was indirect as coffee taxes were very low and there were no special levies during the boom.

Auty (1993) argues that the below average performance of countries with abundant natural resources is largely a result of the ability to postpone reforms due to rents from the primary sector. In fact, it may be more precise to say that resource rents have often allowed the postponement of reforms to import substitution policies, which ironically were often justified as necessary to move away from a dependence on primary products. By the 1970s most of the benefits from import

14

substitution policies had been obtained and reform was urgent. While resource abundance allowed countries to postpone such reforms, resource price or quantity booms often helped to deepen them.

Recognition of the popularity and prevalence of import substitution policies is essential for an understanding of the effects of resource booms from the 1960s to the 1980s. It is then no longer surprising that (non-boom) agriculture rather than manufacturing was the sector most affected by the boom. Domestic manufacturing, highly protected under the import substitution policy, was a de facto non-traded sector. Hence, it stood to gain from the spending effect and, in many cases, the availability of cheaper imported inputs. Agriculture, however, had to bear the entire burden of the changes in internal relative prices and the appreciation of the exchange rate. Note that as agriculture was not a sector with significant dynamic externalities, there was no Dutch disease.

In fact, if the manufacturing sector was the source of learning by doing and other externalities, a resource boom coupled with an ongoing import substitution policy should have led to a large increase in the growth rate of a country, perhaps even a Rostowian type of take-off. That this rarely happened suggests strongly that the types of policies used to promote manufacturing were ineffective or that the dynamic learning effects in manufacturing are much smaller than commonly believed. The evidence of East Asia in the promotion of export oriented manufacturing suggests that the former is more likely the case. For the purposes of this paper, the important point is that resource booms allowed the continuation or bolstering of policies which had obviously become ineffective. In general, the manufacturing sector lobbied governments for greater protection or subsidies to offset the effects of the appreciation of the exchange rate.

More detrimental than the continuation or expansion of import substitution policies was the direct entry of the state into the manufacturing sector. It seems likely that in two of most negative experiences associated with resource booms, Nigeria and Trinidad and Tobago, the single most detrimental policy was the creation of state owned industrial enterprises. Not only did these failed undertakings eat up substantial portions of the boom, they left a legacy of debt and losses in the post-boom years, adding substantially to the fiscal deficits of the countries. In both of these cases, the goal was to promote industrialization by giving a big push to an import substitution strategy. In retrospect, the results were not surprising.

In order to give their industrialization strategies an even bigger boost, many oil-rich countries borrowed heavily on international financial markets on the strength of their resource booms. While these loans put even greater pressure on the exchange rate, they also led to disastrous results when the oil boom came to an abrupt end at a time of very high real international interest rates, ending up in the well-known debt crisis of the early 1980s. To exacerbate matters, many oil boom countries were also recipients of large capital inflows, which of course flowed out of the country as quickly as possible when the boom was over and the debt crisis loomed. Capital flight was a serious problem in many countries in the post-boom years, especially if there was little confidence in the government policy to meet the new challenges. In Venezuela, Gelb (1988: 314) states that this type of non-confidence vote was particularly acute: “much of the second windfall was used to compensate for the drop in private and non-oil public savings and to permit massive capital exports by the private sector.”

15

In many countries consumer subsidies ate up a significant portion of the windfall, especially with respect to oil booms and energy prices. The subsidies were often a way of redistributing the windfall more equally; just as often they were a way of protecting the government’s urban power base by keeping down urban consumer prices. When the boom was over, it was politically difficult to remove these subsidies, putting further pressure on the fiscal deficit. In Ecuador consumer subsidies were equivalent to 10 percent of household income by 1981, causing Gelb (1988: 187) to conclude: “The build-up of Ecuador’s foreign debt may therefore be attributed entirely to its domestic transfer and credit programs.” Attempts to end petroleum subsidies and raise gasoline prices since 1981 have regularly led to strikes and violent protests.

Civil service jobs and pay raises were the other most common way for the government to share the spoils, while simultaneously strengthening its power base. In Cote d’Ivoire, the coffee and cocoa booms led to a 50 percent increase in the size of the civil service from 1976 to 1981, while even Cameroon, a relatively successful boom recipient, gave very high wage increases to the public sector. When the boom ends, it is often difficult, if not impossible, to cut back on the size of the civil service, so any adjustment which does occur comes through lower wages. The ultimate consequence of these two developments is often excessive public employment at low real wages. It is always another source of pressure on the fiscal deficit.

Natural resource booms since the 1960s have also usually taken place in countries with repressed financial systems. That is, interest rates were highly controlled by the government, who also used or directed a high percentage of available credit. Moreover, capital flows were also highly or completely restricted, with the private sector having little or no legal means to obtain foreign assets. Financial repression is often an integral part of an import substitution policy. It is used to keep lending rates low and, more importantly, to direct credit to the key sectors; that is, those with the greatest potential of dynamic learning effects or other externalities.

We saw in the last section that the best remedy for Dutch disease may well be to place a large part of the windfall in foreign assets, which can be decumulated as needed when the boom is over. A repressed financial system does not rule out this option if most of the windfall goes to the government. However, if the private sector obtains a large part of the windfall, low real domestic interest rates and lack of assets to foreign financial markets may result in an investment boom of the type discussed above. Moreover, financial repression is likely to lead to excessive real exchange rate appreciation, a further incentive to concentrate the investment in non-tradeable activities such as construction.

To this point, we have not mentioned the problem of corruption. Bates (1994) argues that natural resource booms weaken the state as its traditional functions give way to redistributing revenue. In particular, finance ministries lose much of their power due to the abundance of riches. Hence, rent-seeking and patronage dominate the agenda of the government rather than socio-economic development. Harberger (1994) stresses the need for a great deal of autonomy for the fiscal and monetary authorities to resist the pressure to spend the windfall in manners which are not only wasteful, but often have post-boom repercussions.

16

In sum, we believe that the irreversibility of government expenditure and restrictions on buying and selling foreign financial assets are the most important factors behind the many negative outcomes of countries which experienced a resource boom. Whether an increase in government expenditure is linked to a poor policy framework when the boom hits, inadequate policy responses to the boom, or just plain corruption, the end result is often the same. When the boom is over, the government cannot quickly adjust this expenditure to the new level of tax revenues, there are large fiscal deficits financed by money creation, inflation rises, and, if the exchange rate is fixed, there is a further, completely perverse appreciation of the foreign exchange rate. Nigeria, for example, had a 60 percent appreciation of its exchange rate from 1980 to 1983, after the boom, when all of the major economic indicators were increasingly unhealthy.

It is often argued that governments should use some type of stabilization fund for commodities which contribute a large portion of fiscal and/or export revenues. In essence, a stabilization fund is nothing more than a permanent plan to accumulate and decumulate foreign assets as the price or quantity of an important commodity fluctuates. There has been a great deal of analysis of the financial return to various stabilization plans, a large part of which is attached to the ability to forecast future prices. Unfortunately, most commodities seem to follow a random walk; that is, past information on price trends does not help predict future price movements. The best guide to tomorrow’s price is today’s price.11 In practice, this means that the fund will often run out of resources unless the rules are such that in times of good prices, a very large part of the windfall goes into the fund. In general, from a pure economic viewpoint, stabilization funds are not profitable. As an alternative to or in conjunction with a fund, the use of hedging in commodity markets is often recommended. Studies by Basch and Engel (1991) and Larson and Coleman (1991) find, for example, that if the markets exist, hedging is preferable to the use of stabilization funds. If futures markets are not long (or deep) enough, then a combination of hedging and a stabilization fund is preferable to the fund by itself.

Much of the literature on the use of funds does not, however, pay enough attention to their primary purpose. That is, they are intended to prevent governments from increasing expenditure in good times which is difficult to reverse in bad times. The quantitative analysis of stabilization funds does not compare the performance of a country with the fund to a country without a fund subject to irreversible government expenditure. Of course, the funds are not immune to political tampering and corruption.

To conclude this sub-section, natural resource booms have often had disastrous consequences due to political economy factors. It would be difficult to call these results either weak or strong Dutch disease as they have rarely if ever resulted in a shrinking of the manufacturing sector or deindustrialization. In fact, the repercussions have often been of a macroeconomic nature, affecting all sectors of the economy sooner or later. Although it may still be possible to speak of a natural resource curse, which can make the government of a resource abundant country act in a perverse if not destructive manner, it does not take the shape of Dutch disease.

11 See, for example, Claessens (1994) for a thorough discussion of stabilization funds.17

IVb. Investment Booms

As noted in Section III, “unproductive” investment booms caused by positive resource shocks cannot happen in a neutral policy setting. If a large part of the windfall goes to the private sector, an unproductive investment boom will only occur if the financial system is repressed. The construction boom form of the investment boom also depends highly on restrictions on imports of capital goods by the private sector and/or scarce management talent. In both cases there is added pressure on the relative prices of non-tradeables and the real exchange rate to rise. Note that import substitution policies, especially if they restrict imports of intermediate and capital goods, will make this relative price increase and exchange rate appreciation more likely.

If the public sector receives a large part of the windfall, an investment boom arises when the government invests most of the windfall immediately rather than put it into foreign assets (or a stabilization fund). Generally, the absorptive capacity of the economy cannot sustain such a large increase in investment, much of which turn out to be unproductive. An investment boom led by the public sector is much more likely to happen if an import substitution strategy is being followed. The positive resource shock permits the government to do more quickly what it wanted to do in any case. However, the outcome is usually an even larger and more highly protected manufacturing sector, a significant portion of which is now owned and managed by the public sector. The large increase in investment may also be due to political pressures, with politicians from every region (or county) of the country wanting their constituents to receive a share of the windfall. Once again, in retrospect the results are not surprising, whatever is the driving force behind the surge in investment.

The investment boom and its corollary, the construction boom, are dependent on a unique set of policies being in place at the time of the positive resource shock—that is, various forms of financial repression and, although not absolutely necessary but very helpful, import substitution policies. The reality of the last 30 years is that most countries receiving large resource shocks had these policies in place, making what would seem to be a rare outcome quite commonplace. Note, however, that although unproductive investment booms have occurred frequently, the negative consequences were not so much due to the appreciation of the relative price of non-tradeables, as often emphasized in the literature, but more the result of the general uneconomic use of resources and the repercussions for future fiscal deficits of the government.

It is clear that public sector led investment booms could easily have been avoided by accumulating financial assets and postponing investments. The avoidance of private sector led investment booms would have required major policy change by the government. Collier and Gunning (1996) argue against the government playing a custodial role in the case of positive resource shocks. As discussed further in the next section, they find that the private sector does a better job of saving windfalls than the government. There is no evidence that the latter is more far-sighted and if it adopts appropriate policy, the negative consequences of excessive and low-return investments can be avoided. In particular, the government must accommodate the portfolio demands of private agents, accumulating and decumulating foreign assets on their behalf if it is difficult for private agents to hold foreign assets directly.

18

Table 1: Summary of Dutch Disease and Investment Boom Theories

“Pure” Economic Effects Political Economy Effects

Dutch Disease Resource movement from NBT and NT to BT sector.Increased demand for goods as income rises.Prices of NBT are fixed at international levels, but prices of NT increase due to higher demand (spending effect).Real exchange rate appreciation.Resource movement due to spending effect from NBT to NT.Loss of externalities due to lower production in NBT.After boom is over, growth is lower than in absence of boom due to loss of externalities.Possible deindustrialization if boom is long enough.

Country is following (or would like to) an import substitution strategy. NBT (manufacturing) highly protected.Financial markets are highly repressed.Boom causes relative price of NT to increase and real exchange rate appreciation. Latter exacerbated by financial repression.NBT lobbies for more protection and higher subsidies.Public sector invests directly in industrial enterprises.Portion of windfall used for consumer subsidies.Civil service employment and real wages increase.Corruption often increases due to abundance of riches in public sector budget.Government borrows on strength of boom.(Non-boom) Agriculture usually hit hardest as not protected and prices fixed at international levels.Manufacturing sector often has grown but under heavy protection. Few obvious externalities.Boom ends but there are large recurrent costs, losses of soe's, debt financing, larger civil service, and consumer subsidies. Difficult to reverse expenditures.External borrowing or money creation (with fixed exchange rate) keep exchange rate high or appreciating.Foreign exchange needed for inputs to manufacturing sector is rationed.Further damage to unprotected sectors. "Deagriculturalization" often occurs.

Investment Windfall leads to increase in savings Country is following (or would like to)

19

Boom and saving rates of both public and private sectors.Real interest rates fall to very low levels.Foreign financial assets not available and few domestic financial assets, so private sector invests heavily in projects with expected low rate of return and construction.Large increase in public investment, often in industrial enterprises.Large increase in investment results in higher prices of NT.NBT suffer due to spending effects.Public sector investment often of low productivity.After boom, high recurrent costs and losses of state owned enterprises lead to large fiscal deficits.

an import substitution strategy. NBT (manufacturing) highly protected.Financial markets are highly repressed.Windfall is largely saved by public and private sectors.Real interest rates fall to very low levels.Public sector increases or "big pushes" the IS strategy with large infrastructure and industrial investments.Due to financial repression, private sector has few options and invests in construction and marginal projects. Government often subsidizes private sector investment.Due to real exchange rate appreciation, NBT lobbies for more protection and higher subsidies.Unprotected sectors, especially (non-boom) agriculture usually hit hard.Public sector invests directly in industrial enterprises.Due to political pressures, public sector invests in most regions (counties) of the country.Government borrows on strength of boom and "pushes" even harder.Public sector investment of low productivity.Boom ends but there are large recurrent costs, losses of soe's, debt financing, and subsidies to NBT.External borrowing or money creation (with fixed exchange rate) keep exchange rate high or appreciating.Foreign exchange needed for inputs to manufacturing sector is rationed.Further damage to unprotected sectors.

NT - Non-tradables; NBT - Non-boom tradables; BT - Boom tradables; soe - state-owned enterprise

Source: Author’s elaboration

20

V. A Review of Cross-Country Studies

In this section we will review the results of eight cross-country studies that have looked at different aspects of the natural resource curse. Although the review is meant to be indicative, rather than exhaustive, given the similar findings of most of the studies, we think that some strong inferences can be made about the empirical reality of theories of the natural resource curse. Of the eight studies, four consist of a number of country case studies and a general synthesis, two contain statistical analysis of a large number of shocks, and two analyze large samples of countries to see if GDP growth rates in resource abundant countries were different than in resource scarce countries. Table 2 contains a summary of the results.

The analysis in the articles by Auty (1994) and Cuddington (1989) as well as the books by Gelb (1988) and Auty (1993) is based on a number of country studies. While both of the Auty studies focus on the natural resource curse in general, Cuddington and Gelb focus explicitly on the effects of positive resource shocks.

Gelb (1988) analyzes the performance of six oil boom countries—Algeria, Ecuador, Indonesia, Nigeria, Trinidad and Tobago, and Venezuela, of which Indonesia was the only country to manage its windfall well. The emphasis of the analysis is on poor investments undertaken by the public sector, which on average captured 80 percent of the windfall. Not only were many of the investments questionable in themselves, in general construction and operating costs were much higher than foreseen. Gelb (1988: 13) puts a great deal of emphasis on the policy priorities that were in place at the time of the shocks for the final outcomes. "One of the main findings of the study is that oil incomes were committed rapidly and allocated to uses determined by a country's pre-shock priorities and its political and economic institutions." In particular, the use of the booms were very much directed by the import substitution strategies being followed in most of the countries. He says that only Nigeria and Tobago and Trinidad seemed to suffer from Dutch disease effects, although it should be noted that in both cases the manufacturing sector expanded. In fact, the agriculture sector was hit hard in both countries, but unless there are dynamic learning externalities in this sector, it is difficult to say that there was Dutch disease. Indonesia’s success was largely due to its prudent macroeconomic management, both with respect to the fiscal budget and the foreign exchange rate. In particular, the windfall was spread over a long period of time. In sum, the “curse” mentioned in the title of the book certainly dominated the blessing, but it was a policy induced affliction caused largely by a perverse investment boom.

Cuddington (1989) focuses on the experience of five countries which received export booms in the 1970s—Cameroon, Colombia, Jamaica, Kenya, and Nigeria. He finds that the performance of each country depended heavily on the fiscal behaviour of the government. Overspending in a boom has a ratchet effect as it is very difficult to bring back down. Jamaica and Nigeria both borrowed on the strength of their booms, exacerbating the fiscal problem. Moreover, government borrowing to sustain the effects of the boom kept the exchange rate at high levels in Colombia and Nigeria. Cuddington (1989: 162) concludes: "Countries that managed booms well were typically those that (a) did not allow fiscal variables, exchange rates, agricultural producer prices, and wages to get badly out of line, (b) avoided indulging in wasteful and inefficient investment or investment

21

that involved burdensome recurrent (such as ongoing maintenance) costs, (c) limited increases in government spending to levels consistent with long-term trends in revenue collection, and (d) maintained prudent external borrowing and foreign exchange reserve policies." Note that he found no evidence of what we have called strong Dutch disease. With the possible exception of Colombia (discussed in more detail in the next section), he also found no evidence of weak Dutch disease. In sum, good performance depended on good policies.

Auty (1993) analyzes the experience from approximately 1970 to 1990 of six mineral (non-oil) economies—Bolivia, Chile, Jamaica, Peru, Papua New Guinea, and Zambia—although the emphasis of the book is on the four countries from the Latin American and Caribbean region. The four principal countries of the study all entered the 1970s attempting to diversify their economies via import substitution strategies. Given their initial conditions in the early 1970s, he ranks the countries according to their ability to resist external shocks, but he finds no correlation with subsequent performance. Moreover, “there is little correlation between the timing and size of external shocks and the overall economic trajectory of the four countries.” (Auty 1993: 50) While all four countries had much lower GDP growth than the average developing country from 1973 to 1988, it should be noted that starting in the mid-1980s growth picked up significantly in Chile and in the 1990s both Peru and Bolivia have improved their economic performances substantially. The strong movement away from an import substitution strategy—which began much earlier in Chile—is the most important reason for these turnabouts. In sum, there is little evidence of a resource curse in these case studies, except in the sense that resource abundance allows countries to continue with poor policies for longer than otherwise. The experience of these countries was very similar to other countries following import substitution strategies in the 1970s, resource abundant or otherwise. The performance of the economies changed when the policy changed.

Auty (1994) is in a similar vein to his previous work, although it analyzes the performance of six large newly industrializing countries—Brazil, China, India, Korea, Mexico, and Taiwan. The main result of the paper is that countries performed much better when they moved from import substitution industrialization to a competitive industrial strategy, focusing on manufactured exports. Countries with larger markets and more resources are able to delay reforms for a longer period. Thus, the modest market and resource deficient countries of South Korea and Taiwan reformed first and have performed the best. Brazil and Mexico were able to delay reforms longer due to both large internal markets and resource abundance. Once again, the only curse that abundant resources carries is the ability to postpone reforms for a longer period of time.

Davis (1983) and Collier and Gunning (1996) both examine the response of a large number of countries to resource shocks. Both papers pay particular attention to the distribution and use of the windfall gains.

Davis (1983) analyzes the effects of the coffee and cocoa positive price shocks to ten producers in the period from 1976 to 1978—Burundi, Cameroon, Colombia, El Salvador, Ethiopia, Haiti, Ivory Coast, Kenya, Rwanda, and Uganda. He calculates the share of total and increased export revenues going to the government and notes that well over 50 percent of the windfall goes to the government. He also acknowledges that, especially due to increased imports, these figures greatly underestimate the government’s share of the windfall. The rest of the article focuses on how

22

the government managed its windfall. Very large increases in capital spending was the general response, much of which included commitments well after the boom was over, leading to fiscal and balance of payments problems. Inflationary problems were frequent after the booms as governments substituted monetary expansion for the external stimulus to the economy. In six of the ten countries in the sample the cumulative rise in imports from 1976 to 1978 was greater than the increase in export revenues! On average, more than half of the increase in imports was for capital goods. Although he does not address the problem, there is no evidence of Dutch disease in his sample. In sum, his analysis points to public sector led investment booms, which were based on overly optimistic expectations with respect to both the length of the booms and the rate of return on the investments.12

Collier and Gunning (1996) undertake an analysis similar to Davis, with the major difference being that they calculate both the direct and indirect windfalls. Their sample includes 23 shocks, 17 of which were positive. The emphasis of their paper is on the different reactions of the private and public sector. As in all of the other studies, for the positive shocks they generally found very large increases in investment with a low return by the public sector, as well as subsequent fiscal deficit problems. They note that in 17 beverage booms the fiscal deficits increased by 2.1 percent during the boom and a further 1.6 percent after the boom! In most cases the private sector saved a higher percentage of the windfall than the public sector, although it was often constrained with respect to portfolio choices. While they note that the public sector should invest in foreign assets and stretch out the investment boom, the central argument of the paper is to put in question the role of government as custodian during resource shocks. As much of the windfall as feasible should be left in the hands of (or transferred to) the private sector, given that it is likely to behave more appropriately than the public sector, at least if it is given access to a wide enough variety of financial and physical assets. In sum, if there is such a thing as a natural resource curse, it may be due to the ability of governments to appropriate most of the windfall and the subsequent sub-optimal policy performance.

Both Davis (1995) and Sachs and Warner (1995) try to determine if there is a natural resource curse by dividing large samples of countries into resource and non-resource abundant and comparing their GDP growth rates. We will report their quite different findings and see if they are reconcilable.

Davis (1995) compares the median and mean performance of mineral (including oil) economies versus non-mineral economies. The countries in his sample of 91 countries considered to be mineral economies are those where the mineral sector accounts for at least eight percent of GDP and 40 percent of merchandise exports. In his initial year, 1970, 23 countries fit this definition, while in his end year of 1991, there were 33 such countries, 11 of which were new. He finds that whether he compares the mean or median levels of initial GDP, growth rates, or the human development index, the 22 economies that were mineral based in 1970 and 1991 outperformed the 57 economies which were non-mineral economies in both years. Even if the fuel economies are omitted, the other mineral economies generally did better than the non-mineral economies. While there are many examples of mineral abundant countries that have done poorly,

12 There is no analysis of how the private sector used its share of the windfall. Hence, there is no discussion of, for example, the construction boom in Kenya that Bevan, Collier, and Gunning (1992) have stressed.

23

the main conclusion arising from his evidence is that the resource curse is not a general phenomenon, and, in fact, a country is more likely to better off due to abundant resources than worse off.

Sachs and Warner (1995) compare the effects of all resource abundance (not just mineral) on growth by doing regressions on a sample of 96 countries from 1970 to 1989. The proxy for resource abundance is the share of primary exports in GDP in 1971. They find that that there is a strong, significant negative relationship between growth and initial resource abundance. They then do a number of other regressions and find that growth is positively related with openness, investment, and bureaucratic quality. The resource abundance measure is always significant. When they estimate a small structural model, they find resource abundance is positively correlated to a closed economy, but it has no correlation with bureaucratic quality. In sum, they find strong evidence for some sort of resource curse, although the underlying reasons are still not clear in this exploratory work.

Are the works of Davis and Sachs and Warner reconcilable? While the latter study is more sophisticated, it suffers from at least one serious reservation. By using the primary export share of GDP in 1971 as the proxy, this means that the countries which had huge oil booms in the 1973 and 1979 would have been much more resource abundant on average than their initial positioning suggested. Although this could have changed the results either way, it would be interesting to see if the results are similar if the average share of exports in GDP over the entire period was used as the proxy. Of course, it would also be useful if the Sachs and Warner regressions were carried out using the share of mineral exports in GDP or if Davis extended his definition of resource abundance to include all primary exports.

The papers reviewed in this section suggest that resource abundance can lead to poor policy choice and, hence, lower growth rates or mismanagement of booms. While there is little evidence of the existence of Dutch disease, over-ambitious public sector led investment booms seem quite common. Nevertheless, it is not clear that resource abundant countries on average would have been better off without the resources, or that resource abundant countries are more prone to bad policies than resource deficient countries.

24

Table 2: Results of Cross-Country Studies

Study Dutch Disease Investment Boom Resource Curse

Gelb (1988)6 countries

No. 5 countries. Yes in 5 countries. Policy induced. Linked to import substitution policies.

Cuddington (1989)5 countries

No. (Perhaps weak form in Colombia.)

3 countries. Yes in 3 countries. Policy induced. Linked to poor fiscal behavior.

Auty (1993)6 countries

No. No. Yes in 2 countries. Poor performances linked to import substitution policies.

Auty (1994)6 countries

No. 1 country. Yes in 1 country. In general delayed policy reform away from import substitution.

Davis (1983)10 countriesAll beverage booms

No. 6 countries. Yes in 4 countries.13

Collier and Gunning (1996)23 countries

No. Yes. Number of countries unspecified.

Yes in unspecified number of countries. Linked to role of government as guardian of windfall.

Davis (1995)91 countriesAll mineral abundant

Not relevant. Not relevant. No. Countries more likely to benefit from resource abundance.

Sachs and Warner (1996)96 countries

Not relevant. Not relevant. Yes. Resource abundant countries had slower growth.

Source: Author’s elaboration.

13 Davis (1983) does not address this question directly. In 4 countries both the fiscal and balance of payments position worsened during the boom.

25

VI. Four Mini-Case Studies

This section contains four case studies of countries which received major positive resource shocks in the 1970s and/or 1980s. We will begin with the case of Botswana and diamonds, widely acknowledged as one of the most successful examples of boom management. The relatively successful management by Colombia of the mid-1970s coffee boom follows. The third case is the disastrous oil boom in Trinidad and Tobago. We finish with the “super” disastrous oil boom in Nigeria. The cases were chosen both to show different types of results and to include large and small countries. In addition, given their almost identical populations, a comparison of Trinidad and Tobago with Botswana helps to highlight some of the conclusions of this study. Table 3 contains a summary of the results.

VIa. Botswana14

The diamond boom in Botswana, which began in 1965, was due to the discovery and development of large amounts of high quality diamonds, not a price increase. That is, it was a quantity boom, not a price boom. From 1966 to 1989, Botswana’s annual GDP growth rate of 8.5 percent was the highest of any country in the world. From 1985 to 1995, its GDP growth rate of 6.1 percent was only lower than Thailand, Korea, and Singapore. It could be argued that the high growth rate was due to the fact that the boom had to support a very small population—approximately 1.5 million in 1995. However, there are numerous instances of small countries not benefiting from booms, of which the case of Trinidad and Tobago below is a prime example. It could also be argued that Botswana’s success was due to the fact that, with the exception of a short period in 1981 and 1982, it has not yet had to face up to the “bust”. However, as has been demonstrated in the preceding sections and will be emphasized in the last two case studies, many boom countries were doing poorly or headed towards disaster even before the boom ended. For example, 8 of the 10 countries in Davis (1983) had budget deficits in the last year of the beverage boom, and the cumulative rise in imports was greater than the cumulative rise in exports in 6 of the 10 countries in the three booming years.

Botswana has not been affected by Dutch disease or other variations of the natural resource curse. Its manufacturing production even doubled, albeit from a small base, in the 1980s, with manufacturing employment growing by 190 percent, to the point where it was three times as large as mining employment. There was a significant rise in construction prices in the late 1980s, but there was no spillover of inflation to the rest of the economy or the foreign exchange rate. Agricultural production seems to have suffered somewhat, but as the most important activity in this sector was traditional livestock, there is no question of a loss of dynamic learning by doing.

The reasons for Botswana’s successful management of the diamond boom are not difficult to find. First, a large part of the windfall was put in foreign savings and only used when the absorptive capacity of the economy was deemed sufficient. Harvey (1992: 343) notes that: "With the exception of a period in the late 1980s, government spending policy paid close attention to two constraints: the availability of skilled manpower, and the future recurrent costs of development

14 This case study is largely based on the works of Hill and Mokgethi (1989), Harvey (1992), and Norberg and Blomstrom (1993).

26

spending in relation to revenue forecasts." Second, very close attention was paid to the foreign exchange rate. On one hand, via management of the windfall savings, it was not allowed to appreciate widely. On the other hand, to help manage inflationary pressures caused by the boom, appreciation of the Botswana pula was allowed periodically. Third, the government never went on an investment spending spree, except for a brief period in the late 1980s when the extremely high level of foreign exchange reserves put pressure on the government to increase its capital spending. Finally, and implicit in the above, Botswana had very good economic management teams at senior levels, including the Presidents.

When revenues from diamonds fell dramatically in 1981-82, the government responded quickly. Domestic credit growth was cut dramatically, interest rates were increased, wages and salaries were frozen, and the pula was devalued by 10 percent. According to Hill and Mugkethi (1989: 186), the idea was to spread the impact of the adjustment over the whole economy and not put the burden on a small segment. The policies were generally successful and the drop in GDP was only 2.4 percent despite a 16 percent drop in total export revenues and 43 percent drop in diamond revenues.

Finally, while income distribution in Botswana is highly skewed, it would be unfair to say that the general population has not benefited from the boom. As Harvey (1992) describes, there have been enormous increases in access to health and education, literacy, life expectancy, infrastructure, and water facilities. Although the country was one of most least developed in Africa at independence, in 1995 life expectancy in Botswana was 68 years versus 52 for Sub-Saharan Africa. Literacy in Botswana in 1995 was 70 percent versus 57 percent for Sub-Saharan Africa.

VIb. Colombia15

Colombia was one of a number of countries which benefited from an enormous increase in coffee prices in the years 1976 to 1978. There was also a quantity boom, starting in 1976, due to major technological improvements in the production of coffee. Production of coffee increased by 30 percent from 1975 to 1978 and by 63 percent from 1975 to 1981. Coffee export revenues in constant dollars increased by 145 percent from 1975 to 1978, the peak earning year. The country’s foreign exchange reserves increased tenfold from 1976 to 1980.

When the boom occurred, Colombia was vigorously promoting exports in order to reduce its dependence on coffee, although industrial promotion was done under the umbrella of an import substitution program. In fact, from 1967 to 1974 non-coffee exports had grown by 14.3 percent annually, while manufacturing exports had grown 27.1 percent from a small base. Coffee export revenues had fallen from 75 percent of total export proceeds from 1961 to 1970 to 44 percent in 1974.

Colombia was a rather unique boom country in that a large part of the windfall went to the privately owned Coffee Federation, which uses the funds to smooth out producers’ incomes. To

15 This case study is largely based on the works of Davis (1983), Kamas (1986), Cuddington (1989), and Wunder (1992).

27

moderate the effects of the boom on the exchange rate, the government followed very tight monetary policy as well as tight fiscal policy at the beginning of the boom. Nevertheless, the foreign exchange rate appreciated from 40 percent from 1975 to 1980. However, given that it had depreciated nearly 30 percent after the first oil boom, the net effect was an exchange rate that was only 11 percent higher in 1980 than in 1970.

Some authors, notably Kamas (1986), believe that Colombia suffered from the Dutch disease as the result of the boom. She notes that the coffee share of export revenues had increased back to 60 percent by 1980 and that the growth rate of manufacturing had decreased. That is, the manufacturing sector did not decline; only its relative position in the economy had declined and the growth rate of manufacturing exports had slowed down from a furious pace of 27.1 percent from 1967 to 1974 to 10.5 percent from 1975 to 1980. It is impossible to call this deindustrialization; it is difficult to even call it weak Dutch disease. As the effects of the government’s export promotion policy were bound to slow down sooner or later, it is likely that a large part of the decrease was due to the fact that the easiest choices had already been made. The data are not consistent with what Cuddington (1989: 149) calls the collapse of non-traditional exports. In fact, by 1982 the coffee share in total export revenues had already fallen back down to 51 percent.

If we look at the larger picture, the case for Dutch disease weakens further. Colombia was the only major Latin American country for whom the 1980s was not the “lost decade”. GDP growth was a respectable 3.7 percent per year. Moreover, the industrial sector growth rate was 5.0 percent per year, and the share of the manufacturing sector in export revenues increased from 20 to 40 percent from 1980 to 1993! It is difficult to reconcile these facts with Dutch disease.

Colombia also managed to avoid an unproductive investment boom. As noted above, the government followed tight fiscal policy at the beginning of the boom and tight monetary policy during and directly after the boom. Public investment fell from about 4 percent of GDP to 2 percent during the boom. However, in the early 1980s there was a large increase in public investment to 7.5 percent of GDP. A large part of these investments, however, were related to the exploitation (with foreign partners) of new discoveries of coal and oil, which have had substantial benefits in the 1990s. They do not seem directly linked to the coffee boom, although the build-up of foreign exchange in the last half of the 1970s may have accelerated their introduction. In essence, it could be argued that a large part of the coffee boom was used quite wisely to diversify the export base.

The only significant poor policy response with respect to the coffee boom in Colombia was the large increase in government recurrent expenditure, beginning in 1977 and continuing after the boom was over. These difficult to reverse increases led to substantial and unnecessary fiscal deficits in the first half of the 1980s.

In sum, although some authors have argued differently, we find no convincing evidence that Colombia suffered from the Dutch disease as a result of the coffee boom. There is also no evidence of a perverse investment boom, and, given the performance of Colombia relative to other Latin American countries in the 1980s, the balance is tilted towards a resource blessing rather than a resource curse.

28

VIc. Trinidad and Tobago16

Given the ultimate disastrous results, it is surprising to note that the first response of Trinidad and Tobago to the 1973 oil boom was precisely what has become the standard optimal policy prescription: it invested 70 percent of the windfall—which was equivalent to 39 percent of non-mining GDP from 1974 to 1978—in foreign assets. Foreign exchange reserves rose from $47 million to $1.8 billion from 1974 to 1978. However, beneath the surface there were some disturbing currents, whose full effect was only felt in the 1980s.

A few years before the boom struck, Trinidad and Tobago had set out on an import substitution strategy in order to diversify the economy. The oil windfall allowed the government to expedite full implementation of this strategy, and in 1975 it began a plan of gas-based industrialization. Shortly after, the government began to use part of the windfall to acquire a large number of declining industries. It is important to emphasize that they were not failing due to Dutch disease effects, as the foreign exchange rate appreciated only five percent from 1974 to 1978 and there was considerable unemployment, suggesting that neither the resource movement nor spending effects were important. In fact, most of these industries were in decline long before the oil boom.

Public pressure to share the benefits of the boom also led to large consumer subsidies for food, fuel, and utilities. By 1981 these were equivalent to five percent of GDP. The political difficulty in cutting back on these subsidies was an important element in the economic collapse of Trinidad and Tobago in the 1980s, when annual GDP growth was -2.5 percent.

The second oil windfall was equivalent to 34.7 percent of non-mining GDP from 1979 to 1981. A large portion was used for further industrial investment, although few additional declining industries were acquired. Subsidies to consumers and the loss-making enterprises acquired with the first windfall also ate up a large portion of the windfall. Unfortunately, most of the new investment was very unproductive due to very large cost over-runs during construction and poor implementation after completion.

The end result was that foreign exchange reserves fell from $3.3 billion in 1981 to $1.1 billion in 1983 in order to finance large fiscal deficits. From 1982 to 1987 annual growth in the non-boom traded goods and non-traded goods sectors was -3.8 percent and -5.7 percent respectively. The share of manufacturing in GDP was the same in 1990 as in 1973.

Did Trinidad and Tobago suffer from Dutch disease? Although the production of non-boom traded goods declined after the second oil boom, it actually expanded 3.8 percent during the two booms. In addition, the small manufacturing sector was the same relative size when "the dust had settled" in 1990 as before the boom. While the 20 percent appreciation of the currency from 1979 to 1982 could be blamed for the subsequent decline of industry, the fact that the decline was generalized across the economy suggests that the large decline in foreign exchange needed to buy

16 This case study is largely based on Auty and Gelb (1986) and Harrison (1994).29

imported inputs plus the end of the public sector investment boom were likely of far more importance. As in many countries the main casualty in Trinidad and Tobago was the agriculture sector, which virtually collapsed, falling 1.8 percent per year in the 1970s and 5.8 percent per year in the 1980s!

Clearly, Trinidad and Tobago suffered from an unproductive investment boom led by the public sector. The use of the resources for new investment was part of the government's industrialization strategy. Their deployment to buy existing firms was due to the strength and militancy of the trade unions. In addition to the large increase in public employment, this same power base was able to obtain large wage increases, hastening the demise of the low-wage agriculture sector.

In sum, natural resource abundance was indisputably a curse for Trinidad and Tobago, although there were some short-term consumption benefits in the 1970s. The windfalls allowed the government to follow questionable policies and succumb to the pressure of special interest groups. When the boom ended, the recurrent and "irreversible" expenditures associated with these choices meant that not only had little been gained from the boom, much was about to be lost.

A comparison of the experience of Trinidad and Tobago with Botswana, two small countries with similar booms, clearly reveals that policy choices led to the starkly different results.17 First, Botswana took into account recurrent expenditures in all of its policy decisions, thus avoiding both fiscal deficits and external debt. Second, it did not undertake investments which were beyond its absorptive capacity, concentrating on small and medium sized projects. Third, it followed a much more prudent exchange rate management policy. While it might be argued that Botswana's boom has not ended, it seems fairly clear that even before the second oil boom ended, Trinidad and Tobago was in a difficult situation due to poor investments and policies. In contrast, if the boom in Botswana ended tomorrow, the country would be relatively well-placed.

VId. Nigeria18

The Nigerian experience with the two oil booms of the 1970s is akin to a hyperactive Trinidad and Tobago. Not surprising, the results were very similar but even more dramatic. Like in Trinidad and Tobago, the Nigerian government was intent on following an import substitution industrialization policy when the boom struck, a boom that was magnified by the large increase in oil production in the 1970s.19 Similar to Trinidad and Tobago, Nigeria began by accumulating foreign assets and soon began massive public investment for industrialization and subsidies to private industrial investment. The period of asset accumulation was, however, much shorter, while

17 See Harrison (1994) for an in-depth comparison of resource boom management in Trinidad and Tobago and Botswana.

18 This case study is largely based on Gelb and Bienen (1988), Bevan, Collier, and Gunning (1992), Nyatepe-Coo (1994), and Collange (1995).19 While it is difficult to measure the exact size of the windfall in Nigeria, due to the combined effects of recovery from the civil war and exploitation of new discoveries, it was between 20 percent and 30 percent of non-mining GDP each year from 1974 to 1978.

30

the investment plan was quicker and larger. From 1970 to 1976 public capital spending rose from 3.6 percent to 29.5 percent of non-mining GDP. One consequence of this spending was that Nigeria actually had an average fiscal deficit of nine percent of non-mining GDP from 1975 to 1978. The new investment was highly protected, with an average effective rate of protection of 177 percent in the manufacturing sector. In both countries the investment turned out to be quite unproductive. From 1980 to 1986 annual GDP growth in Nigeria was -3.8 percent.

The second oil price shock did nothing to alleviate the situation. The budget remained in deficit, averaging 12.3 percent of non-mining GDP from 1981 to 1983. The naira, which was 29 percent higher in 1978 than 1970, appreciated another 60 percent from 1981 to 1983. As the new industries were highly dependent on intermediate and capital goods imports and consumers had become accustomed to cheap imports, the government did not devalue. However, the real appreciation of the exchange rate meant that the government's fiscal revenues, over 80 percent of which came from oil exports, were falling in naira terms, putting more pressure on the budget.

The agriculture sector was hard hit by the movement of resources to manufacturing and services. Moreover, the largely policy induced exchange rate appreciations hit agriculture, the only unprotected non-boom tradable sector, very hard. From 1970 to 1982 the share of agriculture in GDP fell from 49 percent to 22 percent. If the investments had been productive, this new structure of production would have been optimal. The results of the 1980s show that this was not the case.

Nigeria, like Trinidad and Tobago, suffered from an unproductive investment boom rather than Dutch disease. The industrial sector grew rapidly during the boom, while agriculture, the declining sector, was not a sector of dynamic learning externalities.

In sum, Nigeria made all of the same mistakes as Trinidad and Tobago, except on a larger scale. Gelb and Bienen (1988: 259) note that Nigerians did not even enjoy the consumption boom of other unsuccessful countries, such as Trinidad and Tobago, Venezuela, and Ecuador. Moreover, the standard of living in 1986 was probably less than in 1970. The only substantial benefit was that massive investment in public education led to a very large increase in average levels of education.

31

Table 3: Summary of Mini-Case Studies

Country Weak Dutch Disease

Strong Dutch Disease

Unproductive Investment Boom

Resource Curse

Botswana No No (Strong growth in manufacturing)

No Strong blessing

Colombia No No (Strong, albeit slower manufacturing growth)

No Weak blessing

Trinidad and Tobago

No No (Large decline in agricultural production)

Yes (Public sector led)

Strong curse

Nigeria No No (Large decline in agricultural production)

Yes (Public sector led)

Very strong curse

Source: Author’s elaboration

32

VII. Conclusions and Recommendations

The empirical evidence strongly suggests that Dutch disease, which appears somewhat unlikely from a theoretical point of view, is a rare phenomena in developing countries. It was difficult to find even one conclusive case of what we called "weak" Dutch disease—slower overall growth due to the neglect of the industrial sector. Strong or classic Dutch disease resulting in deindustrialization was not present in any of the studies that we reviewed, which included most of the well-known cases of resource booms since 1970. In fact, the exact opposite was often true. More often than not the industrial or manufacturing sector grew during the boom, even in some of the most dramatic cases of failed boom management. When one sector was particularly hit hard, it was almost always the agricultural sector. Given that the premise of the Dutch disease is slower growth due to resource allocation away from the sectors with strong externalities or dynamic learning effects, the movement out of agriculture should have increased growth rates. The failure to do so was due to the policies in place when the boom hit and the policy responses after the boom, not Dutch disease.

In theory, true Dutch disease would only be a real danger to a developing country when its industrial policy is based on export-led competitiveness. However, there are no good examples of such a development. In contrast, there are countries, such as Thailand, Indonesia, and Malaysia, which have had successful export led industrialization despite natural resource abundance, once again suggesting that the curse is policy induced and not an inherent feature of resource abundant countries

Nevertheless, it is clear that the economies of many resource abundant countries performed very poorly, and it is difficult not to conclude that they would have been better off without the resource boom. The curse of resource abundance most commonly manifested itself as a perverse or unproductive investment boom. While this investment attracted resources and put pressure on the price of non-tradables, its most important negative feature was the recurrent costs, including subsidies to loss-making state owned enterprises, that the country faced after the boom had ended. Unproductive investment booms were usually led by the public sector and often accompanied by large increases in "irreversible" current expenditure. The levels of government expenditure from both the investment and consumption side were unsustainable when the boom ended. In some cases, the increase in government expenditure in the boom years was higher than the windfall!

There were four main reasons for so many unproductive investment booms, the first three of which had strong historical foundations. First, from approximately 1955 to 1980 industrialization via import substitution dominated development thinking. When the booms took place, many of the resource abundant countries were trying to diversify away from a reliance on primary production to a more modern industrial economy, with manufacturing as the leading sector. Resource price or quantity booms allowed the government to accelerate the process of import substitution far beyond the absorption capacity of the economy. Large investments, often owned and managed by the public sector, were undertaken in industry, a policy which was consistent with the pre-boom priorities. It is not a coincidence that the worst examples of failed resource boom management occurred in countries where the public sector owned the resource and received most of the windfall directly.

33

Second, countries following import substitution strategies generally had repressed financial systems in order to direct credit to the priority sectors and prevent capital outflows. When the boom took place, there were few profitable investment options that the private sector could do with its part of the windfall. Thus, a large part of the private sector windfall was either transferred to the government via high implicit taxes on domestic financial instruments or spent on durable consumer goods and construction projects.

Third, for many countries that previously were not following import substitution policies, or had very limited programs, the boom windfall was an irresistible temptation to change course. The end result was similar to, and often worse, than that suffered by countries who had long been on the import substitution path.

Fourth, political pressure on the government to spread investment across all of the regions of the country or to support failing industries was often an important factor behind the perverse investment booms.

Unproductive investment booms were partly due to inherent bad investments and partly due to too many investments at the same time. If a substantial part of the windfall had been put in foreign assets and investment was stretched out over a period substantially longer than the boom, it seems likely that many of the problems would have been avoided. If the private sector received a large part of the windfall, it was essential that profitable asset choices were available. In many cases, especially for beverage booms, the government would probably have had to channel the savings to foreign financial assets.

The biggest danger facing resource abundant countries is not that dynamic sectors atrophy but that government expenditure commitments made during boom (or good) years is not consistent with likely revenues during bust (or bad) years. This problem occurs if there are high levels of recurrent costs associated with public investments during boom years or if social welfare programs, including consumer subsidies, are not sustainable over the long term but difficult to reverse. High recurrent costs after the boom have generally led to large fiscal deficits, which in turn has led to some combination of high inflation and high levels of external debt to finance the deficit. Moreover, if the exchange rate is fixed or on a peg system, there often is a large real appreciation of the currency after the boom, which can give the appearance of Dutch disease. Policymakers would do well to be pessimistic about the length of booms as the costs of being incorrect are much lower than if they are optimistic and wrong.

Of course, bad policies are not just the result of mistakes. They are often due to the efforts of rent-seekers, both within and outside of the public sector, special interest groups, and corruption. Consequently, it is advisable to make the fiscal and monetary authorities as free from political pressures as possible.

Even for the most well-intentioned policymaker it is not an easy task to determine the optimal savings rate and period over which to invest the public sector portion of the windfall. Three rules of thumb to follow are: (i) Do not invest beyond the absorptive capacity of the economy,

34

defined as the level at which severe skill shortages would arise or there would be strong inflationary pressures on non-traded sectors. (ii) Assume that all recurrent costs associated with new investment and programs are irreversible. Be pessimistic about the length of the boom. Restrict new investments and programs to levels that are consistent with the expected long-term revenue flows. (iii) Given the inherent risks of any new projects, only invest when the expected rate of return is considerably above that which can be earned in riskless foreign assets. Be sure to use proper shadow prices, especially for foreign exchange.

Finally, if the private sector is more likely to behave in an optimal manner than the public sector, as suggested by Collier and Gunning (1996), it is best to leave the windfall in the hands of the private sector when it lands there and to redistribute towards the private sector otherwise. In contrast to the conventional wisdom that the government should be the custodian of the windfall, they argue that the private sector is more likely to do a better job. As noted previously, such policies assume that private agents can adjust their portfolios in an optimal manner. At a minimum, financial markets must be relatively free. It must be emphasized, however, that even if the private sector directly receives the entire windfall and the tax system is left untouched, a large part of the windfall is likely to end up in the hands of the public sector. Hence, the government rarely has the luxury of just sitting back and watching as nature takes its course.

To conclude, natural resource booms have often reinforced or led to poor economic policies in developing countries. It is not enough, however, for governments to do nothing, especially as the existing policy framework may not be complementary to the boom. While it may be optimal for the government to stay away from micro management of the windfall, it will usually have a strong role to play in macro management. Its main role should be to ensure the long-term health of the economy. The most important mistakes to avoid in this regard are unsustainable budget deficits and external borrowing. By either directly saving the windfall or by making it easy for the private sector to save the windfall, it can avoid these two mistakes and smooth out the effects of the boom on other sectors. In particular, policies which avoid deficits and debt will usually moderate the appreciation of the foreign exchange rate, lessening the need for adjustment in the non-boom sectors of the economy.

35

Bibliography

Auty, Richard M. (1993), Sustaining Development in Mineral Economies: The Resource Curse Thesis, Routledge, London and New York.

Auty, Richard M. (1994), "Industrial Policy Reform in Six Large Newly Industrializing Countries: The Resource Curse Thesis," World Development, v. 22(1), 11-26.

Auty, Richard and Gelb, Alan (1986), “Oil Windfalls in a Small Parliamentary Democracy: Their Impact on Trinidad and Tobago,” World Development, v. 14(9), 1161-75.

Basch, Miguel and Eduardo Engel (1991), "Shocks Transitorios y Mecanismos de Estabilización: El Caso Chileno," mimeo, Inter-American Development Bank, Washington, D.C.

Bates, Robert H. (1994), “Algunas Reflexiones sobre la Economía Política del Petróleo,” Planeación y Desarrollo, mayo, 197-203.

Bevan, David; Paul Collier; and Jan Willem Gunning (1990), "The Macroeconomics of External Shocks," WP, Oxford University.

Bevan, David; Collier, Paul; Gunning, Jan Willem (1992), “Anatomy of a Temporary Trade Shock: the Kenyan Coffee Boom of 1976-79,” Journal of African Economies, 271-305.

Bevan, David; Collier, Paul; Gunning, Jan Willem (1992), “Nigeria: Policy Responses to Shocks, 1970-1990,” Country Studies #11, ICEG, San Francisco, USA.

Bevan, David; Paul Collier; and Jan Willem Gunning (1993), “Trade Shocks in Developing Countries: Consequences and Policy Responses,” European Economic Review, v. 37, 57-65.

Bienen, Henry and Gelb, Alan (1988), “Nigeria: From Windfall Gains to Welfare Losses?” in Gelb and Associates, Oil Windfalls: Blessing or Curse?, 227-261.

Borpujari, Jitendra G. and Melhem F. Melhem (1990), "Adjustment in Major Oil Exporting Countries," Finance and Development, Sept., pp. 39-43.

Claessens, Stijn and Varangis, Panos (1994), "Oil Price Instability, Hedging, and an Oil Stabilization Fund: The Case of Venezuela," Policy Research Working Paper #1290, World Bank, Washington, D.C.

Collange, Gerald (1995),"Deux Pays Petroliers: l'Indonesie et le Nigeria," Afrique Contemporaine, No. 176, 5-17.

Collier, Paul and Gunning, Jan Willem (1996), “Policy Towards Commodity Shocks in Developing Countries,” Working Paper WP/96/84, IMF, Washington, D.C.

36

Conrad, Joseph (1976), Nostromo, Penguin Modern Classics, U.K. (First published in 1904.)

Cooper, Richard N. (1991), "Economic Stabilization in Developing Countries," International Center for Economic Growth Occasional Paper #14, San Francisco.

Corden, W.M. (1984), "Booming Sector and Dutch Disease Economics: Survey and Consolidation," Oxford Economic Papers v. 36, 359-380.

Corden, W. Max and Neary, J. Peter (1982), “Booming Sector and De-Industrialization in a Small Open Economy,” The Economic Journal, v. 92, 825-48.

Cuddington, John (1989), "Commodity Export Booms in Developing Countries," The World Bank Research Observer, v.4(2), 143-65.

Davis, Graham A. (1995), "Learning to Love the Dutch Disease: Evidence from the Mineral Economies," World Development, v. 23(10), 1765-79.

Davis, Jeffrey. "The Economic Effects of Windfall Gains in Export Earnings, 1975-1978," World Development 11:2 (1983), 119-139.

Fardmanesh, Mohsen (1991), "Dutch Disease Economics and the Oil Syndrome: An Empirical Study," World Development, v. 19(6), pp. 711-717.

Gelb, Alan and Associates (1988), Oil Windfalls: Blessing or Curse?, Oxford University Press and the World Bank.

Harberger, Arnold (1994), “Lecciones de Experiencias con Bonanzas Petroleras,” Planeación y Desarrollo, mayo, 191-6.

Harrison, P. (1994), “The Impact of Oil on Trinidad and Tobago, 1966-1990,” Working Paper Series No. 171, Institute of Social Studies, The Hague, Netherlands.

Harvey, Charles (1992), "Botswana: is the Economic Miracle Over?" Journal of African Economies, v. 1(3), 335-68.

Hill, Catherine B. and Mokgethi, D. Nelson (1989), "Botswana: Macroeconomic Management of Commodity Booms, 1975-86," Successful Development in Africa: Case Studies of Projects, Programs, and Policies, EDI Development Policy Case Series, Analytic Case Studies #1, World Bank, Washington, D.C.

Kamas, Linda (1986), "Dutch Disease Economics and the Colombian Export Boom," World Development, v. 14(9), 1177-1198.

Knudsen, Olin and John Nash (1990), "Domestic Price Stabilization Schemes in Developing Countries," Economic Development and Cultural Change, v. 38(3), 539-558.

37

Knudsen, O. and Parnes, A. (1975), Trade Instability and Economic Development, D.C. Heath and Co., Lexington, Mass.

Larson, Donald F. and Coleman, Jonathan (1991), "The Effects of Option-Hedging on the Costs of Domestic Price Stabilization Schemes," PRE Working Paper WPS 653, World Bank, Washington, D.C.

Laszlo, E.; Baker, R.; Eisenberg, E.; and Raman, V. (1978), The Objectives of the New International Economic Order, Pergamon Press, New York, USA.

MacBean, A. (1968), Export Instability and Economic Development, George Allen and Unwin, Ltd., London, U.K.

Meier, Gerald and Seers, Dudley (1984), Pioneers in Development, Oxford University Press, Oxford, U.K.

Norberg, Helene and Blomstrom, Magnus (1993), “Dutch Disease and Management of Windfall Gains in Botswana,” in Blomstrom, Magnus and Mats Lundahl (eds), Economic Crisis in Africa: Perspectives on Policy Responses, Routledge, London and New York, pp. 162-178.

Nyatepe-Coo, Akorlie A. (1994), "Dutch Disease, Government Policy and Import Demand in Nigeria," Applied Economics, v. 26, 327-36.

Prebisch, Raúl (1984), “Five Stages in My Thinking on Development,” in Meier and Seers (eds), Pioneers in Development, 175-204.

Sachs, Jeffrey D. and Warner, Andrew M. (1995), “Natural Resource Abundance and Economic Growth,” Development Discussion Paper No. 517a, HIID, Cambridge, Mass, USA.

Singer, Hans (1984), “The Terms of Trade Controversy and the Evolution of Soft Financing: Early Years in the UN,” in Meier and Seers (eds), Pioneers in Development, 273-312.

Wunder, Sven (1992), "La Enfermedad Holandesa y el Caso Colombiano," Coyuntura Económica, 167-190.

38