POVERTY REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise 53164 FEBRUARY 2010 - Number 1 Natural Resources and Development Strategy after the Crisis Milan Brahmbhatt and Otaviano Canuto, Poverty Reduction and Economic Management (PREM) Recent events have rekindled interest If we view developing countries as a in the role of primary commodities in single aggregate then only about 40 per- development. Was the boom in com- What policies can help cent of their merchandise exports were modity prices from around 2003 through poor countries best primary commodities by value in 2005­ 2008 just a cyclical event, or does it sug- 07, down from around 50 percent in the gest that prices have entered on a period manage commodity early 1990s. This aggregate measure can of secular strength, driven by factors resources for long-run be misleading, however, because it is such as demand in big, fast-growing de- development? dominated by a few big economies like veloping countries like China? It is no- China that are almost entirely exporters table that, while commodity prices fell of manufactures. sharply from their peak in 2008 with the onset of the global recession, they generally remained much higher than previous recession lows, often as high as in 2005­07, a period of robust world growth. Furthermore, prices have also rebounded smartly over the course of 2009 (figures 1 and 2). If a period of sustained commodity strength is immi- nent, what are the implications for development policies? Development economists have long debated the problems associated with the traditionally high specialization in pro- duction and export of primary commodities of most devel- oping countries. Many argue that dependence on primary commodities has proved to be a poisoned chalice or curse for development, which, given this view, necessarily en- tails structural change and rapid industrialization. Others, however, suggest that sustained high commodity prices could reduce the relevance of an industrialization-focused development strategy for commodity-dependent, low-in- come countries (LICs).1 In this note we briefly review four questions: How dependent are developing countries on primary commodity exports? What is the outlook for pri- A different picture emerges if we take a simple average mary commodity prices? Is there a natural resource across developing countries (that is, giving each country "curse" (or blessing)? What policies can help poor coun- an equal weight). Commodities still comprised a little over tries best manage commodity resources for long-run de- 60 percent of the merchandise exports of the average de- velopment? veloping country in the middle part of this decade, al- though this was down from over 90 percent in the late How dependent are developing countries on 1960s. primary commodity exports? POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Looking at the median, half of developing countries still parently confirming the Prebisch-Singer hypothesis. have commodity export dependence of over 70 percent. However, it is now well understood that attempts to as- Among LICs, commodity export dependence averages sess long-run trends on the basis of visual inspection and around 75 percent. Viewed by regions, Africa, Latin Amer- simple time series models can be mis-leading, especially if ica and the Caribbean, and the Middle East and North Af- the series in question are so-called unit root processes. In rica are the most commodity dependent, while South Asia, this case processes without any deterministic trend can East Asia, and Europe and Central Asia are the least (figure yield apparently significant but actually spurious regres- 3). So, although declining, commodity or natural resource sion results. Cuddington et al. (2007) carefully survey dependence remains a fact of life for a majority of devel- econometric studies of the Grilli-Yang series through oping countries. 1998. Their overall conclusion is that, although there is clear evidence of a structural break in 1921, it is not pos- What is the outlook for primary commodity prices? sible to reject the unit root hypothesis for real commodity In the 1950s the famous Prebisch-Singer thesis argued prices. There is also no evidence of drift, either positive or that real primary commodity prices (for example, relative negative. We find essentially the same results for the Gril to manufactures prices) displayed a long-run declining li-Yang series over the period 1900­2008. trend. Faced with a resulting steady decline in their terms So, based on statistical properties alone, we have little of trade, developing countries should foster industrializa- tion, following, according to the thinking of the time, an import substitution strategy. During the commodity price spike of the 1970s, on the other hand, many analysts argued that permanent natural resource scarcity would result in steadily rising real commodity prices. Based on econometric study of long time series, the present consensus appears to be that real commodity prices do not display any permanent trend or drift over time. Fig- ure 4 shows the Grilli and Yang time series of real non-en- ergy commodity prices (updated by other researchers) for the period 1900­2008. (Grilli and Yang, 1988; Pfaffen- zeller et al., 2007). The series is a weighted index of the nominal prices of 24 non-energy commodities, divided by an index of the unit values of manufactured goods exported from developed to developing countries. Visual inspection of figure 4 suggests a definite downward trend, and this ap- pears to be confirmed by regression of the log of the Grilli- Yang series on a deterministic time trend (modeling the er- ror process as a first-order AR1 process) over the whole period 1900­2008, which yields an estimate that real com- modity prices fall on average by 0.5 percent per year, ap- POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise 2 reason to expect real commodity prices to trend either up monetary conditions lax, as at present, since inventory car- or down in the long run. It is a feature of unit root pro- rying costs are low and there is more incentive to leave cesses, however, that series with this property are highly depletable natural resources in the ground. In the longer correlated over time. So it is quite possible for commodity term, however, general price levels and real interest rates prices to move significantly lower or higher for substantial are low and monetary conditions lax, as at present, since periods even in the absence of a long-run trend or drift, inventory carrying costs are low and there is more incen- such as, for example, the long period of unusually low tive to leave depletable natural resources in the ground. In prices from the mid-1980s through the 1990s. Again, the longer term, however, general price levels and real in- based on statistical properties alone, one would not be sur- terest rates can be expected to rise, removing the over- prised to see a sustained period of high prices following shooting in real commodity prices. the low prices of the 1980s and 1990s. So there are both supply and demand factors that could Are there plausible fundamental economic factors to support the present relatively high level of real commodity support such an outlook? The price of commodities rela- prices in the medium term, although these factors will tend tive to the price of manufactures can be usefully analyzed to dissipate in the longer term. Current World Bank fore- in terms of demand and supply: the demand for primary casts are consistent with this scenario, projecting only a commodities relative to the demand for manufactures, and gradual easing in real commodity prices from existing lev- the supply of commodities relative to the supply of manu- els by 2015. Forecast real prices in this period are in fact factures. squarely in the range that prevailed from the 1920s through On the supply side, if long-term productivity growth in the early 1980s (figure 4). If correct, this means that com- agriculture and minerals is less than in manufacturing modity exporters are likely to face a more benign medium- then, other things equal, one would expect agricultural and term price environment than in the 1980s and 1990s. mineral prices to rise relative to those of manufactures. But there is little evidence to suggest that productivity Is there a natural resource "curse" (or blessing)? growth in commodities sectors is significantly different The short answer is "no" or rather "it depends." A sur- from that in manufactures, so this is unlikely to influence vey of the large and rapidly growing empirical research in relative prices either way (World Bank, 2009). It is true, this area suggests that, in the words of a recent World Bank however, that investment in new capacity in energy and report, natural resources are "neither curse nor destiny" minerals was cut substantially when prices were low in the (Lederman and Maloney, 2007). Studies of the relation- 1980s and 1990s and is recovering only slowly due to skill ship between natural resource abundance and growth have, shortages, technical difficulties in developing new reserves however, often tended to generate disparate and some- (for example, deep offshore), and political uncertainty in times contradictory results. The influential study by Sachs regions with new reserves. Bio-fuel subsidies have also and Warner (1995) is representative of results which find helped switch grain acreage away from food to fuel use, that natural resource abundance has a strong negative im- providing a major reason for the steep grain price hikes pact on growth. Lederman and Maloney (2007), on the from 2005 through the early part of 2008. Over the longer other hand, challenge the Sachs and Warner findings on term, though, one would expect a more copious supply re- measurement and econometric grounds and find natural sponse, as skill shortages and technical difficulties are resource abundance to have a positive effect on growth. overcome, and new reserves and acreage are brought into A recent effort to reconcile such apparently disparate production. research findings (Collier and Goderis, 2007) observes Relative demand for commodities could also rise in the that, first, negative long-run growth effects are mostly re- medium term to the extent world growth after the financial lated to oil and minerals--concentrated "point source" re- crisis is more dependent on developing countries and de- sources that can easily become the object of rent-seeking mand in these countries is more commodity intensive than and redistributive struggles (including armed conflict). On elsewhere. In the longer term, however, production pro- the other hand, there is little evidence of negative growth cesses in developing countries will continue to become effects related to high prices for agricultural commodities, more efficient in terms of raw material consumption, ap- which are generally more open to competitive entry. Sec- proaching closer to developed country levels, while rela- ond, high oil and mineral prices mostly have a negative tive final demand for commodities like food will continue impact on long-run growth in exporting countries with bad to decline due to low income elasticity relative to things governance. They have a significant positive impact on like services. There is also evidence that real commodity growth in exporters with good governance. This finding prices are affected by monetary conditions (Frankel, suggests that continued high commodity prices in the next 2008). Since commodities are traded in flexible price mar- few years could provide valuable resources to accelerate kets, their prices tend to overshoot in response to monetary economic and social development in commodity export- changes relative to general manufactures and services ing countries with good policies and governance. prices, which adjust more sluggishly. Commodity prices There are several considerations to keep in mind when will tend to be high when real interest rates are low and evaluating the ways in which natural resource abundance POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise 3 can lead to worse economic performance, especially under consumption. When prices plunged in the 1980s these conditions of poor governance. countries were left with balance of payments crises and First, because of political economy reasons, countries unsustainable external debt levels. Again, it is critical to with weak governance are more likely to adopt poor eco- note that the actual extent of Dutch Disease effects, vola- nomic policies to manage commodity booms, contributing tility, and over-borrowing will depend to a large extent on to significant misallocation and mismanagement of re- policies--for example, on the extent to which cautious fis- sources. For example politicians may expand public cal policies are able to moderate aggregate demand pres- spending and employment excessively and too rapidly, sures, smooth volatility in government revenues, and curb with the aim of increasing their patronage networks and external over-borrowing. improving their chances of staying in power, while re- Lastly, in addition to problems of short-run economic sources shift out of productive activity into unproductive management, natural resource­abundant countries also rent-seeking activity (Mehlum et al., 2006; Robinson et face important longer-run questions about the optimal al., 2006.) Poor fiscal policy indeed appears to be at the pace at which to deplete their resume today and how much heart of economic mismanagement in the wake of natural to save for the welfare of future generations. An important resource booms. Studying natural resource boom episodes metric here is whether the country's economic strategy is in the 1970s and 1980s Gelb (1988) concluded that "the sustainable, meaning one that transfers sufficient capital to most important recommendation to emerge from this study future generations to allow them to achieve at least the is that spending levels should have been adjusted to sharp same level of welfare as current generations.3 From this rises in income levels more cautiously than they actually perspective natural resources can be viewed as part of a were."2 country's overall capital stock, alongside its physical capi- Second, natural resource booms create complicated tal stock (such as existing machinery and buildings) and problems in macroeconomic management that are chal- intangible capital (including human capital, social capital, lenging even in economies with good governance and ca- and other factors such as the quality of its institutions). To pable institutions, and much more so in economies without increase its overall capital stock, a country's investment in these advantages. One of these problems is the so-called its physical, human, and other capital must be larger than Dutch Disease effect: increased domestic income from the the depreciation of its physical and other capital, including booming natural resource sector generates higher spend- the depletion of its natural resources. This measure of ing on domestic goods (as well as imports), leading to countries' adjusted net savings rates is shown on the verti- higher prices and output in the non-tradables sector. Wages cal axis of figure 5. The horizontal axis shows countries' in the economy also tend to rise, squeezing profits in sec- annual depletion of their natural resources (principally oil tors of the economy that are internationally tradable but and minerals, together with a measure of forest depletion). which are not based on natural resources, such as manu- Figure 5 suggests that countries with high rates of natu- facturing, where prices are largely fixed at international ral resource depletion are often on unsustainable develop- levels. With increased inflation in non-tradables prices ment paths: they are not saving enough to cover the deple- there is an appreciation of the real exchange rate and an tion of their natural resources, resulting in negative output contraction in non-resource tradables sectors like adjusted net savings rates. manufacturing. These adjustments are of concern if one believes that sectors like manufacturing have some special What policies can help poor countries best manage characteristics that stimulate long-run growth, for example commodity resources for development? increasing returns to scale, learning by doing, or abundant First, given the evidence that problems with governance technological spillovers. Evidence that manufacturing are at the root of economic problems associated with natu- possesses these special characteristics is mixed, but there ral resource abundance, efforts to enhance transparency is fairly robust evidence for the more general proposition and strengthen checks and balances concerning all aspects of a negative relation between real exchange rate over- of natural resource extraction and use are clearly vital for valuation and growth (Rodrik, 2007; Aguirre and Calder- ensuring accountability. These aspects include the terms of on, 2005.) contracts with companies engaged in resource extraction There are also problems because volatility of primary or operation, ongoing monitoring of operations, and the commodity prices and revenues can drive volatility in collection and use of government taxes and other revenues government spending and real exchange rates, with the re- from natural resources. Broad global efforts like the Ex- sulting uncertainty damaging investment and growth. An- tractive Industries Transparency Initiative can play a part, other related way in which commodity price volatility may as, at the domestic level, can anti-corruption reforms, mea- affect growth is by fostering over-borrowing. High com- sures to improve transparency and scrutiny by civil society modity prices in the 1970s encouraged many resource- and media, procurement reforms, strengthening of formal abundant countries to borrow heavily from abroad, to fi- audit, parliamentary scrutiny, and so on. Equitable sharing nance large investment projects and high public of benefits across regions, ethnic groups, and so forth can POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise 4 ing foreign debt with the rates of return available on various kinds of investment, for example the return on foreign assets, the return from reducing foreign debt (not generally the same thing in developing countries), and returns to domestic public invest- ments. A commonly used benchmark for fiscal pol- icy in a natural resource­rich economy is the per- manent income rule. Under this rule the country should save all resource revenues over and above a certain permanently sustainable increase in the lev- el of consumption, which is equal to the annuity value of the country's natural resource wealth.4 In practice the rule often leads to a recommendation to establish a Natural Resource or Sovereign Wealth Fund that invests in foreign assets, the returns from which can support spending on the government's non-natural-resource fiscal budget. It is worth noting that the permanent income approach addresses several of the key issues associated with natural help reduce the danger of civil strife over resources. resource fiscal management. It is by definition a sustain- An institutional innovation that has attracted much able policy that converts a temporary, exhaustible stock recent attention is the use a separate (extra-budgetary) of natural resources into a stock of financial assets that Natural Resource Fund (NRF) to facilitate good manage- generates a permanent income stream. Since the policy ment of revenues. However, experience suggests that the calls for saving a substantial proportion of natural resource establishment of such funds is no substitute for sound revenues, it reduces the pressure of rising domestic de- overall fiscal and economic management, although in cer- mand that leads to real exchange rate appreciation and tain circumstances it may buttress the right policy mix. Dutch Disease effects. By smoothing expenditures, the While Natural Resource Funds are sometimes created to policy also moderates the problems caused by volatility in protect resource revenue from political pressure and po- natural prices and revenues. tential waste and corruption, and this argument has its There is nevertheless something anomalous about view- merits, an NRF of itself will not prevent such waste and ing the permanent income rule as a long-run development abuse unless it is part of a broader effort to strengthen gov- strategy, with poor capital-scarce countries financing in- ernance and integrate the fund with an overall fiscal policy vestments in rich countries through sovereign wealth framework. funds. Several analysts have recently argued that the per- Second, attention also needs to be paid to the actual manent income rule is optimal only under special circum- substance of economic policy decisions about the alloca- stances that do not apply to most developing countries; tion of natural resource revenues between consumption essentially these conditions are the ability to freely borrow and savings of various kinds. These decisions will help and lend at the world rate of interest, which would result in determine how well the country is able to handle the mac- foreign and domestic rates of return becoming aligned ro management problems associated with natural resource (Collier and Venables, 2008; Van der Ploeg and Venables, abundance, such as the Dutch Disease and commodity 2009). Most developing countries, however, are character- price volatility, as well as the impact of natural resources ized by restricted access to world capital markets, capital on the country's longer-run growth and poverty reduction scarcity, and potentially high rates of return on domestic efforts. Figure 6 provides a schematic of basic choices investment, especially if the government is able to effi- open to the government, for example whether to return ciently supply scarce public infrastructure and to improve revenu revenues to private citizens (via tax cuts or trans- the investment climate so as to raise returns on private in- fers, which will then be reflected in increased private con- vestment. Under these circumstances a more optimal strat- sumption and investment), or to retain resource revenues egy would be to devote a larger portion of resource reve- in public hands, which then need to be allocated between nues to high-return public domestic investments, leading public consumption and various kinds of public invest- to higher growth and, ultimately, a higher value of con- ment (or net asset accumulation). sumption than under the permanent income strategy. At a very general level these decisions need to be guid- Evidently, much of the success of a strategy oriented ed by a comparison of the government's social discount more toward domestic investment will depend on how ef- rate (which measures the value it puts on consumption to- ficiently public investment funds can be allocated and day versus consumption at later dates) with the rates of managed to achieve high returns in practice. So, thirdly, return available on various kinds of investment, for ex- reforms to strengthen public investment management, cost ample the return on foreign as sets, the return from reduc POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise 5 benefit analysis, monitoring and evaluation, and budget evidence for this hypothesis is mixed. processes and institutions provide another crucial element 3. Heal (1996) discusses alternative interpretations of of a successful natural resource­based development strat- sustainability. egy. To the extent that it will take time to develop a pipe- 4. The permanent income approach to fiscal policy in line of good projects and to strengthen public investment natural resource­abundant economies is studied in more management capacity, it may be prudent for the country to detail in Van Wijnbergen (2008), Davis et al. (2003), and initially continue to invest most of its revenues in foreign Barnett and Ossowski (2002). assets, but to then increase the proportion invested domes- tically, in line with domestic absorptive capacity. The About the authors country will also likely continue to flexibly shift resources Milan Brahmbhatt is an Economic Advisor and Otaviano Ca- into or out of its natural resources fund, depending on the nuto is the Vice President and Head of the Poverty Reduction need to address price volatility and Dutch Disease type and Economic Management (PREM) Network. pressures. However, investment climate reforms, support for innovation, and high return domestic infrastructure in- vestments can also help alleviate Dutch Disease pressures by increasing the supply capacity of the economy. We conclude that booming commodity revenues raise difficult challenges that, if not adequately addressed, can harm long-run development. However, with good policies, governance, and management, such revenues can also be a valuable resource that helps accelerate overall economic and social development. Notes 1. See, for example, Oxford Analytica International, 2009, "Commodities Force Re-think on Growth," August 18. 2. A stronger version of the political economy channel argues that natural resource booms can even lead to a worsening of governance, for example a "voracity effect" as political actors race to seize and spend natural resource revenues before others do, provoking more intense politi- cal, bureaucratic, and even violent conflicts for control of natural resource revenues (Tornell and Lane, 1999). The POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise 6 References Mehlum, Halvor, Karl Moene, and RagnarTorvik. 2006. "Institu- tions and the Resource Curse." Economic Journal 116(508): Aguirre, Alvaro, and Cesar Calderon. 2005. "Real Exchange 1­20. Rate Misalignments and Economic Performance." April. Central Bank of Chile Economic Research Division. Pfaffenzeller, Stephan, Paul Newbold, and Anthony Rayner. 2007. "A Short Note on Updating the Grilli and Yang Commod- Barnett, Steven, and Rolando Ossowski. 2002. "Operational As- ity Price Index." World Bank Economic Review 21(1): 151­163. pects of Fiscal Policy in Oil-Producing Countries." IMF Working Paper WP/02/177. IMF, Washington, DC. Robinson, James A., Ragnar Torvik, and Thierry Verdier. "Politi- cal Foundations of the Resource Curse." Journal of Development Collier, Paul, and Benedikt Goderis. 2007. "Commodity Prices, Economics 79(2): 447­68. Growth and the Natural Resources Curse: Reconciling a Conun- drum." August. Working Paper 276. Centre for the Study of Afri- Rodrik, Dani. 2007. "The Real Exchange Rate and Economic can Economies. Growth: Theory and Evidence." August. John F. Kennedy School of Government, Harvard University, Cambridge, MA. Collier, Paul, and Anthony Venables. 2008. "Managing Resource Revenues: Lessons for Low Income Countries." Paper for the Sachs, Jeffrey D., and Andrew M. Warner. 1995. "Natural Re- Africa Economic Research Consortium 2008 Annual Confer- source Abundance and Economic Growth." NBER Working Pa- ence. per 5398. National Bureau of Economic Research, Cambridge, MA. Cuddington, John T., Rodney Ludema, and Shamila A. Jayasuri- ya. 2007. "Prebisch-Singer Redux." In Daniel Lederman and Tornell, Aaron, and Philip R. Lane. 1999. "The Voracity Effect." William F. Maloney, eds., Natural Resources: Neither Curse Nor American Economic Review 89(1): 22­46. Destiny. Stanford University Press and the World Bank. Van Der Ploeg, Frederick, and Anthony Venables. 2009. "Har- Davis, J.M., R. Ossowsky, and A. Fedilino, eds. 2003. Fiscal nessing Windfall Revenues: Optimal Policies for Resource-Rich Policy Formulation and Implementation in Oil Producing Coun- Developing Economies." Working Paper No. 2571. CESifo tries. Washington, DC: IMF. Group, Munich, Germany. Frankel, Jeffrey. 2008. "The Effect of Monetary Policy on Real Van Wijnberger, Sweder. 2008. "The Permanent Income Ap- Commodity Prices." In John Campbell, ed., Asset Prices and proach in Practice." Unpublished paper. March. PREM Network, Monetary Policy. University of Chicago Press. World Bank, Washington, DC. World Bank. 2009. Global Eco- nomic Prospects and the Developing Countries. Washington, Gelb, Alan and Associates. 1988. Oil Windfalls: Blessing or DC: World Bank. Curse? World Bank and Oxford University Press. World Bank. 2009. Global Economic Prospects and the Devel- Grilli, Enzo R., and M.C. Yang. 1988. "Primary Commodity oping Countries. Washington, DC: World Bank. Prices, Manufactured Goods Prices, and the Terms of Trade of Developing Countries: What the Long Run Shows." World Bank Economic Review 2(1): 1­47. Heal, Geoffrey.1996. "Interpreting Sustainability." Paine Webber Working Paper Series in Money, Economics and Finance PW- 95-24. May. Columbia Business School, New York, NY. Lederman, Daniel, and William F. Maloney, eds. 2007. Natural Resources: Neither Curse nor Destiny. World Bank and Stanford University Press. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise 7