Chapter 5. Managing Oil Windfalls in the CEMAC

Bernardin Akitoby, and Sharmini Coorey
Published Date:
August 2012
  • ShareShare
Show Summary Details
Frederick van der Ploeg

Natural resource rents worldwide exceed $4 trillion annually, amounting to some 7 percent of global GDP. Nonrenewable resource revenue is a dominant feature of 50 economies with a combined population of 1.4 billion. In 24 countries, resources make up more than three-quarters of exports. Resources account for at least 40 percent of GDP in 13 countries and more than half the fiscal revenue in 18 countries (IMF data 2000–05). Some countries (e.g., Botswana, Chile, and Malaysia) have grown fast on the basis of this revenue, but others (Nigeria and Cameroon, for instance) have not, and have been labeled as countries suffering from the “resources curse.” Turning attention to the Central African Economic and Monetary Community (CEMAC), Table 5.1 shows oil revenue as a percentage of GDP for each of the member countries.

Table 5.1Oil Revenue in the CEMAC Countries(Percent of GDP)
Central African Republic0.
Congo, Republic of20.321.018.920.821.631.837.932.140.120.731.837.435.834.129.927.124.8
Equatorial Guinea15.322.522.623.226.431.737.533.934.637.125.925.924.522.120.618.616.9
Note: CEMAC = Central African Economic and Monetary Community.
Note: CEMAC = Central African Economic and Monetary Community.

Oil revenue as a share of GDP ranges from zero for the Central African Republic to about 17 percent for Gabon, 23 percent for Chad, and 37 percent for the Republic of Congo in 2011. Figure 5.1 indicates that CEMAC oil production rose until 2005, dropped slightly through 2011, and is projected to fall rapidly during the next two decades. As a result of this oil depletion, Table 5.1 shows that CEMAC oil revenue as a percentage of GDP is anticipated to fall during the coming years, in sharp contrast to a country such as Iraq, whose oil revenue is projected to last well beyond the foreseeable future. From the point of view of the permanent-income hypothesis, Iraq should consume all of its oil windfall every year, while the countries of the CEMAC should save a certain portion to generate a sufficient return on sovereign wealth to finance a sustained increase in consumption. From a development perspective, however, investing in sovereign wealth (e.g., U.S. Treasury bills with a low rate of interest) seems unlikely to be better than using the oil revenue to pay off foreign debt (which typically carries a higher interest rate) or invest in domestic infrastructure, education, or health projects (with typically a much higher rate of return).

Figure 5.1CEMAC Oil Production

Sources: IMF African Department database; and IMF staff estimates.

Note: CEMAC = Central African Economic and Monetary Community.

Figure 5.2 shows that the CEMAC’s near-term economic prospects are not particularly good despite recent reserves accumulation. GDP growth is projected to be flat at very modest levels, and the share of investment in GDP gradually falls. The predictions for the CEMAC region suggest that very little of the windfall, which is forecast to continue until 2015–16, will be used for investment; instead, it will be used to maintain historically high government spending while keeping the burden of foreign debt as a percentage of GDP fairly low. Trade surpluses are expected to fall as oil revenue flattens out. Despite projected surpluses for the trade balance and falling foreign debt, the current account will show modest deficits. Government budgets are also expected to stay in modest surplus.

Figure 5.2Macroeconomic Characteristics of the Central African Economic and Monetary Community

Sources: IMF, International Financial Statistics; IMF, World Economic Outlook, April 2011; and IMF staff estimates.

The crucial question is whether the CEMAC countries could have harnessed their oil windfalls better and whether they can do so more successfully as their streams of oil revenue decline. After all, controlling for initial income per capita, investments in physical and human capital, trade openness, and rule of law, natural resource dependence (measured by the ratio of natural resource exports to GDP) has a strong and significant negative effect on growth of GDP per capita (Table 5.2, column 1). All things equal, an increase in the ratio of resource exports to GDP of 10 percentage points depresses average growth of GDP per capita by 1.1 percent per year. However, empirical evidence supports the hypothesis that countries with good institutions receive a positive growth effect from resource dependence, while those with bad institutions are adversely affected. Increasing the ratio of natural resource exports to GDP by 10 percentage points increases average growth by a mere 0.1 percent per year in countries with good institutions (a weighted index of various indicators measured on a scale from zero to one) but decreases annual growth by 1.43 percent in countries with bad institutions (Table 5.2, column 2). The very small effect of windfalls on growth if institutions are good suggests that the effect of institutions on growth may well be asymmetric—windfalls hurt economic performance if institutions are poor but do not necessarily improve performance if institutions are good. It has also become clear that volatile commodity prices are a key driver of the natural resource curse; natural resource dependence, unrestricted international capital flows, being landlocked, and ethnic tensions boost the volatility of per capita growth and thus depress growth prospects (van der Ploeg and Poelhekke, 2010).1

Table 5.2Cross-Country Empirical Evidence for the Natural Resource Curse(Annual growth in real GDP per capita)
Sachs and Warner (1997)Mehlum, Moene, and Torvik (2006)
Initial income–1.76 (8.56)–1.26 (6.70)
Openness1.33 (3.35)1.66 (3.87)
Resource dependence–10.57 (7.01)–14.34 (4.21)
Rule of law0.36 (3.54)n.a.
Institutional qualityn.a.–1.3 (1.13)
Investments1.02 (3.45)0.16 (7.15)
Interaction termn.a.15.40 (2.40)
Number of countries7187
Adjusted R20.720.71
Source: Author’s compilation.Note: n.a. = Not available.
Source: Author’s compilation.Note: n.a. = Not available.

The fundamental problem faced by the oil-rich economies of the CEMAC is how to transform their oil wealth into a portfolio of other assets—human capital (education and health), domestic physical capital (both private and public), and perhaps foreign financial assets—that yields a continuing and growing flow of income to their citizens. The World Bank’s (2006) estimates of adjusted net saving or genuine saving2 provide a measure of the extent to which many oil-rich countries have failed to do this, and the CEMAC countries do not appear to be an exception: the oil-rich Cameroon, Republic of Congo, and Equatorial Guinea have saving rates of –2.5 percent, –14.6 percent, and –35.7 percent, respectively; the Central African Republic has no oil, but has a slightly positive genuine saving rate of 0.5 percent (World Bank, 2006). This suggests that the CEMAC countries were not effectively transforming their oil wealth into productive assets. A country successfully converting its oil reserves into physical, human, or financial capital or sovereign wealth does not run down its genuine natural wealth. The projections presented in Figure 5.1 suggest that the genuine saving rate may have become even more negative in recent years. Indeed, negative genuine saving rates have been persistent for large parts of sub-Saharan Africa, indicating that these countries have been running down their national wealth.

The present analysis offers a theoretical framework for addressing the question of how the oil-rich CEMAC countries can successfully transform their oil wealth into productive assets and harness their oil windfalls for economic development.3 The chapter first discusses the benchmark that corresponds to the orthodox view of putting all the revenue from diminishing oil wealth into a sovereign wealth fund. It then argues that investment should be made in the domestic economy rather than exclusively in a sovereign wealth fund, because many developing economies face the twin problems of capital scarcity and declining efficiency as public investment in physical and human capital (i.e., infrastructure and health and education) is ramped up. The argument is then refined by making the case for “investing to invest,” which is necessary for overcoming the absorption problems resulting from Dutch disease and the difficulty of shunting private and public capital from the traded sectors to the nontraded sectors to release capital and other factors of production to make expansion of the latter sectors feasible. Because absorption problems may be most acute in the short run, temporarily parking some of the oil revenue in a sovereign wealth fund might be wise until the economy has sufficient capacity to satisfy the increased demand for investment and consumption goods. Finally, the chapter summarizes a brief policy conclusion.

Benchmark: Put Oil Revenue in a Sovereign Wealth Fund

The orthodox policy view has been to accumulate the windfall in a sovereign wealth fund and live off the interest from the fund afterward (e.g., Barnett and Ossowski, 2003). This view gives rise to the bird-in-hand policy, which presumes no increase in consumption ahead of the windfall and a gradual buildup of consumption during the windfall. After the windfall, the sovereign wealth fund is gradually depleted, which leads to a withering away of the earlier increases in consumption. If future oil revenue can be used as collateral for borrowing, the permanent-income policy comes into play, whereby borrowing occurs ahead of the windfall, then during the windfall the debt is paid off and sovereign wealth is accumulated sufficient to sustain a permanent increase in consumption. Although the bird-in-hand policy is often advocated on grounds of prudence, the volatility of consumption under that policy compared with under the permanent-income policy leads to large welfare losses.

To illustrate the permanent-income rule and determine how much of the oil windfall to save and how much to consume, the analysis supposes that households receive exogenous production income Y and government transfers or citizen dividends T and have no access to the international capital market, so that their consumption is C = Y + T. All foreign assets A are held by the government and earn a return equal to the world interest rate r. The economy’s budget constraint is thus given by

in which N is the size of the exogenous oil windfall. The first part of equation (5.1) says that the current account must equal the increase in assets of the nation; the second part indicates that the government surplus equals the increase in government assets. The size of the sovereign wealth fund of this economy is thus given by A. The present value of the oil windfall plus sovereign wealth should cover the present value of government transfers. Alternatively, we have

where the permanent values of production income and the oil windfall are given by, respectively, Yp(t)rter(st)Y(s)dsandNp(t)rter(st)N(s)ds. In situ oil wealth is NP/r. Private utility is

where ρ = r > 0 is the rate of time preference and σ > 0 is the coefficient of intertemporal substitution. The government chooses the time paths of government transfers and saving to maximize private utility (5.3) subject to the budget constraints (5.1) and (5.2). This yields optimal consumption and government transfers

Government transfers, and thus private consumption, respond to the permanent component (i.e., the annuity value of the stream of present and future natural resource revenue) and not the actual component of the oil windfall. This is achieved by borrowing ahead of a windfall, saving during the windfall, and using the interest on the accumulated sovereign wealth fund to finance the sustained increase in transfers and consumption after the windfall. The government surplus and current account thus respond to the temporary component of the windfall, Ȧ = NNP + YYP. The nonwindfall primary deficit (NȦ) is driven by the permanent component of the windfall.

If oil revenue declines exponentially at an annual rate of 10 percent and r = 0.025 (as indicated by the “Windfall” line in the left panel of Figure 5.3), one-fifth of the windfall is the permanent component, which is consumed (“Consumption” in the left panel of Figure 5.3), and four-fifths is saved in sovereign wealth and shows up in temporary current account surpluses. It can also be seen that as oil wealth is depleted (“In situ resource wealth” in the right panel of Figure 5.3 or the falling permanent oil income line indicated by “Permanent oil income” in the left panel of Figure 5.3), the permanent-income rule builds up corresponding sovereign wealth (right panel of Figure 5.3) to leave total wealth unaffected. This ensures a sustained increase in citizens’ consumption.

Figure 5.3Permanent-Income Prescription for a Temporary Windfall

Source: Author’s calcluations.

The orthodox bird-in-hand and permanent-income policy rules are popular and may be suited to the interests of mature, oil-rich economies, but they are wholly unsuitable for the countries of the CEMAC. In developing economies, the separation theorem, which states that domestic investment decisions should be completely independent of domestic saving and windfalls of foreign exchange, is simply not valid because these countries face capital scarcity and are not well integrated into world financial markets (van der Ploeg and Venables, forthcoming). Developing economies thus need to depart from the partial equilibrium framework and confront the arduous task of harnessing oil revenue for growth and development. Before turning to that topic, the chapter discusses the problems of capital scarcity and the absorption problems that result from the inevitable inefficiencies that arise when ramping up public investment in physical and human capital in developing economies.

Two Obstacles: Capital Scarcity and Inefficiencies From Ramping up Public Investment

The CEMAC countries face two big obstacles to economic development, and any strategy for harnessing oil windfalls must tackle these obstacles. First, these countries are badly integrated into international capital markets. Borrowing for domestic investment typically requires payment of a much higher interest rate than the world interest rate. This is a pity, because many public investment projects, in both physical infrastructure and human capital such as education and health, could generate a much higher rate of return than the world interest rate. To capture the effects of this obstacle, the analysis draws on previous empirical work that finds that the interest premium increases with the degree of capital scarcity and decreases with the ability to pay as measured by the size of the oil windfall (Akitoby and Stratmann, 2008). A semi-elasticity of the natural log of the spread is estimated with respect to the debt-to-GNI ratio of 1.9. This implies that a 10 percentage point increase in the debt-to-GNI ratio pushes up the interest differential by 6.9 percentage points if the economy starts out with a debt-to-GNI ratio of 100 percent (or 1.3 percentage points if it starts off with no foreign debt). To the extent that oil income is part of GNI, it becomes easier for the country to service its debt, so creditworthiness will be higher.4 This empirical finding suggests that it is necessary to depart from the permanent-income policy by increasing consumption less strongly, using the remainder of the windfall to alleviate capital scarcity and gradually boost investment in the domestic economy, and thus speed up the process of economic development (van der Ploeg and Venables, 2011b).5 The section on capital scarcity uses these ideas to examine the optimal public investment paths after an oil windfall in a framework that also deals with the second obstacle.

The second obstacle to economic development is that absorption constraints limit the use of an oil windfall to rapidly scale up public investment. Capacity, institutional, legal, and training constraints lengthen the time it takes to implement public investment projects efficiently, both in infrastructure and in human capital. Given that public investment can be half or more of total investment in developing economies, this is a serious constraint on potential growth. It takes a long time to recognize, implement, and realize a public investment project, especially if it is large. For example, making a sizable parcel of land suitable for modern agriculture, which requires investment in large-scale irrigation, may involve many years of negotiation with local chiefs to get permission to use the land. Years might be needed for the bureaucracy, local government, and national government to agree to undertake a particular project. And sufficient capacity to supply the necessary investment goods might be lacking. For all these reasons, not all the money spent on public sector investment will result in increases in the public sector capital stock, and public investment will be more costly in the early stages of economic development when interest rates are high. Recent studies suggest that only 40 to 60 percent of spending on public investment gets delivered and leads to effective accumulation of public sector capital, and low-income countries such as most of those in the CEMAC are likely to be at the lower end of this range (Dabla-Norris and others, 2011). The public investment management index (PIMI) illustrates this problem. The PIMI is defined as the ratio of the part of public investment that yields effective public sector capital accumulation to total spending on public investment. PIMI data were obtained through surveys about the effectiveness of project appraisal, selection, implementation, and evaluation in many countries. The PIMIs that result from these detailed country-level data were then aggregated for 40 low-income, 31 middle-income, and then all 71 countries, and are presented in Table 5.3 for the period 2007–10. Table 5.3 indicates that, on average, only about half of public investment effort translates into productive public capital.

Table 5.3Public Investment Management Index by Income Group, 2007–10
Country groupPIMIAppraisalSelectionImplementationEvaluation
Low income (n = 40)0.470.210.280.300.20
Middle income (n = 31)0.570.210.300.280.22
All countries (n = 71)0.510.
Sources: Dabla-Norris and others, 2011.Note: PIMI = public investment management index.
Sources: Dabla-Norris and others, 2011.Note: PIMI = public investment management index.

The PIMI captures the four stages of the project investment process: (i) appraisal, (ii) selection, (iii) implementation, and (iv) evaluation. The PIMI-adjusted measure of public capital is only 30 percent of GDP compared with 71 percent for the unadjusted measure in low-income countries (Dabla-Norris and others, 2011). The big declines in adjusted public capital in low-income countries are attributable to the low efficiency of new investments; in middle-income countries, the low efficiency of new investments has been offset by large investment efforts. These declines are in sharp contrast to the rise in the unadjusted measure of public capital. Despite the very high investment rates in much of sub-Saharan Africa, including the CEMAC countries, the projects have not delivered the expected growth and welfare results (e.g., Tabova and Baker, 2011). However, if the efficiency-adjusted measure of public sector capital is used, cross-country empirical evidence suggests that public capital is a significant determinant of economic growth (Gupta and others, 2011).

In line with Table 5.3, the analysis here assumes that the PIMI is roughly 0.47. As the rate of public sector investment is ramped up, the efficiency of public investment deteriorates (Berg and others, 2011). Thus, the analysis introduces internal costs of adjustment for public investment and calibrates the adjustment cost coefficient in such a way that the economy replicates the investment rates and PIMI of the low-income countries. The internal costs of adjustment of public investment also capture the increasing costs incurred when rapidly scaling up public sector investment (van der Ploeg, forthcoming). The absorption problems that frustrate rapid economic development when scaling up public investment are thus captured. Use of the internal costs of adjustment and the PIMI also result in higher returns on public investment, and thus more realistically calibrate the model to developing economies. The next section shows the optimal way to harness a temporary windfall in a small economy suffering from capital scarcity.

Capital Scarcity: Use Oil Revenue to Boost Public Investment

In the developing countries of the CEMAC, the separation theorem does not hold. Optimal public investment is, in fact, dependent on available finances and thus on the size of the oil windfall. Therefore, it would be a serious mistake for the CEMAC countries to put all oil revenue in a sovereign wealth fund when domestic investment needs are so high. To illustrate the differences between a strategy designed to harness oil revenue for economic development and a permanent-income rule, this analysis offers several policy simulations for the same temporary oil windfall discussed in the “Benchmark” section above. These policy simulations assume that the world interest rate and discount rate are 2.5 percent; an annual depreciation rate of public investment of 2.5 percent (corresponding to an expected lifespan of public projects of 40 years); a production elasticity of public capital of 0.17 (the ballpark estimate reported in Bom and Ligthart, 2010); shares of labor and private capital in value added of 0.7 and 0.3, respectively; and an elasticity of intertemporal substitution σ equal to 0.5. Furthermore, the policy simulations suppose a PIMI of 0.47 and a semi-elasticity for the debt-to-GNI ratio on the natural log of the interest spread of 1.9, as discussed in the “Two Obstacles” section. For the sake of concreteness, it is supposed that the economy starts off with an initial debt-to-GDP ratio of 100 percent and an initial public sector capital stock equal to 30 percent of GDP, and that the current size of the oil windfall is equal to current GDP. It is also assumed that all private capital is foreign direct investment for which the world interest rate is the main determinant.

For public investment, the much higher domestic interest rate is the main determinant. The public investment rate is given by I/S = (q – 1)/ϕ, where I is public investment, S is the efficiency-adjusted stock of public capital, q is the value of public investment, and ϕ = 34.4 is the adjustment cost parameter, which is calibrated to fit a PIMI of 0.47. Total spending on public investment includes internal adjustment costs and is given by J = I + 0.5 ϕ I2/S. In the steady state, the PIMI is I/J = 1/(1 + 0.5ϕδ) = 0.70, but in the early stages of development only 47 percent of investment outlays are delivered because current public investment rates (I/S) are much higher. The value of public investment q follows from the arbitrage equation, which says that the increase in private output plus the reduction in internal adjustment costs resulting from a marginal increase in the efficiency-adjusted public capital must equal the rental charge plus the depreciation charge minus the expected rate of change in the value of public capital.

Maximizing utility (5.3) subject to the equations describing asset dynamics (5.1) and public capital dynamics, = I – δS, yields the following two optimality conditions (see van der Ploeg, forthcoming, for more detail):

and the reduced-form debt and public capital dynamics are given by

where non-oil output is given by the production function Y = EKα’Sβ, 0 < α’ = 0.3 < 1, β’ = 0.17 > 0, with K indicating private capital (FDI) or, equivalently, the reduced-form production function Y = ESβ with EE11α(αr+μ)α1α,ββ1α>β, and μ indicating the depreciation rate of private capital.

equation (5.7) is the upward-sloping Keynes-Ramsey rule (also known as the Euler rule) in which growth in consumption responds to the gap between the social interest rate and the discount rate. Note that growth in consumption responds positively to the interest premium on debt, П(D/Y), П’>.0, so that it is optimal to consume less and pay off debt. Of course, the interest premium on debt should not be taken too literally, but should be seen as a metaphor for capital scarcity.6equation (5.8) is the arbitrage equation for public investment, which states that the social user cost of public capital, consisting of the social interest charge plus the depreciation charge minus the expected rate of increase in the social value of public capital, must equal the marginal product of public capital plus the marginal reduction in adjustment costs.

Equations (5.5) through (5.8) fully model capital scarcity, public investment, and windfall revenue management. They can be solved with a multiple shooting algorithm. The policy simulations can easily be done for each CEMAC country with the specific coefficients and parameters describing each country’s oil windfall, but the qualitative features will be like the ones discussed here. One caveat should be made: the simulations start with a relatively large government debt while the CEMAC countries have almost no government debt. The simulations are thus meant to be purely illustrative, with government debt not indicating so much the burden of government debt but rather the extent of the capital scarcity problem many developing countries face. A proper calibration to the CEMAC countries must be left for another occasion, or one should refer to the more careful empirically calibrated analysis of Berg and others (2011).

Figure 5.4 presents the simulations and shows both the no-windfall paths and the paths with a temporary windfall. The difference between the lines indicates the effects of the windfall. Figure 5.4 shows that the oil windfall speeds up the process of economic development, because the efficiency-adjusted public capital stock, and thus production, rise much more quickly toward their unchanged steady-state values. The temporary public sector investment boom is triggered by a temporary spike in the value of public sector capital fueled by the oil-induced boom in demand. This also leads to a temporary boost to consumption. The public investment boom has three noticeable features. First, on impact the social value of capital jumps and thus the rate of public investment rises modestly and the PIMI declines modestly.7 Second, ramping up public investment worsens absorption constraints and increases the inefficiency of public investment (see the substantial and persistent falls in the PIMI triggered by the windfall-induced investment boom in the upper left panel of the figure). Third, net government assets (i.e., the value of public capital, qS, minus sovereign debt) are not predetermined, because the shadow value of public capital jumps on impact from 2.36 to 3.19. Although the initial jump in the value of net government assets is small, the initial jump from 2.36 to 3.19 in the value of public capital is considerable.

Figure 5.4Harnesing Windfall with Capital Scarcity and Rising Cost of Ramping Up Public Investment

Source: Authors’ calculations.

Note: “x” axis (horizontal) 5 years; BIH (dashes and dots) = Bird-in-hand policy; PIH (dots) = Permanent-income hypothesis policy; No windfall (dashes); Optimal policy (solid lines).

The boom in public sector investment is associated with bigger government surpluses. This results in more rapid and substantial drops in sovereign debt, which rapidly turn into sovereign wealth. Interest spreads and the cost of borrowing decline, which gives the government more scope to provide transfers to its citizens and to ramp up public sector capital despite the lower efficiency of public sector investment. An interesting feature of the no-windfall simulations reported in Figure 5.4 is that the shadow value of public capital, q, and thus the public sector investment rate, (q–1)/ϕ (which is thus proportional to q), and the PIMI overshoot. This reflects an initial phase in the development process in which the value of public capital and the investment rate rise, and the PIMI falls before these movements are reversed further along the development path. In the windfall simulations, private consumption overshoots as well, so that in the initial phases consumption is kept low and rises gradually to make room for a rapid rise in public investment.

The Advantages of Investing to Invest

Oil windfalls can alleviate capital scarcity and thus boost public investment. This outcome is in sharp contrast to the separation result that prevails in the permanent-income rule discussed in the “Benchmark” section, which is most relevant for countries that are well integrated into world capital markets and do not suffer from capital scarcity. The permanent-income rule is a partial-equilibrium rule, given that no account is taken of the effects of the oil windfall on wages, prices, exchange rates, and quantity variables. The optimal policy rule does take account of general-equilibrium interactions—it is a one-sector model and thus only relative intertemporal prices (i.e., the interest rate) matter while relative intratemporal prices (such as the real exchange rate or the relative price of nontradables) do not play a role. (See the next section for a brief discussion of the effect of oil windfalls on relative intertemporal prices and Dutch disease effects.)

Figure 5.4 also compares the optimal investing-to-invest rule with the permanent-income rule in the private consumption and sovereign debt panels. Under the permanent-income rule, the increment in private consumption equals the permanent value of the windfall at the time the windfall starts, that is, 0.144. Over time, the return on in situ oil wealth (i.e., the permanent value of the oil windfall) falls while the increase in interest income on the balance of the sovereign wealth fund rises by an equivalent amount until it has reached 0.144—see the right panel of Figure 5.3. However, the enormous boost to sovereign wealth under the permanent-income rule does not boost economic development because public capital, private capital inflows, and non-oil production are assumed to be exogenous under this application of the permanent-income rule. The oil windfall under the permanent-income rule does sustain a permanent boost to government transfers and thus to consumption, but fails to put oil funds into public investment.

During the initial phases, the optimal strategy (see the solid lines in Figure 5.4) devotes much less to sovereign wealth or to paying off government debt and thus leads to a much bigger increase in private consumption than does the permanent-income rule (indicated by the steeply falling line in the left-bottom panel of Figure 5.4). After about 35 years, the consumption increment under the optimal strategy falls below that under the permanent-income rule. The optimal strategy thus permits much more consumption upfront and leads to a boost in public capital instead of parking that much-needed revenue in a sovereign wealth fund. Application of the permanent-income rule to developing countries such as those in the CEMAC is far from optimal, does not boost development, and does not serve the interests of citizens.

The impacts on consumption and sovereign debt if the bird-in-hand rule is used (again, not taking account of general-equilibrium interactions) are also shown in Figure 5.4. Under this rule, all windfall revenue is accumulated into a sovereign wealth fund, and no transfers and resulting boosts to private consumption are allowed until sovereign wealth is accumulated. The withdrawal rate is set at 4 percent from the balance of the fund (as in Norway). Consumption does not jump on impact but rises only gradually. Eventually, consumption reaches a higher value than under the permanent-income rule before falling back to the no-windfall path in the very long run.8 The bird-in-hand rule performs even worse than the permanent-income rule. It does not offer any prospect for improved economic development and yields larger consumption increments than the permanent-income rule only after 10 years, and higher consumption increments than the optimal rule only after 23 years, before dropping off to zero in the very long run.

The celebrated Hartwick rule states that any depletion of subsoil oil assets must be exactly offset by accumulation of other assets such as sovereign wealth and public, private, or human capital (Hartwick, 1977). In other words, genuine saving must be zero. Indeed, for the permanent-income rule, Figure 5.3 indicates that the optimal accumulation of sovereign wealth exactly equals the decline in oil wealth, so that genuine saving is zero. In contrast, the bird-in-hand rule is more conservative than the permanent-income rule in that it only consumes a fraction of the stock of accumulated wealth. The bird-in-hand rule, therefore, leads to excessive accumulation of sovereign wealth, which gives rise to positive genuine saving increments. However, developing economies with capital scarcity and increasing costs associated with ramping up public investment require negative genuine saving increments to speed up the process of growth and development so that the positive increment in net assets (public capital minus sovereign debt) at each point in time is less than the negative increment in subsoil wealth.

Another reason for negative genuine saving is anticipation of better times, which occurs if oil exporters anticipate either future reductions in the costs of extracting oil or future increases in the world oil price. It is then better to borrow on the international capital market and postpone oil depletion so that oil can be extracted more cheaply or fetch a higher price. Genuine dissaving then occurs equal to the sum of expected extraction cost reductions and expected capital losses on subsoil oil wealth.

The simulations presented in Figure 5.4 are meant to be an illustration of the main qualitative policy message: an investing-to-invest policy is better than naively applying the permanent-income or bird-in-hand rule for developing countries facing capital scarcity and large domestic investment needs. To get a better quantitative understanding of the policy message for the CEMAC countries, at least three adjustments need to be made. First, it is better to take account of the country characteristics of each member state of the CEMAC when designing the optimal policy response; for example, the Central African Republic has no oil at all. Second, the size and duration of the windfall for each oil-rich CEMAC country may vary and will generally be smaller than the windfall reported in Figure 5.4 by a factor of about four. Although the effects on growth will be correspondingly smaller, the effects for the country will still be substantial. Third, and most important, a clear distinction must be made between capital scarcity and sovereign debt. Although Figure 5.2 indicates that the CEMAC countries have little sovereign debt, they still suffer from capital scarcity because the interest payments that have to be made on domestic investment projects—as well as the potential return on such projects—are typically much higher than world averages. Hence, the policy messages derived from Figure 5.4 also apply to a capital-scarce country with almost no sovereign debt. To capture this situation adequately, the functional specification of the interest rate premium is modified to П=П(DY+N+Δ),П>0, where the constant Δ > 0 indicates capital scarcity reflected in an interest premium even in the absence of sovereign debt.

Absorption Problems and Dutch Disease: Temporarily Park Oil Revenue in a Fund

No absorption problems arise if public capital can be imported from abroad or if airports, roads, and the like are delivered by foreigners (as Chinese firms have often done in Africa and Brazil). However, for most developing economies of the CEMAC, the majority of investment goods will have to be produced at home. Nurses are needed to train nurses, teachers to train teachers, roads to make further roads, and not all of these jobs can be accomplished by importing Chinese labor and capital. Furthermore, political pressure will be brought to bear to create jobs for local people. Thus, given that oil windfalls will be geared toward domestic consumption and investment, they will boost demand and put pressure on the nontradable sectors to expand and thus lead to hikes in the price of nontradables and to reallocation of labor and capital from the tradable to the nontradable sectors (Corden and Neary, 1982). The bird-in-hand policy and the permanent-income policy fail to deliver an optimal response to such Dutch disease effects because they are derived from a partial-equilibrium framework and do not take account of changes in the real interest rate or exchange rates. It may be optimal to smooth appreciation of the real exchange rate over time and thus have a small, long-lasting rather than a large, temporary decline in the tradable sectors. Persistence in preference allows for some “addiction” to high levels of public spending even after the windfall has ceased, and may help to explain the volatile behavior of real exchange rates.

The optimal public investment and real exchange rate strategy in a fully specified, general-equilibrium, two- or three-sector model of a small, open economy must take into account capital scarcity and the rising cost of public investment. The optimal response should also take into account the need to “invest to invest.” Developing countries need teachers to train teachers, nurses to train nurses, and roads to make roads. Homegrown capital produced by the nontradable sector rather than imported from abroad is thus needed for successful economic development, but such capital takes time to deliver and leads to a different set of temporary absorption problems in the nontradable sector. In such a setting, it is optimal to temporarily invest some of the oil windfall in a sovereign wealth fund until the nontradable sector is able to deliver the investment goods necessary for economic development (van der Ploeg and Venables, 2011a). The inefficiency problem encountered when scaling up investment, as discussed in previous sections, implies that even more oil revenue has to be stored temporarily in a sovereign wealth fund.

Detailed policy simulations are reported in van der Ploeg (forthcoming). These simulations show that the price of nontradables jumps on impact at the news of the oil bonanza. Over time, as labor shifts from the tradable to the non-tradable sector, the capacity of the nontradable sector expands. The stock of public capital expands and, as the demand for nontradables is met, the initial appreciation of the real exchange rate is reversed. The absorption constraints resulting from Dutch disease are more severe if a greater part of consumption and public investment has to be produced at home. It then takes much longer for the economy to move along its development path. Again, the efficiency of public investment is reduced considerably as public investment is ramped up, which aggravates the absorption constraints resulting from Dutch disease, and the windfall is also used to bring down sovereign debt more quickly. The oil bonanza increases consumption of both nontradables and tradables, but output of tradables falls considerably to make room for a boost in nontradables output. The oil windfall finances the resulting current account deficits and the more rapid reduction in sovereign debt.

With a putty-clay technology for public capital, public investment can occur in either the nontradable or the tradable sector, but public capital cannot be shifted between sectors once installed. The only way the nontradable sector can expand if all capital is used in the tradable sector is for the tradable sector to gradually wind down its stock of public sector capital stock via wear and tear, thus bringing down its stock of private capital and releasing labor, which can then move to the nontradable sector. This adjustment mechanism also leads to a temporary appreciation of the real exchange rate and a gradual reallocation of workers from the tradable to the nontradable sector, but the root cause of this adjustment process is wholly different if public sector capital is only used in the nontradable sector, as in the simulations illustrated by Figure 5.4. Whether the Dutch disease dynamic is primarily driven by winding down capital in the tradable sector or by increases in the price of nontradables and sectoral reallocation depends on whether the tradable or the nontradable sector is more capital intensive. A detailed analysis of these two adjustment mechanisms to Dutch disease, that is, gradual accumulation of homegrown capital versus the gradual erosion of capital in the tradable sector, is much more fully analyzed within the comprehensive theoretical model in van der Ploeg and Venables (2011a).


The oil-producing countries of the CEMAC would be unwise to follow either the permanent-income rule or the bird-in-hand rule and put their already declining oil windfalls into sovereign wealth funds. The main concern of the CEMAC countries is to boost economic development, not to engineer a sustained increase in consumption financed by interest earned on U.S. Treasury bills. The CEMAC countries’ main challenges are to deal with, on the one hand, capital scarcity, and on the other hand, declining inefficiency of investment projects in public infrastructure, health, and education as public investment projects are scaled up using what is left of the oil bonanza. It is far better for the CEMAC countries to use their oil windfall to steadily ramp up public investment, tolerate a temporary fall in the efficiency of public investment, and gradually boost the efficiency-adjusted stock of public capital and non-oil output. As a result of this strategy, wages and consumption would rise in the initial phases of economic development much more than they would under the permanent-income or the bird-in-hand rules. The optimal policy for harnessing the oil bonanza requires negative genuine saving so that the positive increment in net assets (public capital minus sovereign debt) at each point in time is less than the negative increment in subsoil oil wealth.

In practice, absorption constraints are aggravated because private demand and especially public sector demand are strongly biased toward nontradable products. This bias results in appreciation of the real exchange rate and Dutch disease. Nevertheless, the optimal harnessing strategy yields a sustained increase in private consumption and non-oil gross national product. The bigger the component of public capital that has to be homegrown, the bigger the absorption constraints. This is the challenge of “investing to invest.”

Important inputs into the decision-making process can be obtained from financial programming models and from full-scale, empirically calibrated, dynamic stochastic general equilibrium models of oil-rich developing economies. Such models allow a richer derivation of optimal policy proposals for managing oil windfalls, especially with regard to how much of the windfall to consume upfront; how much sovereign debt to redeem or sovereign wealth to accumulate; and most important, how much investment in public infrastructure, health, and education to undertake, and how the optimal response is affected by changes in the real exchange rate and Dutch disease effects. The crucial point is that the prevalence of macroeconomic and microeconomic absorption constraints in ramping up investment in public infrastructure and human capital requires oil-rich, developing countries to temporarily park part of their oil windfalls in sovereign wealth funds.


    Akitoby, B., and T.Stratmann,2008, “Fiscal Policy and Financial Markets,” Economic Journal, Vol. 118, No. 533, pp. 1971–85.

    Barnett, S., and R.Ossowski,2003, “Operational Aspects of Fiscal Policy in Oil-Producing Countries,” in Fiscal Policy Formulation and Implementation in Oil-Producing Countries, ed. byJ.Davis, R.Ossowski, and A.Fedelino (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Berg, A., R.Portillo, S.-C. S.Yang, and L.-F.Zanna,2011, “Government Investment in Resource Abundant Low-Income Countries (unpublished, Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Bom, P., and J.E.Ligthart,2010, “What Have We Learned from Three Decades of Research on the Productivity of Public Capital?” Working Paper No. 2206 (Munich: CESIFO).

    • Search Google Scholar
    • Export Citation

    Cordon, W.M., and J.P.Neary,1982, “Booming Sector and De-Industrialisation in a Small Open Economy,” Economic Journal, Vol. 92, No. 368, pp. 825–48.

    • Search Google Scholar
    • Export Citation

    Collier, P., F.van der Ploeg, M.Spence, and A.J.Venables,2010, “Managing Resource Revenues in Developing Economies,” IMF Staff Papers, Vol. 57, No. 1, pp. 84–118.

    • Search Google Scholar
    • Export Citation

    Dabla-Norris, E., J.Brumby, A.Kyobe, Z.Mills, and C.Papageorgiou,2011, “Investing in Public Investment: An Index of Public Investment Efficiency,” Working Paper 11/37 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Gupta, S., A.Kangur, C.Papageorgiou, and A.Wane,2011, “Efficiency-Adjusted Public Capital and Growth,” Working Paper 11/217 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Hartwick, J.M.,1977, “Intergenerational Equity and the Investment of Rents from Exhaustible Resources,” American Economic Review, Vol. 67, No. 5, pp. 972–74.

    • Search Google Scholar
    • Export Citation

    Mehlum, H., K.Moene, and R.Torvik,2006, “Institutions and the Resource Curse,” Economic Journal, Vol. 160, No. 508, pp. 1–20.

    Sachs, J.D., and A.M.Warner,1997, “Natural Resource Abundance and Economic Growth,” in Leading Issues in Economic Development, ed. byG.Meier and J.Rauch (Oxford: Oxford University Press).

    • Search Google Scholar
    • Export Citation

    Tabova, A., and C.Baker,2011, “Determinants of Non-Oil Growth in the CFA-Zone Oil-Producing Countries: How Do They Differ?” Working Paper 11/233 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    van der Ploeg, F.,forthcoming, “Bottlenecks in Ramping Up Public Investment,” International Tax and Public Finance.

    van der Ploeg, F.,2011, “Natural Resources: Curse or Blessing?” Journal of Economic Literature, Vol. 49, No. 2, pp. 366–420.

    van der Ploeg, F., and S.Poelhekke,2010, “Volatility and the Natural Resource Curse,” Oxford Economic Papers, Vol. 61, No. 4, pp. 727–60.

    • Search Google Scholar
    • Export Citation

    van der Ploeg, F., and A.J.Venables,2011a, “Absorbing a Windfall of Foreign Exchange: Dutch Disease Dynamics,” OxCarre Research Paper 52, (Oxford, UK: OxCarre).

    • Search Google Scholar
    • Export Citation

    van der Ploeg, F., and A.J.Venables,2011b, “Harnessing Windfall Revenues: Optimal Policies for Resource-Rich Developing Economies,” Economic Journal, Vol. 121, No. 551, pp. 1–31.

    • Search Google Scholar
    • Export Citation

    van der Ploeg, F., and A.J.Venables,forthcoming, “Natural Resource Wealth: The Challenge of Managing a Windfall,” Annual Review of Economics.

    • Search Google Scholar
    • Export Citation

    World Bank, 2006, Where is the Wealth of Nations? Measuring Capital for the 21st Century (Washington).


A detailed survey of the evidence for and causes of the natural resource curse is given in van der Ploeg (2011).


These add to the usual definition of saving a measure of education spending to reflect investment in human capital and subtract depreciation of physical and human capital, the use of natural resources (making special allowances for renewable resources such as fish and forests), and the deterioration in environmental quality (mainly arising from carbon dioxide and fine particles pollution).


The cross-country empirical evidence suggests a very weakly significant direct positive effect of natural resource exports on interest rate spreads, indicating, if anything, that resources worsen creditworthiness. This may be the result of the adverse impacts of resources on governance, political stability, and the risk of conflict.


The oil windfalls of some CEMAC countries yield substantial revenue but are declining. Hence, the CEMAC oil windfalls are insufficient to pay off the whole external debt to remove the problem of capital scarcity altogether.


Using the empirical estimate of 1.9 for the semi-elasticity of the natural log of the spread with respect to the debt-to-GNI ratio (van der Ploeg and Venables, 2011b), we specify П(d) = 10–4 exp(6.294) [exp(1.9d)–1] for the interest spread schedule where 6.294 is the mean log of the spread. This implies that a 10 percentage point increase in the debt-to-GNI ratio pushes up the interest differential by 6.9 percentage points if the economy starts out with a debt-to-GNI ratio of 100 percent (or 1.3 percentage points if it starts off with zero foreign debt).


If there are also adjustment costs associated with changing the rate of public investment (rather than adjustment costs for the stock of public capital), there would be no discrete jumps on impact in public investment or the PIMI.


This does not show up in Figure 5.4 because it takes a much longer time than the 40 periods indicated by the horizontal axis.

    Other Resources Citing This Publication