Information about Sub-Saharan Africa África subsahariana
Journal Issue
Share
Chapter

Chapter 4 The Macroeonomic Impact of Scaling Up Aid

Author(s):
Matthew Gaertner, Laure Redifer, Pedro Conceição, Rafael Portillo, Luis-Felipe Zanna, Jan Gottschalk, Andrew Berg, Ayodele Odusola, Brett House, and José Saúl Lizondo
Published Date:
March 2012
Share
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

A few questions stand out when examining the macroeconomic impacts of scaled-up aid. The first is overarching: can aid-financed public investment lead to higher growth? The question of whether and to what extent foreign aid helps poor countries grow remains controversial. The aid-growth literature provides mixed conclusions on the effectiveness of aid in stimulating GDP growth, as opposed to other microsocial outcomes, for which the aid impact is clearer.23 Recent studies include Rajan and Subramanian (2008), which concludes that “it is difficult to discern any systematic effect of aid on growth” (p. 660). Clemens, Radelet, and Bhavnani (2004) argue that aid intended for investment purposes does show a growth impact. More recently, Arndt, Jones, and Tarp (2009) claim that “aid has a … significant causal effect on growth over the long term…. [and] remains an important tool for enhancing the development prospects of poor nations” (p. 1).

Aid can lead to growth, depending on how it is used and on country-specific factors. In general, aid is used, in part, for public investment in capital that will increase the marginal productivity of capital (both private and public), and thus stimulate higher growth. However, the impact of aid-financed public investment on growth depends on country-specific factors, including, among others, (i) the extent to which aid-associated spending is directed to projects that, if implemented well, yield a sound return; (ii) the degree to which spending is relatively efficient ($1 spent does not necessarily buy $1 more of roads or schoolbooks, but some fraction that depends on administrative costs, corruption, and the like); and (iii) the existence of supporting institutions, such as the legal system and financial services, that enable a private sector response to public investment.24 These country-specific conditions may play a significant role in determining the actual growth impact, as suggested by the empirical literature about the link between public investment and growth. (Box 4.1).

Box 4.1.Public Investment and Growth

Public investment (in roads, education, health care, and so on) is clearly critical to development. But how much will overall output increase in response to a given increase in public investment spending? The evidence from the empirical literature does not provide clear-cut answers.a Although many studies seem to suggest a positive relationship, particularly for infrastructure investment, the magnitudes of these contributions vary considerably from one study to another, presumably because of differences in the econometric methodology employed, the level of data aggregation, and the type of public investment considered. A majority of recent studies—especially those using physical indicators such as miles of road and phones per capita to measure investment in infrastructure—find significantly positive effects of public capital on growth. In contrast, the findings are less robust in those studies that use public investment flows (or their cumulative value), likely a reflection that investment spending may be a poor proxy for the accumulation of productive assets, owing to waste, inefficiencies in public procurement, corruption, or inclusion of current expenditures (e.g., wages and salaries) in reported investment figures.b

The lack of unanimity in the results may be explained by country-specific factors. The empirical literature suggests that the impact of public investment on growth may depend on, among others, (i) the financing source of public investment; (ii) the institutional context within which investment decisions are undertaken; (iii) the quality of project evaluation, selection, and management; and (iv) the regulatory and operational framework in which infrastructure services are provided.

aSee the recent surveys of this literature by Romp and de Haan (2007) and Straub (2008a, 2008b) and the references therein.bSee Easterly and Rebelo (1993) and Pritchett (2000). The recent work by Arslanalp and others (2010), however, finds a positive impact for public capital—built from investment flows—on economic growth, using a data set that includes high-, middle-, and low-income countries

The second question is a corollary to the first: can aid inflows hurt growth through real exchange rate appreciation that undermines growth-promoting export industries, the so-called Dutch disease? Another question is, if the Dutch disease phenomenon is present, would it diminish the positive effects of public investment temporarily or permanently? These effects have traditionally been linked to revenue inflows associated with discoveries of oil or other natural resources, but interest has risen about the similar impact of increased aid inflows (Box 4.2).

Box 4.2.Dutch Disease

One of the major issues in the debate on scaling up aid has been the possible adverse macroeconomic consequences. An increase in aid is similar to a commodity windfall in that it results in large flows into the economy, which can cause an appreciation of the real exchange rate, directing resources into the more profitable nontradables sector and adversely affecting the tradables sector. The adverse impact on the tradables sector is often dubbed “Dutch disease.”

Whether a real appreciation is, in fact, a disease—that is, whether it is harmful—depends in large part on the role played by the tradables sector in productivity growth. Some microeconomic evidence indicates that exports growth may have a larger positive impact on productivity than other types of growth, generating positive externalities that increase overall growth.a These externalities might arise, for example, as exporting firms “learn by doing” and the new know-how spreads to other domestic firms. As a general matter, the experience of many countries with export-led growth, as well as the empirical evidence on the related natural resource curse, suggests caution in discarding this concern too rapidly. However, public investment in infrastructure, health, and education is also critical to growth, and aid can help finance this spending, so even with the positive externalities of the tradables sector, a balanced approach to investment is important.

Critically, the real exchange rate appreciation and contraction of the tradables sector need not be a “disease.” These occurrences may be just a necessary counterpart to freeing up the resources required to make aid-financed investments. Even if exports create positive externalities, if aid is spent well and the policy response is appropriate, the impact on competitiveness would be temporary, with the medium-term returns to growth outweighing any negative externalities associated with a temporary decline in the tradables sector. Indeed, if the tradables sector creates positive externalities and public capital is invested to further stimulate its productivity (for example, through better transportation networks), the increased aid would have synergistic effects on competitiveness, exports, and productivity over the medium term, a phenomenon Berg and others (2010) label “Dutch vigor.”b

The literature regarding the potential impact of aid on real exchange rates and manufacturing is growing. But the empirical record is mixed. Rajan and Subramanian (2011), for instance, conclude, by measuring the relative growth rates of exportable industries, that aid inflows seem to have systematic adverse effects on a country’s competitiveness. It underscores that a channel for these negative effects is the real exchange rate appreciation caused by aid inflows. Christiansen and others (2009), however, find that foreign aid is progressively absorbed over time through net imports, and is associated with a more depreciated real exchange rate in the long term. The mixed empirical results are not surprising given the country-specific factors that condition the effects of aid, such as the efficiency and nature of investment, the responses of policymakers and the private sector, and so on, as emphasized in the text.

aSee, for instance, Van Biesebroeck (2005), who finds that exporting has raised productivity in manufacturing firms in SSA.bSee also Torvik (2001) and Adam and Bevan (2006).

Can fiscal and monetary policy responses shape the macroeconomic effects of aid surges? Concerns about competitiveness and real appreciation have frequently caused authorities to accumulate some portion of the aid flows as international reserves—partial aid absorption—while still spending the local currency counterpart of these flows (Berg and others, 2007). When public investment is productive, partial aid absorption helps limit short-term real appreciation of the currency but at the expense of lower medium-term growth because private consumption and investment are crowded out. Similarly, countries that seek to maintain the level of the nominal exchange rate to avoid losing competitiveness are likely to still experience (i) real appreciation through inflationary pressures, when aid inflows are not sterilized; or (ii) crowding out of the private sector, thus constraining growth, when there is full sterilization.25 Therefore, policies can exert an important influence on the macroeconomic impact of aid and, to that extent, should be designed to help reap the growth benefits of aid. However, macroeconomic policies cannot compensate for poor investments or absorptive capacity constraints.

The macroeconomic analysis undertaken by the IMF for these 10 countries focused on these questions. In general terms, by setting country-specific parameters, the studies traced how an aid increase would affect public investment spending, how public investment spending would translate into an accumulation of public capital (spending efficiency), how this would affect the marginal productivity of public and private capital, and how private investment would respond. Public and private capital accumulation would drive growth in the medium term. At the same time, based on specified fiscal and monetary policy responses, the studies assessed how an aid shock might affect inflation and the real exchange rate, and thus move private investment toward the nontradables sector, away from the potentially more dynamic tradables sector. Special attention in the adjustment process was given to the possibility of having Dutch disease or Dutch vigor effects.

Design and Features of the Model

Most of the analyses of the pilot countries were conducted using a common framework developed at the IMF by Berg, Gottschalk, Portillo, and Zanna (Berg and others, 2010). The framework is based on a small open-economy quantitative model in the tradition of dynamic stochastic general equilibrium (DSGE) models, in which short-term growth is driven by changes in aggregate demand, while longer-term growth is driven by accumulation of capital (Box 4.3).

Box 4.3.DSGE Models

Dynamic stochastic general equilibrium (DSGE) models provide a coherent framework for policy discussion and analysis. Thus, they are playing an important role in the formulation and communication of monetary policy at many central banks in both developed and developing economies.a

The benchmark DSGE model for an open or closed economy is micro-founded, includes real and nominal rigidities, and emphasizes agents’ intertemporal choice (Christiano, Eichenbaum, and Evans, 2005; and Smets and Wouters, 2003). Because current agents’ choices depend on future uncertain outcomes, the model is inherently dynamic and allows for agents’ expectations to influence current macroeconomic outcomes.

The basic structure is built around three interrelated blocks: a demand block, a supply block, and a policy block. The blocks capture the behavior of the economic agents of the economy—households, firms, and the government. These agents interact in markets that clear every period, reflecting general equilibrium features. The demand block describes households’ decisions. They consume, decide how much to invest, and are monopolistic suppliers of labor, which allows them to set wages. The supply block captures firms’ behavior. Firms hire labor, rent capital, and set prices because they are monopolistic suppliers of goods. The policy block models government decisions. Fiscal policy is usually restricted to a Ricardian setting, while monetary policy is conducted through an interest rate feedback rule, in which the interest rate is set in response to deviations from an inflation target and an output gap. This basic structure is enriched with a stochastic structure of demand, supply, and policy shocks.

The DSGE-type model used for most of the Gleneagles scenarios is an open-economy version with tradable and nontradable goods that has been specifically developed to analyze scenarios in which aid is scaled up. The model has a number of low-income-country-specific features. The economy includes “hand-to-mouth” consumers without access to financial markets, amplifying the effects of demand shocks. Domestic and foreign assets are imperfect substitutes, so the capital account can be closed. And as described in the text, the model contains a number of features designed to capture the role of separate and uncoordinated monetary and fiscal policy responses to aid surges, public investment efficiency, absorptive capacity constraints, and the risks of Dutch disease.

This category of models is well suited for macroeconomic analysis and for addressing policy issues of interest in low-income countries. The models may be helpful for analyzing the macroeconomic effects of particular shocks in these countries such as those to terms of trade, remittances, and aid. Admittedly, these models, as complicated as they are, contain many rough approximations and shortcuts, and will not on their own produce accurate forecasts. However, they can help organize thinking, offer a way to systematically incorporate various sorts of empirical evidence, and provide a vehicle for transparently producing alternative macroeconomic scenarios, which can be compared across countries. These features make them suitable for analyzing aid scaling-up scenarios.

aFor a review of the issues and challenges surrounding the use of DSGE models at central banks, see Tovar (2009).

The model was designed to analyze both the short- and medium-term macroeconomic effects of aid surges by capturing the main mechanisms and policy issues of interest in low-income countries. The model incorporates a role for public capital in production so that government spending can raise output directly as well as induce private investment. Specifically, an increase in public investment from an aid increase leads to an increase in the public capital stock, which results in higher GDP growth. Firms respond to the positive impact of public investment on the marginal productivity of private capital by increasing investment, thus expanding private capital and growth.

But the model allows for varying degrees of public investment efficiency, allowing for less than full conversion of investment into useful public capital. This, in turn, affects both the private investment response and growth.

The model includes a learning-by-doing externality in the tradables sector to capture the notion that real exchange rate appreciation may harm productivity in this sector and hinder overall growth. However, even if such externalities exist, the gains from well-invested aid can outweigh these negative effects. Indeed, if aid is invested well, the externalities raise the productivity of the tradables sector, such that aid can produce even greater gains in growth—higher public capital accumulation induces crowding-in effects on private capital accumulation, which, over the medium term, helps raise tradables output above its trend, producing Dutch vigor. In contrast, when aid is not invested well, these externalities induce declines in productivity in the tradables sector that suggest that aid can harm growth—less public and private capital accumulation might not fully offset the negative effects of real appreciation on productivity, resulting in an adverse impact on growth.

On the policy front, the model allows for separate and possibly uncoordinated fiscal and monetary policy responses to an aid surge, permitting a variety of policy combinations to help contain inflation and real exchange appreciation. The fiscal authorities’ decision corresponds to a view on whether to spend or save the aid. The monetary authorities’ decision, however, is about whether to “absorb” the aid and finance a higher current account deficit by selling the aid-associated inflows in the foreign exchange market, or to accumulate the aid as international reserves with resulting crowding out of the private sector (Box 4.4). The model also allows for issuance of central bank bonds to sterilize increases in the domestic money supply that may occur as a result of reserves accumulation. Such sterilization policies lead to increases in the real interest rate and, therefore, to crowding out of private sector activity.

The model facilitates an examination of the impact of the efficiency of public investment on growth. First, it is important to distinguish between the efficiency of public investment based on aid-surge funds and the historical efficiency of public investment. It is the efficiency of the aid-surge-related public investment relative to historical investment efficiency that determines how much impact aid-financed public investment has on growth. Inefficient aid-surge-related investment will create only a small amount of additional public capital—however, historically low efficiency also means that previous public investment spending created very little public capital to begin with, so even a small addition to public capital can make a big difference to output. This conclusion has some important and perhaps surprising implications. If a country has a lot of trouble converting investment spending into useful capital—in other words, both historical and aid-related investment efficiency are low—the growth effects of a given aid surge are hard to determine.

Box 4.4.Spending and Absorption

The management of aid flows occurs through the interaction of fiscal and monetary policies, through spending and absorption. Spending is defined as the widening of the fiscal deficit associated with the use of an increment of aid. Absorption is defined as the widening of the current account deficit (excluding the inflow of the aid) resulting from foreign exchange purchases associated with the aid.

Absorption measures the extent to which aid engenders a real resource transfer through higher imports. For a given amount of fiscal spending, absorption depends on exchange rate and monetary policies. The government spends some aid dollars on imports and can sell the remainder of the aid dollars to the central bank in exchange for a local currency counterpart to finance spending on domestic goods. The degree of absorption depends on the central bank’s response. The central bank can sell the foreign exchange to mop up the local currency counterpart spent by the government, thus allowing the private sector to import commensurately.

If the central bank accumulates the foreign exchange as international reserves instead of selling it, the real exchange rate is unlikely to appreciate to the same extent. However, the authorities will have attempted to use the same aid flow twice—once to finance spending and once to increase reserves—and this can have unintended negative effects. Once the aid money is accumulated as reserves, it is not available to finance the aid-related spending. As this spending takes place, the money supply increases. If the monetary authorities do not react, the resulting inflation will in effect tax domestic residents to pay for the spending. Central bank efforts to issue bonds instead to soak up (“sterilize”) the increase in the money supply will potentially raise interest rates and crowd out sector activity.

Berg and others (2007) provide evidence of SSA countries that, when confronted with aid surges, spent most of the aid but limited aid absorption in an effort to contain the impact on the real exchange rate and competitiveness. In several of the countries studied, aid was spent, but was not fully absorbed, to buffer reserves and out of concerns about exchange rate appreciation. This “double use” of the foreign exchange from aid inflows—for both spending and to buffer reserves—led to higher inflation in some cases and to an increase in interest rates and the domestic debt burden in others. The simple lesson is that the reactions of monetary, fiscal, and reserves policy must be coordinated to avoid unintended negative consequences from aid surges.

But if a country’s investment efficiency either declines with the aid surge (perhaps because it cannot handle the larger aid volume), or increases (for example, because of improved investment planning and financial management practices), a measurable effect will result. Finally, the model can be calibrated to each country. Because the model is micro-founded, some parameters can be based on microeconomic evidence, such as the efficiency of investment. Other parameters depend on steady-state ratios that can be determined from national income accounts, public and private sector balance sheets, and input-output matrices, or can be informed by more or less structural macroeconometric estimates, for example, money demand parameters. Other parameters describe the policy response to aid or the policy regime in place and can, therefore, be treated as free parameters and modified according to the policy experiment.

Synthesis of Country Case Studies

Simulation results from the 10 country studies suggest that increased aid can have a significant positive long-term impact on economic growth, and that negative effects on inflation and real exchange rates should be manageable. However, this outcome depends crucially on the macroeconomic policy response, as well as on the efficiency of public investment and the response from the private sector. Effective use of increased aid presents the main challenge, given constraints on absorptive and administrative capacity in most of the countries.

In most of the case studies, a significant increase in aid flows would be required for the Gleneagles commitments to be met. Benin, Liberia, Rwanda, and Zambia are currently receiving relatively high levels of aid. Ghana, Tanzania, and Togo would require a scaling up of aid of at least 5 percent of GDP, and more than 10 percent of GDP would be necessary in the Central African Republic, Niger, and Sierra Leone.26 The case studies assumed that the increases in aid were used to scale up public spending on investment, broadly defined, as prioritized by each country’s national poverty reduction strategy. Most of the case studies assumed that the additional aid was in the form of grants, with no additional debt created as a result of the scaling up. This is a critical assumption—if additional debt were to be created, the fiscal implications of future interest payments as well as debt sustainability would need to be taken into account, lowering the growth payoff of the higher aid.

All country cases showed a positive growth impact from an aid surge, although the scale of the macroeconomic effects varied according to the specific country circumstances and to the magnitude and duration of the aid increase required to meet an ODA level of $85 per capita (2004 prices). The assessments for Benin, the Central African Republic, Niger, Rwanda, Sierra Leone, Tanzania, and Togo show a significant effect on GDP growth due to the higher investment in public infrastructure and the resulting positive impact on private sector investment. Even in countries with lower aid increments, the results are nonnegligible: in Benin, for example, the incremental increase in aid in 2008-15 is estimated to boost growth by 0.8 percent of GDP on average relative to the benchmark; consequently, per capita income would be 29.4 percent higher in 2015, reaching $574 in 2007 constant prices.

The positive growth return occurs despite higher inflation and real exchange rate appreciation in several cases in which increased aid has a temporary adverse impact on the tradables sector (Table 4.1). In 7 of 10 cases, an aid shock resulted in real exchange rate appreciation of more than 2.5 percent (and up to 15 percent for Niger and the Central African Republic) against the steady state, but in all cases the tradables output recovered over the long term and GDP was permanently higher than in the steady state. In Togo, for example, the aid surge was quite large, about 9.5 percent of GDP. The real exchange rate appreciated by about 9 percent and inflation was 5 percent higher than in the steady state, with nontradables production higher and tradables production lower. However, over the long term, the tradables sector recovers with production higher than in the steady state despite the appreciated exchange rate, showing evidence of Dutch vigor—higher productivity resulting from public investment outweighed the effect of the real exchange rate appreciation, adding about 1 percentage point to the real GDP growth rate. A similar effect is seen in one of the scenarios for Tanzania, for which aid increased by 6 percentage points of GDP and the real exchange rate appreciated by 4 percent: the tradables sector was initially adversely affected, but tradables output was higher over the long term, again adding about 1 percentage point to the real GDP growth rate.

Table 4.1.Summary of Baseline Scaling-Up Scenarios for 10 Pilot Countries(percentage deviation from benchmark, unless otherwise noted)
CountryAid shock

(% of GDP)
REER

appreciation

t+2
Tradables

production

t+2
Nontradables

production

t+2
Annual average

increase in real

GDP level

t+2 to t+12
Sierra Leone14.66.0-6.06.02.7
Niger14.015.0n.a.n.a.2.5
Central African Republic14.015.0-11.010.03.5
Togo9.59.0-9.011.01.5
Tanzania6.04.0-2.03.01.0
Liberia14.91.50.00.50.2
Ghana4.52.70.01.01.0
Benin2.45.9n.a.n.a.0.8
Zambia0.90.6-0.61.10.3
Rwanda0.91.4-1.30.50.1
Source: Authors’ calculations.Note: n.a. = Not applicable; REER = Real effective exchange rate; t+2 = 3 years from the latest year for which data were available in the respective country.

Because Liberia already receives more than $85 per capita in ODA, an ODA of 4.9 percent of GDP was assumed.

Source: Authors’ calculations.Note: n.a. = Not applicable; REER = Real effective exchange rate; t+2 = 3 years from the latest year for which data were available in the respective country.

Because Liberia already receives more than $85 per capita in ODA, an ODA of 4.9 percent of GDP was assumed.

The magnitude of the real exchange rate appreciation, inflation, and the growth impact depended significantly on the macroeconomic policy response to increased aid. The Niger, Rwanda, Tanzania, and Zambia cases contain scenarios with partial absorption policies (Box 4.4). These cases show that partial absorption could help limit short-term real appreciation as compared with full absorption of the aid increase, but at the expense of lower growth over the long term because private sector activity is crowded out. The Tanzania case goes further to show a scenario with partial absorption in which the increase in the domestic money supply is not fully sterilized; in this case, inflation and interest rates spike sharply with a further adverse impact on private sector investment.

Fixed exchange rate regimes can present difficult issues for macroeconomic policymakers. In cases in which the use of the aid requires a real appreciation, a fully absorbed aid surge will lead to a period of inflation: with the stable nominal exchange rate, the real appreciation requires a rise in the price of nontradable goods. The Central African Republic (fixed exchange rate regime) and Sierra Leone (floating exchange rate regime) cases show similar magnitudes of increased aid in their scaling-up scenarios, with almost full absorption in both cases. In the Central African Republic, the real exchange rate appreciates by 14 percent, with a temporary increase of 10 percentage points in the inflation rate. There is a pronounced but relatively temporary effect on tradables output. In Sierra Leone, more is spent on imported goods, resulting in lower exchange rate appreciation (6 percent). The exchange rate regime also makes a difference for inflation—with a nominal appreciation of the rate, inflation initially falls rather than rises, but stabilizes quickly.

Policymakers should be careful not to misinterpret—and overreact to—the inflation that may result from an aid surge in a fixed exchange rate regime. If they attempt to fight it, for example, by keeping the money supply stable in the face of the aid inflow, they may succeed in narrow terms in keeping inflation low, and thus in stabilizing the real exchange rate. But like any partial absorption strategy, higher domestic interest rates and crowding out of private investment and consumption would ensue, as if the additional spending were domestically financed. In the end, inflation may be a necessary corollary to the effective use of the aid. Of course, policymakers need to be aware that inflation may increase for other reasons, such as excessively loose fiscal and monetary policies, in addition to the aid surge.

The country cases reinforce the point that higher aid efficiency is important for maximizing the return to growth and mitigating the effects on inflation and the real exchange rate, both in magnitude and speed. Inefficient use of aid raises the risk that Dutch disease effects would dominate, limiting the positive impact on medium-term growth. Most of the country studies contained scenarios examining the impact of a decline in aid efficiency. Almost all of these scenarios, however, showed that the growth return to public investment eventually outweighed temporary negative impacts on inflation and competitiveness.

A decline in the marginal productivity of public investment or the diversion of aid to finance unproductive government consumption would reduce the boost to medium-term growth from aid, and raise the risk that the effect of a real exchange rate appreciation could outweigh the impact of the increased capital stock on growth, creating a Dutch disease scenario. Indeed, the Niger and Togo studies showed scenarios in which efficiency was lowered to a point such that Dutch disease effects dominated and the tradables sector was unable to recover, yielding an outcome in which productivity and growth were lower than in the baseline aid shock scenario (where parameters were set at historical and/or standard levels). Whether Dutch disease effects are important—or even whether there is a disease at all—depends on two factors: the extent of externalities in the tradables sector and the productivity of the aid-financed investment spending. Gains in productivity associated with growth in the tradables sector can amplify the positive effects of aid on tradables output and real GDP (the aforementioned Dutch vigor), but this effect can also magnify the negative impact of a contraction in the tradables sector due to real appreciation resulting from increased aid inflows (Dutch disease). Whether the learning-by-doing externalities intensify the positive or negative impact of aid depends on the efficiency of public investment and on the extent to which increased public capital accumulation is successful in crowding in private investment. Such crowding in helps ensure that the initial fall in tradables output fades quickly rather than becoming protracted. Given the problems across the region with effective execution of public investment projects in the past, public financial management reforms must occur ahead of the scaling up of aid to ensure effective use of the increased resources.

The impact of scaled-up aid on growth is likely to be reduced in countries in which the private sector remains underdeveloped or constrained by an unfavorable business climate. Although the case studies did not incorporate a more muted private sector response to public capital accumulation—one proxy in the DSGE model would be to restrict access to financial services even further, or to raise the cost private firms face in adjusting their investment rate—several cases pointed to the investment climate as an important factor in the response of growth to aid. The medium-term growth impact of increasing aid is likely to be stronger if the increase in public investment is reinforced by an expansion of private capital in response to increased marginal productivity; a muted response leads to a higher risk of Dutch disease.

The caveats in interpreting the results of the country studies should be considered. Data limitations remain a critical constraint in calibrating the model for each country, requiring that a number of assumptions be made in setting the country-specific parameters (Box 4.5). In addition, the region is undergoing rapid structural change, and many of the parameters that describe key relationships in these economies may be changing over time. Finally, although the model is designed to reflect the dynamics of low-income countries as closely as possible, the analysis might not sufficiently capture channels of transmission relevant to Africa.27

Box 4.5.Assumptions for Calibrating the DSGE Model

The DSGE model was calibrated using a number of assumptions for country-specific parameters. The broad categories included historic averages of GDP components; basic macroeconomic variables; public sector assets; other basic characteristics of the economy; and the size, duration, and use of the aid increase. After the model was calibrated, a baseline aid shock was examined. Thereafter, parameters were modified to simulate different aid and policy scenarios.

For example, the baseline calibration for Tanzania is depicted in the table below.a

GDP components (% of GDP)
Consumption83
Consumption of tradable goods50
Investment27
Government spending25
Government consumption16
Government spending on tradable goods9
Government spending on nontradable goods16
Trade balance-35
Exports14
Imports49
Inflation, growth, interest rate (%)
Annualized inflation5
Trend growth7
Annualized real interest rates2
Aid process (% of GDP)
Benchmark aid8
Persistence of aid shock (fraction of previous year’s shock that persists in current year)0.95
Aid increase5
Duration of aid increase (number of quarters)20
Policy rule parameters (%)
Amount of increased aid spent100
Persistence of real debt accumulation90
Portion of aid shock spent on nontradables60
Portion of aid shock spent on investment50
Sale of foreign exchange from aid surge100
Weight of exchange rate target1.5
Inflation-targeting coefficient28
Assets (% of GDP)
Real money balances10
Private sector net foreign assets4
Reserves13
Government deposits at the central bank7
Government debt held by the central bank4
Government debt held by the private sector4
Total government debt8
aGiven data constraints, parameters on basic characteristics of the economy (e.g., technology, elasticity of substitution, wage adjustment, capital mobility, financial market access) were kept at standard levels for low-income countries in the model for Tanzania.

    Other Resources Citing This Publication