Tackling Inflation During Reforms in Africa

International Monetary Fund. External Relations Dept.
Published Date:
January 1991
  • ShareShare
Show Summary Details

The rapid increase in the rate of inflation in a large number of economies in Sub-Saharan Africa (SSA) has been a cause of much concern, not least because of its damaging social and economic consequences, and few African countries have the institutions to manage inflation, either through indexation or through safety net programs. Moreover, with persistently high inflation there is the danger that inflation may become self-generating by strengthening changing expectations about future inflation. While few African countries have experienced the triple-digit inflation often observed in Latin America, the average inflation rate has increased markedly, sometimes fivefold in many countries.

From 1975-89, average inflation in SSA was about 16.7 percent per annum, much above the average rate of 6.4 percent per annum in industrial countries (see chart and table). But this average masks considerable disparities. In particular, since 1977 inflation has been much lower in countries belonging to the CFA Franc Zone, where it has closely followed French inflation, whereas triple digit inflation has been recorded in at least four countries—Ghana, Sierra Leone, Uganda, and Zaire.

A more detailed discussion is available in “Africa’s Rising Inflation: Causes, Consequences, and Cures,” by the author, PRE Working Paper, No. 577, The World Bank, Washington, DC.

It has been argued by some that devaluation, combined with price and interest rate liberalization—an integral part of many World Bank and IMF-supported structural adjustment programs—has led to “cost-push” inflation, that is, inflation that is created and sustained by increases in the cost of production, these increases being independent of the state of demand. Without a sufficient supply response (i.e., the extent to which output increases in the export or import substituting sectors), these price changes have had little positive real effects, while generating higher inflation. This view is supported by the relatively low inflation rates in the CFA Franc Zone countries, which have a fixed exchange rate system and whose monetary policy has to follow certain rules. But the relationship between exchange rate policy and inflation is more complex, and depends on the presence of parallel markets and whether appropriate monetary and fiscal policies are pursued as part of the reform effort.

This article presents the major lessons for an anti-inflationary policy, and is based on a World Bank research project covering Algeria, Cote d’lvoire, Ghana, and Zimbabwe. Each of these countries was selected as a prototype of the different macropolicy regimes described in the box. In particular, this article discusses the impact on inflation of a realignment of exchange rates, monetary and fiscal policies, and price liberalization.

Exchange rate policy

There are two approaches to effecting a change in the real exchange rate—through a nominal exchange rate devaluation or through a deflation of the economy. Some economists have argued that the latter approach is preferable because it avoids the costs of high inflation. A somewhat different, but related, critique has come from the Economic Commission for Africa (ECA), which argues that a nominal devaluation is very quickly wiped out by inflation, with no effect on the real exchange rate. This arises from the low supply response and the high share of monetary financing of the budget in many African countries. But what does the evidence tell us?

First, there is no doubt that there is a direct cost-push effect from exchange rates to domestic prices. This is confirmed by the work on Ghana and Zimbabwe as part of this project, as well as by numerous studies in Africa and other parts of the world. In general, the immediate effect of exchange rate changes on inflation is likely to be high, even if imports represent a small proportion of GDP. This is due to the importance of key imports in the production process and the inability of domestic industry to substitute for these imports. In Ghana, where the share of imports in GDP is about 0.20 percent on average during the estimation period, we find that a 1 percent increase in the exchange rate leads to a 0.40 percent increase in prices.

Second, the initial effect of exchange rate changes on domestic prices is small when there is a substantial domestic capacity to produce goods that can substitute for imports. The effect is reduced even further when price controls and subsidies are introduced. This has been demonstrated in Zimbabwe, where both factors lead to a very small initial effect of exchange rate changes on domestic inflation. In the extreme situation, where there are complete price controls, as in Algeria, exchange rate changes have no effect on domestic prices, indicating that domestic inflation is primarily a function of changes in price controls. However, real or “virtual” inflation in the system is much higher. Moreover, the use of subsidies to avoid the full impact of exchange rate changes on domestic prices often leads to fiscal deficits whose financing has overall inflationary implications.

Inflation in Sib-Sahahran Africa, 1961-89

Source: Wortd Bank data.

Third, the effects of devaluation on domestic inflation depend to a large extent on accompanying fiscal and monetary policies. A simulation exercise on Zimbabwe, under alternative assumptions about export response and the share of monetary financing of the budget, shows that under normal circumstances, about a third of the nominal devaluation would get offset by an increase in domestic prices. The full effect of a devaluation on domestic prices is larger than the initial effect, because the budget deficit in Zimbabwe increases with a devaluation. A higher budget deficit that is financed mainly by money creation leads to higher inflation. In Zimbabwe, the net effect of a 10 percent increase in the rate of devaluation is an increase in inflation by 3.33 percent. When the supply response of exports is low in the short run (elasticity 0.1), and a high proportion of the budget is financed by monetary creation, devaluation is completely offset by a rise in domestic prices, and there is no effect on the real exchange rate. Because the domestic bond market is not adequately developed for the government to borrow in the domestic market, monetary financing of the budget deficit is quite common in Africa. These results appear to lend some credence to the ECA critique. But their resolution lies in managing the fiscal deficit and pursuing an appropriate monetary policy, not in avoiding exchange rate adjustments.

Infalation in Sub-Saharan Africa: 1975-89(Percent, annual average)
1900-891975-89Highesl recorded1930-891975-89Highest recorded1980-891975-89Highesl recorded
Industrial Countries5.06.4Between 70-20 percentLess than 10 percent
Asia7.97.3Botswana10.511.016,4Burkino Faso4.16.430.0
Sub-Saharan Africa17.216.7Burundi7.610.136.6Cameroon19.49.817.2
Gambia17.915.156.7Central African
More than 20 percentKenya10 411.320.4Republic23.9-14.6
Sierra Leone63.743.9178.7Lesotho13.613.916.0Cote d Ivoire15.39.527.4
Sudan133.125.164.7Malawi16 633.9Ethiopia4.38.328.5
Source: World Bank data

For these countries, averages are for 1980-88 and 1975-88.

For these countries, averages are for 1981-89.

Source: World Bank data

For these countries, averages are for 1980-88 and 1975-88.

For these countries, averages are for 1981-89.

Moreover, exchange rate changes can also often improve the fiscal deficit and will thereby lead to lower inflation. In Ghana, for example, devaluation led to an improvement in the fiscal deficit, and as a result, a declining rate of inflation. Simulations with a macroeconomic model for Ghana show that a slower devaluation would have meant a higher rate of inflation, as well as a more depreciated parallel exchange rate. The fiscal deficit improved, because devaluation resulted in higher receipts in domestic currency of foreign aid and a reduction in exchange subsidies to importers.

Finally, the size of parallel markets determines the degree by which official devaluation will affect prices. Once the misalignment of the exchange rate has gone very far, consumer and producer prices already reflect the parallel market exchange rate—the official devaluation is just the formalization of the status quo. The Ghana study showed through appropriate econometric work that devaluation of the official exchange rate had no direct cost-push effect. This is true also of a large number of countries in the Type III category in Africa, where severe misalignments of the exchange rate have led to the emergence of a large parallel market for foreign exchange. Hence, reforms without devaluation are unlikely to be successful. In these cases, managed official devaluation leads to lower inflation if the budget improves along with devaluation.

These studies make one point clear: there is no unique relationship between devaluation and inflation. Initially, there is always a positive cost-push effect of a devaluation on inflation, as is confirmed by almost all studies of inflation in Africa. The exceptions are countries with complete price controls. But it is clear that these controls are usually unsustainable and their removal typically involves wrenching economic—and often political—change, because of a release of repressed inflation. The subsequent and overall effect of devaluation on inflation, however, depends on the impact of the devaluation on the budget, the adequacy of external funding of the program, the export response, and the accompanying monetary policy.

Monetary and fiscal policies

In Africa, as in other parts of the world, there is a strong link between the growth of the money supply and inflation. Empirical results in Ghana and Zimbabwe show that excess real money balances have a significant and positive effect on inflation. This link is confirmed by a number of other studies on inflation in Africa. In these cases, a reduction in the fiscal deficit, and thereby, in the growth of the money supply, is obviously necessary to reduce inflation. But, as is shown in the case studies, how and where fiscal cuts are made matters a great deal in reducing inflation. Expenditure cuts that affect productive investments will lead to a much smaller impact on inflation and, in the extreme, may even raise it.

The strong link between monetary growth and inflation raises the issue of what causes low inflation in the CFA Franc Zone countries. A recent study of inflation in the CFA Franc Zone showed that in three of the four countries examined, there was no link between domestic credit growth and monetary growth. According to this study and the case study on Cote d’lvoire for this project, excessive domestic credit growth adversely affects the balance of payments instead of resulting in high inflation.

Classification of macropolicy regimes in Africa, 1975-89
Type I
Fixed exchange rates; no capital

controls; none or very few price controls
BeninEquatorial Guinea
Burkina FasoGabon
Central AfricanMali
Côte d’lvoire
Type II
Fixed-but-adjusting exchange rates;

capital controls; selective price controls
Cape VerdeSeychelles
Type III
Dual or multiple exchange rates; capital

controls; selective price controls
GambiaSao Tome & Principe
Type IV
Dual or multiple exchange rates; capital

controls; widespread price controls
Source: “Africa’s Rising Inflation Causes. Consequences, and Cures,” by Ajay Chhibber.

Although Algeria is not in Sub-Saharan Alrica, it is included here for discussion.

Countries such as Angola. Ethiopia. and Mozambique have extensive price controls bul are unable to enforce them.

Source: “Africa’s Rising Inflation Causes. Consequences, and Cures,” by Ajay Chhibber.

Although Algeria is not in Sub-Saharan Alrica, it is included here for discussion.

Countries such as Angola. Ethiopia. and Mozambique have extensive price controls bul are unable to enforce them.

The benefit of membership in the monetary union is obviously lower inflation. This has very little to do, however, with the fixed nominal exchange rate. It arises primarily from the common monetary policy and no capital controls. But their experience in the 1980s demonstrate that there are some costs as well. The consequences of a fixed exchange rate have been high variability in the real exchange rate in these countries and lower growth.

Liberalizing prices

In examining the relationship between price decontrols and inflation, one has to look at three cases: economies with extensive price controls; economies with selective-but-enforced price controls; and economies with parallel markets.

In the first case, the issue in managing the price decontrol is one of measuring the extent of monetary “overhang” (i.e., excess liquidity) and determining how much can be absorbed by special schemes and how much will be dissipated through inflation. These special schemes could involve the sale of public assets, such as housing. Alternatively, the government could sell bonds that provide a positive real return to absorb money balances that would otherwise drive up the price of scarce commodities. The remaining money supply will probably lead to higher demand for goods and services, driving up inflation. The degree of supply response will also determine the extent of inflation. In Ghana, for example, higher real growth has led to an increase in the real money balances, lowering the inflationary impact from the same level of monetization.

The enforcement of selective controls implies that demand for commodities subject to price controls cannot be met. Typically, price controls are effective when they are applied through public enterprise pricing, but can be easily circumvented by the private sector. In Zimbabwe, low-income consumers are protected by subsidized transport, electricity, and publicly provided housing. Part of the unmet demand spills over into the market for uncontrolled commodities, whose prices rise, and the unspent balance accumulates as in the first case in the form of monetary assets. In this situation, price decontrol typically leads to higher prices in the sectors that were earlier subject to price controls; such price rises may increase costs in other sectors and thus their prices.

Finally, the existence of parallel markets eliminates any excess demand, as demand for controlled commodities spills over into parallel markets, leading to higher prices there. Inflation is unrecorded and price decontrol typically has insignificant inflationary effects.


At first glance, the evidence appears to suggest a strong correlation between exchange rate regimes and inflation. Countries with floating exchange rates (or auction systems for access to scarce foreign exchange) appear to have experienced higher inflation. On the other hand, countries with fixed exchange rates have typically faced lower inflation.

A careful analysis, however, indicates that the story is more complex. In countries such as Ghana, Sierra Leone, Uganda, and Zambia, the high inflation was prevalent prior to the exchange reforms of the 1980s and was prevalent through periods when the exchange rate was fixed. The high inflation rendered the official exchange rate more or less irrelevant and parallel markets emerged. The level of the official exchange rate was important for accounting purposes, but had very little relationship to the true price of foreign exchange—the parallel exchange rate. Detailed analysis of the Ghanaian exchange reforms shows that adjusting the official exchange rate may have actually lowered inflation by reducing fiscal deficits. The underlying cause of inflation was the high fiscal deficits, which were financed primarily through money creation.

The lower and stable inflation in countries with pegged exchange rates, such as those of the CFA Franc Zone, also has its genesis in the underlying monetary and financial arrangements, rather than in the fixed exchange rate. The openness of the capital account between countries of the Zone ensures that the money supply is not a policy variable. Domestic credit expansion does not lead to monetary expansion and inflation, but instead affects the balance of payments.

The key, then, to price stability lies in providing checks on large fiscal deficits and non-inflationary mechanisms for financing them, rather than the exchange rate regime per se. In principle, this can be done through responsible expenditure and revenue policies. In practice, experience shows that it requires institutional arrangements that restrain excessive expenditure. Unrestricted movement of capital provides one mechanism for such a restraint. The other is to effectively separate monetary and fiscal policy, either by joining a monetary union such as the CFA Franc Zone, or by establishing and managing an independent central bank. This independence also needs an institutional structure akin to the Federal Reserve system in the United States. The central bank needs to be given its autonomy.

The costs of joining a monetary union appear to be high: pegging the nominal exchange rate restricts adjustment in the real rate and throws all adjustment effects onto fiscal policies. But it may ensure price stability. The best policy option is to have appropriate monetary and fiscal policies without the rigidity of a fixed pegged exchange rate. This may pave the way for price stability without jeopardizing growth.

The most useful, comprehensive, and detailed set of statistics available on the debt of developing countries.

—Robert H. Cassen, Director, International Development Centre, University of Oxford

World Debt Tables 1990-91

Featuring complete and up-to-date data on debt for more than 100 developing countries

Includes the latest 1989 data and projections through 1999

Plus periodic supplements through the year as new data become available

When you need current and reliable data on developing country debt, you will find no better source than World Debt Tables 1990-91. This new edition gives you instant access to data gathered through the World Bank’s exclusive Debtor Reporting System. You also get from top World Bank economists an analysis of the world debt situation that will help you put the data into perspective.

World Debt Tables 1990-91 Volume 1. Analysis and Summary Tables Volume 2. Country Tables

Complete two-volume set and supplements

□ 686pages / ISBN 0-8213-1725-3 / US$125.00/Order Stock #/1725

Available separately: Volume t. Analysis and Summary Tables

□ 208pages/ISBN 0-8213-1723-7/ US$16.95/ Order Stock 11723

Coupon orders accepted only from the U.S.A. and France. To order from other countries, request a list of World Hank Publications Distributors. Prices may wary by country.

In the U.S. return the coupon to: World Bank Publications P.O. Rox7247 K619 Philadelphia, PA 19170-8619

In France return the coupon to: World Bank Publications 66, avenue d’lena 75116 Paris. France

Other Resources Citing This Publication