II. Monetary and Exchange Rate Policies in Sub-Saharan Africa

International Monetary Fund. African Dept.
Published Date:
April 2008
  • ShareShare
Show Summary Details

A marked improvement in macroeconomic conditions in most sub-Saharan African countries in recent years has reshaped the environment for monetary and exchange rate policy. A number of post-conflict economies are still in a stabilization phase and remain vulnerable, and a much smaller number suffer from severe instability, but in many sub-Saharan African countries higher economic growth has been associated with lower inflation, higher international reserves, and healthier public finances (Chapter 1).

The burden of fiscal dominance for the conduct of policy has been greatly reduced in most sub-Saharan African central banks. With the support of IMF-supported programs and substantial debt relief, and with a favorable external environment, bank financing of the budget deficit is now negligible in most countries, helping anchor inflation expectations. Moreover, pervasive distortions in financial and exchange rate markets have been greatly reduced (Masson and Pattillo, 2004). The risk of fiscal dominance has not completely disappeared, however; a severe worsening of external conditions or political disarray, for example, could renew fiscal pressures on monetary policy.

Monetary policy in Africa was until recently concerned primarily with bringing inflation down to (or near) single digits and anchoring inflation expectations. Given a lack of credibility and the need for a nominal anchor, policy focused on targeting intermediate variables, such as the exchange rate or monetary aggregates. While most African countries initially relied on de jure exchange rate pegs as a nominal anchor, most countries outside the CFA zone shifted to targeting money (Adam and O’Connell, 2006).

With inflation stabilized, many African countries—particularly those with some form of managed floating exchange rate regime—are in the process of adapting their policy regimes. In the long run, stable and predictable inflation contributes to an efficient financial sector, low real interest rates, and a good investment climate (Box 2.1). In the short run, however, there are trade-offs between price, output, and exchange rate stability. These trade-offs become sharper in a low-inflation environment, and monetary policy must strive to minimize macroeconomic volatility while keeping inflation low by identifying and responding to domestic and external shocks. Transitioning toward such a framework, which is the defining feature of modern monetary policy (Goodfriend, 2007), is an important component of the medium-term policy agenda in sub-Saharan Africa.

Box 2.1.Inflation Objectives for Sub-Saharan African Low-Income Countries

Monetary policy should aim to bring about and then preserve price stability. This box analyzes the choice of inflation objectives in sub-Saharan Africa and offers some guidance.

The long-term objective is for inflation to be low enough to allow households and businesses to ignore inflation altogether while accommodating relative price changes without the need for deflation. While country-specific factors and value judgments about policy trade-offs generate some variation, there is a consensus among policymakers (including in low- and middle-income countries without IMF-supported programs) that inflation in the range of 2–5 percent is appropriate for developed countries and somewhat higher—perhaps 5–10 percent—is a reasonable long-run target for low-income countries, given their need for larger and more frequent relative price adjustments.

  • A lower long-term inflation objective may not be desirable, given upward biases in CPI measurements, and may expose countries to the possibility of deflation and a liquidity trap.

  • Higher inflation tends to lower growth, in part because higher and more volatile inflation tends to distort relative prices and depress the quantity and quality of investment.

  • Higher inflation is a poor way to finance higher spending. It corresponds to a tax on holders of cash. It has several disadvantages: it is likely to fall mostly on the poor, who cannot escape through investment in indexed financial assets; it erodes other tax revenue; and it requires high inflation to raise a small amount of revenue. The 2005 review of the IMF’s Poverty Reduction and Growth Facility (PRGF) concluded from a study of 48 PRGF countries that an increase in the inflation rate from 6 to 15 percent would finance only 0.4 percent of GDP in spending, ignoring second-round negative effects on growth.

  • Empirical work on the long-run effects of inflation supports these arguments, though it does not give precise answers about appropriate upper and lower bounds (see Selassie and others, 2006).

  • To maintain competitiveness, countries with fixed exchange rates, such as those in the CFA zone and the Rand Monetary Area, need to keep inflation in line with that of the anchor currencies, adjusted for any trend differences in productivity growth.

Short- to medium-term inflation objectives depend on initial inflation levels. Disinflating from a very high initial rate (say, above 50 percent) may be relatively painless, or even expansionary, provided there is a credible commitment to lowering inflation and fiscal and other policies are supportive (Fischer, Sahay, and Végh, 2002). However, at moderate levels of initial inflation (say, in the low double-digit range), disinflation becomes more costly because most of the adjustment must operate through aggregate demand. When starting at moderate inflation, it is generally advisable to reduce inflation more gradually.

In general, the higher the central bank’s credibility, the less costly disinflation, and the lower the short- to medium-term inflation objective can be. Credibility and reputation are strengthened in part by a central bank’s accepting output costs to controlling inflation. This would suggest that central banks may at times wish to be fairly tough in order to establish credibility. Policymakers’ understanding of this is reflected in the adoption of institutional frameworks thought to support credibility, such as central bank independence. However, the benefits of higher credibility need to be balanced against the costs of acquiring it. And inflation expectations are also determined by the structure and the state of the economy. This limits room for tough policies to lower expectations of future inflation.

Once inflation is within the long-term target range, policymakers should resist persistent increases in inflation because the potential benefits are likely to be more than outweighed by the cost of later stabilization. Purely transitory supply shocks of volatile items such as food need not be counteracted.

Note: This box was prepared by Jan Kees Martijn and Hans Weisfeld.

The purpose of this chapter is to take stock of current practices and provide general policy recommendations. The first section reviews the landscape of monetary and exchange rate policy regimes in sub-Saharan Africa, with a particular emphasis on the role of reserve-money targets and exchange rates in the policy frameworks in the region. The second section discusses policy issues.

While all types of regimes found in the region are discussed, the focus is on managed float regimes. These present a special challenge: while they allow for more policy flexibility, there is little guidance on how policy should be conducted. Moreover, there has been little research on African, particularly low-income, countries with managed floats.

The review of monetary policy in the region reveals that many countries behave to some extent as informal inflation targeters. While money targeting is the official nominal anchor for many countries, there is considerable flexibility in meeting targets. For countries with low to moderate inflation, such target flexibility is not associated with inflation surprises. This suggests not that money target misses reflect unexpected (and undesired) deviations in the monetary policy stance but rather that these targets may not fully determine the stance of monetary policy in the short run. On the other hand, countries with high inflation persistently miss targets by substantial amounts, which suggests that money targets in these countries are one way to identify the substantial deviations in policy that are typically associated with fiscal pressures.

To the extent that there is substantial flexibility in practice in countries with low to moderate inflation, greater clarity about the policy stance and overall framework can contribute to monetary policy effectiveness. Depending on each country’s specific conditions, the choice of policy regime falls along a continuum, from greater emphasis on intermediate targets on one end to a direct focus on inflation objectives on the other. Regardless of where each country stands, there are merits to greater transparency, communication, and a clearly articulated effort to actively respond to shocks while controlling inflation expectations.

Considerable uncertainty surrounds the monetary policy transmission mechanism in Africa, posing a challenge for the conduct of monetary policy. Thin though growing credit markets, pervasive excess liquidity in the banking system, and large structural changes affect the channels through which policy influences prices, notably by limiting the influence of policy interest rates (Box 2.2). Lack of reliable, timely statistics poses an additional challenge in terms of identifying shocks and pinning down the current state of the economy. Some of these uncertainties suggest a potentially important role for monetary aggregates, especially as an indicator variable, although they also confirm the need for flexibility in meeting monetary targets. Ultimately, however, the role that monetary aggregates should play in the policy framework is a country-specific question.

Box 2.2.The Monetary Transmission Mechanism in Sub-Saharan Africa

An effective monetary policy requires an understanding of the ways in which monetary policy affects inflation and other key macroeconomic variables. The transmission mechanism is complex and constantly evolving, even in developed countries. In African countries, the challenge is compounded by structural changes and the stabilization of the economy.

The transmission mechanism differs depending on whether or not monetary policy is dominated by fiscal considerations. Under fiscal dominance, inflation expectations reflect lack of a credible anchor and are sensitive to fiscal news. Lax policies have a rapid effect on inflation, often preceded by nominal depreciations (Fischer, Sahay, and Végh, 2002).

  • Nachega (2005) finds that the need for seigniorage revenue drove inflation in the Democratic Republic of Congo over the period 1980–2003.

  • Unsustainable government spending may lead to inflation even if monetary financing is currently low, as the experience of Kenya, Nigeria, The Gambia, Zambia, and Zimbabwe in the 1990s demonstrates (Buffie, 2003).

During stabilization episodes, the anchoring of inflation expectations is the most important factor driving inflation. Provided that fiscal dominance is credibly eliminated, monetary policy can reduce inflation quickly.

Once inflation has been brought down to low levels, monetary policy operates through three channels: the exchange rate, the interest rate, and credit (Kamin, Turner and Van’t dack, 1998).

  • The exchange rate is often the most important channel in the region. In Kenya, Cheng (2006) finds that policy-driven changes in the short-term interest rate have a considerable impact on the shilling. Primarily through this channel, changes in monetary policy account for about one-fourth of annual inflation dynamics. The effect operates mostly through pass-through rather than expenditure-switching effects; output does not seem to respond to changes in policy.

  • The interest rate and credit channels have been found relevant in some African countries. In Namibia, Uanguta and Ikhide (2002) show that monetary tightening, driven by balance of payments deficits under Namibia’s peg, often lead to increases in real lending rates and a contraction in private sector credit.

Imperfections in the financial sector limit the effectiveness of the credit and interest rate channels:

  • The transmission of the policy stance to interest rates, even at short maturities, is hampered by limited bank participation in shallow or dormant interbank markets (Laurens, 2005).

  • Monetary tightening has limited impact on credit when banks hold involuntary excess reserves, which seems pervasive in local banking systems and may reflect unwillingness or inability to lend (Saxegaard, 2006).

  • The small size of financial sectors in both low- and middle-income countries in the region limits the macroeconomic impact of credit conditions (IMF, 2006a). Over 2000–04, private sector credit averaged 13 percent of GDP for low-income countries and 21 percent for middle-income countries excluding South Africa. Such limited size reflects little access to bank credit (among the lowest in the world) and significant crowding out by public sector borrowing (Sacerdoti, 2005).

Changes in the financial sector imply that the interest and credit channel will become more important over time. Reforms are improving the operation of interbank and government debt markets, enhancing their role in the transmission process. Access to credit is improving and greater participation by foreign banks will likely improve the sector’s efficiency (Cihák and Podpiera, 2005).

Formal inflation targeting is by no means the end goal for every floating exchange rate country in the region. International experience demonstrates that there are many effective ways of conducting monetary policy in a float. Nonetheless, central banks in Africa need to develop domestic financial systems, establish credibility, and improve in-house technical capacity—especially for liquidity forecasting—even as they may continue to target money, at least de jure, in the foreseeable future. However, inflation targeting provides a useful benchmark for understanding and assessing floating regimes, and countries may wish to move to adopt some elements of inflation targeting, especially if they are considering a move to formal inflation targeting in the longer term.

The review of current practice finds that, in some countries with de jure managed float regimes, nominal exchange rate stability is common. In such countries, exchange rates display prolonged episodes of near-complete stability. They also display periodic discrete changes, however, suggesting that there may be only a limited commitment to a given exchange rate level. This assessment complicates the analysis of policy regimes in managed floats: some countries may be acting as de facto pegs, while others may rely on exchange rate stability partly to achieve inflation or other objectives. The chapter does not attempt to classify de facto regimes; rather, it considers alternative ways in which exchange rate stability may enter the policy framework in the region and the challenges this poses for monetary policy.

Given its role in influencing inflation and inflation expectations, management of the exchange rate can help stabilize inflation. However, exchange rate management must remain subservient to price stability in a floating regime, or tensions between exchange rate and inflation objectives will arise. The tensions are bound to become more prominent as capital mobility increases and, combined with the lack of transparency in exchange rate targets, can undermine the credibility of monetary policy.

The Monetary and Exchange Rate Landscape

This section describes the policy frameworks in place in sub-Saharan Africa, with an emphasis on the role of reserve money and exchange rates in countries with managed floats. The section draws on a survey of IMF staff country teams and on empirical evidence from the past decade.


The policy framework can be characterized as the list of objectives and targets that guide policy and the instruments for implementing it. The ultimate objectives of policy usually include price and exchange rate stability, economic growth, and financial sector stability; and they are sometimes translated into final targets—for example, 2 percent inflation. Intermediate targets focus on variables that provide information about the expected path of final targets; they help guide policy without any necessary presumption that they must be met. Operational targets apply to variables that directly influence the money or exchange rate market equilibrium, usually signaling the current policy stance. Finally, policy instruments are the tools to help achieve these targets.

De jure frameworks are a useful starting point for the analysis of monetary and exchange rate policies. Because they reflect how the authorities describe their own regimes, a focus on them can facilitate the policy debate in the region. In addition, tensions between conflicting objectives or targets are easier to identify by comparing the de jure framework with actual practice.

Table 2.1 summarizes the de jure frameworks in sub-Saharan Africa.1 Three main conclusions emerge:

  • Countries with a managed float have a variety of competing policy objectives.2 While the main objective is price stability, the stability of the exchange rate also appears as a policy objective in most countries (19), through either its effect on external competitiveness (8) or the desirability of exchange rate smoothing (11).

  • Reserve money targeting, which is prevalent, is a distinctive feature of monetary policy in Africa. In the rest of the world, the practice of emphasizing money as the nominal anchor has long since disappeared.3

  • The role of the exchange rate in the policy framework is difficult to assess from the de jure description. Some de jure managed float regimes may be pegs in disguise. It is also important to consider other ways in which exchange rate stability enters the policy frameworks in managed floats.4 These issues are explored in detail below.

Implementation of reserve money targeting

Broad money appears as an intermediate target in most countries that target reserve money. While in principle this intermediate target is essential to monetary targeting, it can be bypassed in practice: not all money targeters look at broad money. Moreover, it appears that overshooting broad money targets does not always elicit a policy response: reportedly, only five countries tighten policy when the target is overshot.

Table 2.1.De Jure Monetary Policy Frameworks in Sub-Saharan Africa
RegimesPolicy ObjectivesIntermediateOperational TargetMain Instruments
Pegs (23)Stability of the exchange rate regime (23)Private sector credit (1)Exchange rate (23)Open market operations
Price stability (23)Foreign exchange sales
Economic growth (12)
Money targeting (18)Price stability (all countries)Monetary aggregates (16)Reserve money (18)Open market operations (17)
External competitiveness (5)Foreign exchange sales (18)
Exchange rate smoothing (12)
Economic growth (9)
Inflation targeting (3)Price stability (all countries)Interest rates (3)Open market operations (3)
External competitiveness (1)Foreign exchange sales (3)
Exchange rate smoothing (1)

Misses are common with reserve monetary targets. Over the past two years, most countries have missed targets at least once. The reasons vary, but they include limitations in policy implementation, portfolio inflows and incomplete sterilization, and changes in money demand.

Partial rebasing is also common. When there is a miss, the new money target to some extent preserves the subsequent rate of change, rather than attempting to undo the miss and preserve the previously targeted levels. Money targets were usually partially rebased in at least 11 of the 15 countries where money has been part of IMF-supported program conditionality. In that sense, monetary policy partially lets bygones be bygones.

A closer quantitative look at reserve money operating targets sheds some light on the role these targets play. Countries with high inflation do display consistently higher money growth than expected (Box 2.3). Moreover, in those countries, target misses are in the same direction as misses of the inflation target. In these countries, large and sustained misses tend to signal a breach of stabilization policy, which carries high inflation costs.

Box 2.3.Assessing Reserve Money Targeting in Sub-Saharan Africa

Monetary targets and outcomes were recorded for a sample of 16 countries with de jure managed floats. The selection of countries was based on the availability of data and the explicit targeting of reserve money in the context of an IMF-supported program. Variables used are quarterly reserve money growth (actual and 2–5-months-ahead projections) and end of period annual CPI inflation (actual and one-year-ahead projections). Data were obtained from IMF staff reports of individual countries for 2002–06.1

Reserve money targeting can be assessed in a number of ways. The average reserve money target bias indicates whether targets were missed systematically in a particular direction and by how much. The (squared) mean square error of the reserve money target measures the average size of misses, in either direction. It is also revealing to analyze whether reserve money target misses were correlated with inflation.

Average Bias in Reserve Money Target Versus Average Inflation

Countries with high inflation displayed higher than targeted growth in reserve money. There is a positive correlation between average target misses and average inflation across all countries (first figure). This relationship is driven by the fact that reserve money targets in countries with high inflation (above 11 percent, the median inflation in the sample) behave very differently from those in countries with lower inflation. Countries are thus separated into two groups (high- and low-inflation countries) for the remainder of the analysis.

Sub-Saharan African Countries: Reserve-Money-Targeting Performance
Average RM GrowthRM Target BiasMean Square Error of the RM TargetCorrelation Between RM Misses and Inflation Misses
High-inflation countries11.34.410.30.71
Low-inflation countries5.6-0.64.3-0.06

Correlation of Reserve Money Target Misses and Inflation Misses

In high-inflation countries, money growth typically exceeded targets by 4.4 percent. The median mean square error in this group is quite high. In addition, within each country, higher than average target misses are strongly associated with contemporaneous misses in inflation (table and second figure).

In countries with lower inflation, target misses tend to have no cost in terms of inflation. Money growth was on average roughly on target. Yet misses were common and fairly large, though lower than for higher-inflation countries. Within each country, these misses were either not correlated or negatively correlated with misses in inflation (table and second figure).

1 Given the limited number of observations per country, these results are indicative. Using annual observations for all PRGF countries, however, Selassie and others (2006) find broadly similar results.

In lower-inflation countries there is no clear relationship between adherence to money targets and inflation. These countries make substantial target misses in both directions on any given test date (they tend to average out). However, misses are not generally correlated with inflation surprises. This indicates that, if average inflation is good, there are no costs associated with greater money target flexibility.

Other evidence also suggests that targeting end-of-quarter money stocks in IMF-supported programs may create incentives for temporary adjustments to meet the targets. Compared with countries without IMF-supported programs, those that do have one tend to have a significant and large drop in money growth at the end of the quarter that is undone the following month. This may create a misleading impression about the true importance of the targets, as well as perhaps creating unnecessary interest rate volatility.5

These results indicate substantial flexibility in money targeting. The next subsection elaborates on why flexibility may enhance the effectiveness of monetary policy and also on the challenges it poses for policy transparency.

The role of the exchange rate

A clear conclusion from the staff survey is that monetary authorities include exchange rate stability as one of their policy objectives. This is not novel to Africa: in most developing countries there are concerns about the high pass-through of exchange rate fluctuations and how currency mismatches affect domestic balance sheets (Calvo and Reinhart, 2002).

The concern for exchange rate stability raises the question of the role of the exchange rate in de jure managed floats. There is a range of possibilities: at one extreme the authorities may occasionally attempt to smooth extremes of exchange rate stability in the context of money or inflation targeting; and at the other, the regime may be indistinguishable from a pegged exchange rate, at least for a period of time.

In practice, countries do not generally persist at one well-defined point on the policy spectrum. The regimes and the external environment may evolve, and the shocks to which the authorities respond with exchange rate flexibility may only occur occasionally. Thus, even countries that display substantial exchange rate rigidity, including de jure pegged regimes, typically adjust them with some frequency (Frankel, 2004).

The potential—and frequently observed—flexibility even in countries that display substantial rigidity is of interest from a policy perspective. Thus, while efforts to classify regimes in terms of their de facto behavior are very important for some purposes, this chapter focuses on how to understand the role of exchange rates in a de jure floating regime.6

The complexity of the situation in African de jure floating regimes—and the frequency of changes even in countries that mostly keep their nominal exchange rate stable—can be observed indirectly by dividing de jure managed floating regimes in the region into two groups of countries: (i) those for which changes of more than 1 percent happen frequently; and (ii) those in which the nominal exchange rate often remains unchanged. Figure 2.1 plots the frequency distribution of monthly exchange rate changes for these two groups of countries, with the first group defined as countries for which the exchange rate change is zero in less than 30 percent of months.7

Figure 2.1.Sub-Saharan Africa: Exchange Rate Distribution, 2005–07

Source: IMF, International Financial Statistics.

Broadly, exchange rate behavior in the first group of countries is consistent with the authorities using available instruments to influence the stability of the exchange rate without directly controlling its value. Authorities may conduct monetary policy partly to stabilize the exchange rate in addition to other short-term objectives.

The substantial exchange rate stability observed in practice in the second group of countries substantially complicates the analysis of policy frameworks in those countries. There are several possible reasons why some countries may display stable exchange rates, which may be difficult to disentangle:

  • Some countries may be anchoring their policies on preserving a fixed nominal exchange rate. Of the 14 countries in this group 5 were classified as de facto exchange rate pegs by the IMF Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER) for 2007. There may indeed be a de facto commitment to a given exchange rate level that may dominate other policy concerns, including inflation.

  • The nominal exchange rate may serve as an operating target, possibly along with other operating targets like reserve money or interest rates. In this case, the monetary authorities aim to achieve the desired level of the exchange rate at a given point in time, not as the nominal anchor but in direct support of other objectives. These other objectives could be attempting to reduce nominal (and real) exchange rate volatility as an additional subordinate objective—especially if inflation falls within a certain range—as well as achieving desired inflation rates.8 Unlike a peg, however, there may not be a clear commitment to a particular level, and the exchange rate may be adjusted as necessary in support of the final policy objectives. In some cases, the role of the exchange rate may be analogous to that of the policy interest rate in a developed country float, with the important difference that exchange rate target decisions may lack transparency.9

There is no clear difference between the two groups of countries in terms of inflation or output performance (in levels or volatility). There is also no clear relationship to per capita income, somewhat contrary to the “fear of floating/learning to float” literature for emerging markets, which emphasizes that time and development are required to achieve full flexibility of the exchange rate in a floating regime (IMF, 2004b).

The fact that some managed floats are hard to distinguish from pegs also suggests that lack of clarity about the role of the exchange rate may pose a challenge to the effectiveness and credibility of monetary policy. In assessing the nature of a particular regime, a critical question is which objectives are subordinate and which are dominant, and which may only be revealed by certain shocks (and possibly with a lag). For example, whether nominal exchange rate stability is subordinate to the inflation objective or not may only be revealed when there is a large and persistent shock to the real exchange rate, such as a major shift in the terms of trade.10 The next section elaborates on the advantages of clarity and transparency in the conduct of monetary policy.

Policy Challenges of Implementing Monetary and Exchange Rate Regimes in Sub-Saharan Africa

This section provides general recommendations on monetary policy frameworks for Africa. It draws on specific features of African economies, the extensive literature on monetary policy in developed and emerging market countries, and the international cross-country policy experience of the past 20 years.

Monetary policy in countries with pegs

Countries with hard-pegged regimes—the WAEMU, the CEMAC, and the countries that have pegged their currency to the South African rand—have the benefit of a transparent framework that credibly provides nominal stability and lowers international transaction costs, providing impetus to trade and other forms of international integration.11 The challenge lies in adjusting to shocks that call for shifts in the real exchange rate. The experience of strong growth after the 1994 devaluation of the CFA franc indicates the importance of avoiding prolonged and severe misalignments. In the absence of flexibility in nominal exchange rates, reliance must be placed on supportive measures, such as fiscal policy and structural reforms that encourage wage flexibility and promote productivity growth.

Monetary policy can provide a supportive role, however. With a hard peg, inflation determines the real exchange rate. In the medium to long run, monetary policy cannot influence this relative price.

However, with limited capital mobility, it might make a difference over shorter horizons. Thus, monetary policy needs to be consistent with the demands of the exchange rate regime.12

The handful of countries with de jure conventional pegs benefit from the stability and credibility pegs can provide. These are mostly countries in the process of stabilization or small very open economies, such as Cape Verde. For these countries, the potential benefits of exchange rate stability are likely to outweigh the potential benefits of running an autonomous monetary policy (Mussa and others, 2000). As with hard pegs, these countries must ensure that other policies support required adjustments to the real exchange rate. They may also carefully consider possible vulnerabilities to capital account and other shocks as they become more integrated with international capital markets.

Managing a managed float

The considerable success many sub-Saharan African countries have had in stabilizing inflation over the past few years suggests that current frameworks have worked well for them. Supported by sound fiscal policy, countries using money targets have reduced inflation to near single digits. The relative flexibility in meeting reserve money targets that is seen in practice suggests that the authorities pay substantial attention to other indicators, such as wages, the exchange rate, credit conditions, and inflation itself.

For policy to become more effective, however, greater discretion regarding money targets should be constrained by a clear commitment to price stability. This “constrained discretion” emphasized by Bernanke and others (1999) is the hallmark of inflation targeting, although it can also apply to other regimes with some degree of intermediate target flexibility. It implies that, while policy should not be constrained by mechanical rules, it should be held directly accountable for achieving its end objective (low inflation). Transparency about policy intentions, articulated through a communication strategy that explains policy decisions and helps form inflation expectations, becomes the most effective nominal anchor. On the other hand, unconstrained discretion may greatly limit the effectiveness of monetary policy with little gain in achieving other objectives.

A monetary policy continuum

Depending on the country’s structural characteristics, the appropriate policy framework lies somewhere along a continuum. On one end are regimes that emphasize intermediate targets. On the other are regimes that allow greater flexibility for intermediate targets and emphasize achieving price stability directly (the so-called inflation targeters). In between there is a continuum of regimes that incorporate elements from both ends; many of these could be described as “inflation targeters lite” (Stone, 2003).

When countries are moving away from fiscal dominance but credibility is low, there are important advantages to intermediate targets. Because the relationship between policy and inflation is uncertain, targeting broad money or the nominal exchange rate can help anchor expectations and reduce inflation rapidly. Indeed, the successful stabilization of inflation in postconflict economies like Mozambique and Uganda in the early 1990s and in Burundi, the Democratic Republic of Congo, and Sierra Leone more recently seems to be based at least in part on the role of money targets.

The role of money targets in stabilizing inflation is not to strictly limit policy but rather to signal that stabilization is on track. The evidence from the previous section and the international evidence indicate that countries can be successful in their stabilization process without strict adherence to the money targets.13 What does matter is that the stabilization effort, namely the buildup of fiscal sustainability, stays on track. Money targets have such a “tripwire” role: large and sustained positive target misses indicate a clear policy breach that is often followed by higher inflation.

As inflation stabilizes and credibility increases, the trip-wire role of money declines. As Box 2.4 describes, strict adherence to money targets may have adverse consequences for inflation and output stability. For these reasons, such strict adherence is often not observed, even in countries with successful inflation outcomes. This flexibility also means, however, that money targets may fail to signal or guide policy, as they are only imperfectly related to the de facto policy stance in the short run. This lack of clarity may generate uncertainty and may not help establish or maintain policy credibility. Greater money target flexibility should therefore be complemented with additional innovations in the policy framework.

Box 2.4.The Role of Money

What role can money play in the conduct of a managed float in Africa? We conclude that (i) it can be a useful guide to policy as an intermediate target, particularly as a “trip wire” for major deviations from policy when the independence of the central bank from fiscal and political pressures is uncertain; but (ii) optimal policy is not likely to consist of strict adherence to monetary aggregates; rather, substantial use of other information is also required.

The targeting of monetary aggregates, typically reserve money, as a guide to monetary policy is now nearly unique to sub-Saharan Africa. Most of the rest of the world has transited to some form of inflation targeting or inflation-targeting ‘lite’ (Stone and Bhundia, 2004). However, low-income African countries present characteristics that may make some degree of money targeting desirable: (i) continued risk of fiscal dominance and weak central bank institutions that imply a need for a clear signal as to the direction of monetary policy, and (ii) thin financial markets that make the transmission mechanism hard to understand and may leave a relatively small role for interest rates.

The long-run relationship between money and prices is often presented as a justification for targeting money. While this relationship has broken down in many countries (developed, emerging, and low-income), it has remained stable in others. The argument is that, provided it is stable, controlling one of the variables of that relationship, money, ensures control of the other variable, inflation.

Moreover, many studies conducted in Africa—somewhat in contrast to other regions—find that, historically, deviations from that long-run relationship have led to increases in prices. These studies use vector error correction models that estimate both the long-run or cointegrating relationships (between money, prices, and other variables) and the short-run dynamics that ensure that the system converges back to equilibrium if there are temporary deviations. Money in excess of what is implied by that relationship, “excess money,” Granger-causes inflation (see Mikkelsen, 2008, for an application to Zambia, and Gerlach and Svensson, 2003, for the euro zone).

While of interest, this evidence is not sufficient to conclude that strictly targeting money is the best policy option. Historically, fiscal dominance has been a significant burden for monetary policy in Africa. An implication of such dominance is that money growth was driven mainly by central bank financing of fiscal deficit. In this context, it is to be expected that excess money leads to high inflation. Once fiscal dominance disappears, however, money growth may be driven mainly by shocks to money demand, as it is optimal to accommodate these shocks to prevent unnecessary fluctuations in interest rates. The positive short-run relationship between excess money and inflation would then shrink or disappear. In sum, the empirical relationship between money and prices may not be invariant to changes in the policy regime. This may be the case even if money demand itself remains stable.

Moreover, basing monetary policy on even a stable long-run relationship may not be the most effective strategy. Solely relying on the long-run relationship between money and prices amounts to a minimal information strategy that may be suboptimal because it does not make use of all the available information (Svensson, 2001). These arguments hold even in the absence of shocks or structural breaks in money demand. In addition, money targeting may lead to excessive volatility in inflation and output, as unaccommodated shocks to money demand can generate undesirable interest rate volatility.

Monetary aggregates may have informational content about the state of the economy, however (see Svensson and Woodford, 2003; Coenen, Levin, and Wieland, 2005; and Berg and Portillo, forthcoming, for an application to Africa). Provided money demand shocks are not excessively volatile or persistent and there is a strong contemporaneous co-movement between real money balances and economic activity, monetary aggregates may serve as an economic indicator. In Africa, this role may be important given the scarcity of accurate and timely economic data. Ultimately, though, the role of money as an indicator variable will depend on country-specific characteristics: countries with breaks in money demand, little correlation between money and output, or a relative abundance of economic statistics may find little use for monetary aggregates in the policy framework.

Increased flexibility seems to describe the current frameworks in mature stabilizers like Mozambique, Uganda, and Tanzania (Box 2.5). In some of these countries, the central bank implements a more eclectic monetary policy framework by defining a range of monetary aggregates (rather than a discrete target) and adjusting an interest rate corridor taking into account a variety of real and financial market developments, one of which is money. These countries could be described as informal inflation targeters, in that their policy frameworks are geared toward price stability but the inflation-targeting infrastructure is not fully in place.

Box 2.5.Mozambique and Uganda: Adding Flexibility to Monetary Policy

This box describes the steps taken to add flexibility to the current frameworks in Mozambique and Uganda. In part driven by the gradual deepening of the financial system and large exogenous shocks, these regimes are moving away from strict money targets to a more eclectic framework, with the ultimate goal of implementing an inflation-targeting regime in the medium term.

Uganda and Mozambique have demonstrated a strong commitment to price stability, with core inflation at single digits for five years or more in both countries. Their excellent record reflects responsible monetary policy supported by prudent fiscal policy, in which reserve money ceilings have provided a nominal anchor.

Base money ceilings have recently been exceeded or adjusted upward in both countries in response to an apparent increase in demand for real money balances, due to stronger-than-expected economic activity and demand for currency in rural areas. The ratio of base money to GDP has been generally increasing in both countries for several years, but this process has recently accelerated. Consistent with a shift in real money demand, inflation pressures, other than those stemming from hikes in international food and oil prices, have not surfaced.

The conduct of monetary policy under a managed float has been complicated in both countries by increasingly large foreign exchange inflows. The authorities have reacted by smoothing exchange rate volatility, which has led to the accumulation of international reserves and required more intensive use of open market operations to meet reserve money targets. Mozambique has relied less heavily on open market operations than Uganda, mainly owing to stronger sales of foreign exchange and a less strict adherence to monetary targets. In Uganda, money targets based on monthly averages reduced the authorities’ freedom. The Bank of Mozambique also operates an interest rate corridor by adjusting its deposit and lending facilities, which has helped bestow greater stability on interbank rates, though benchmark treasury bill rate volatility and credit growth remain comparable in both countries.

In view of such developments, the authorities are planning to move toward a more flexible monetary framework, aiming for inflation targeting in the longer term. Both countries are now looking at broader indicators of price pressures in addition to monetary aggregates, and their programs already include money target adjustors based on higher-than-expected growth in currency in circulation. Mozambique is considering introducing bands for monetary aggregates, and coordination between the ministry of finance and the central bank is being buttressed to increase the effectiveness of the monetary program. Uganda has adopted a target on net domestic assets in its IMF-supported program that allows for greater flexibility in money growth, although the base money target remains the operational target for the central bank.

Monetary and Exchange Rate Policy Indicators(Percent of GDP unless otherwise indicated)
Accumulation of central bank net foreign assets1,2-
Change in base money
Nominal exchange rate (12-month appreciation rate)3-21.8-2.7-2.5-2.4
Accumulation of central bank net foreign assets1,
Change in base money10.
Nominal exchange rate (12-month appreciation rate)3-4.3-6.52.313.2
Sources: Bank of Mozambique, Bank of Uganda, and IMF staff.

Cumulative from December of previous year. For month m and year t, change in x is defined as (Xm, tXdec, t –1)/GDPt.

Excludes valuation effect.

Based on monthly exchange rate averages. A minus sign represents a depreciation of the exchange rate.

Sources: Bank of Mozambique, Bank of Uganda, and IMF staff.

Cumulative from December of previous year. For month m and year t, change in x is defined as (Xm, tXdec, t –1)/GDPt.

Excludes valuation effect.

Based on monthly exchange rate averages. A minus sign represents a depreciation of the exchange rate.

Note: This box was prepared by Manrique Saenz.

Finally, some countries have begun to adopt full-fledged inflation targeting (Box 2.6). South Africa has a regime of the sort generally found in a large number of middle-income economies. Much attention is given to the inflation forecast, and policy decisions are clearly explained to the public. By clarifying the objectives of policy and providing a sound anchor for inflation expectations, the inflation-targeting regime has been very successful in stabilizing inflation in South Africa despite substantial exchange rate volatility.

Box 2.6.Inflation Targeting in Ghana

In May 2007, the Bank of Ghana announced its adoption of formal inflation targeting, becoming the first country in sub-Saharan Africa besides South Africa to do so. The move came after several years of preparation and a successful disinflation strategy that brought inflation down from the 40s in 2002 to near 10 percent by mid-2006. Inflation targeting is now an integral part of Ghana’s strategy to accelerate growth and achieve middle-income status by 2015.

The motivations for the move to inflation targeting focused on the advantages of anchoring inflation expectations, particularly given the increasingly weak relationship between inflation and monetary aggregates. Broad money has grown sharply since late 2005, while inflation has continued falling. Ongoing structural changes in the economy have led to an increasingly unstable demand for money. In addition, increasingly focusing expectations on inflation targets should also allow for more flexibility in exchange rate policy. Finally, inflation targeting is seen as improving the accountability of both the central bank and the fiscal authorities, given the joint nature of inflation target decisions.

The institutional and analytical framework developed for inflation targeting provides numerous advantages, particularly in terms of transparency and predictability. The Bank of Ghana, has built the main institutional, analytical, and communications elements of this framework since 2002, although further progress on financial sector development and communications would strengthen it.

  • The 2002 Bank of Ghana Act confirmed the independence of the central bank with the primary objective of price stability.

  • To consolidate freedom from fiscal dominance, central bank credit to the government is limited each year to 10 percent of that year’s total revenue collected, although in practice the government has not resorted to any central bank financing for the past several years. The target range for CPI inflation is set jointly by the government and the Bank of Ghana.

  • The relative stability in the exchange rate between 2004 and mid-2007 has also anchored inflation expectations and appears to be an intermediate operational objective (notwithstanding the cedi’s recent depreciation). The development of market-based liquidity instruments has advanced significantly, although further progress would be desirable.

  • The Bank of Ghana has developed an analytical framework, and the Monetary Policy Committee’s decisions are now informed by the results of an inflation-forecasting model, developed in cooperation with other central banks. The Committee also considers the results of a financial sector stability analysis, which includes stress testing. Surveys of inflation expectations also inform policy decisions.

  • The Bank of Ghana has developed an elaborate communication strategy. The Monetary Policy Committee meets every other month, after which it issues a press release and holds a press conference, under the chairmanship of the governor, at which it explains its announcement and decision. A detailed monetary policy report, including the financial stability analysis, is published.

Initial indications of the effectiveness of the regime are encouraging. Despite the ongoing oil and food price shocks, CPI inflation remained stable at 10 percent from mid-2006 to late 2007, when it ticked up to near 13 percent. Headline inflation has thus remained above the Bank of Ghana’s target range of 7–9 percent for 2007, although core inflation (excluding energy and utilities) has fallen below 9 percent. In explaining its November 2007 decision to increase its prime rate by 100 basis points to 13.5 percent, the Bank of Ghana acknowledged the need to tighten monetary policy in the context of strong domestic demand.

The Bank of Ghana is expected to continue to reinforce the credibility of the regime. The recent increased flexibility of the exchange rate is commensurate with a full-fledged inflation targeting framework. The communication strategy could be enhanced by the adoption of an “open letter” to explain the reasons for potential target misses. Bank of Ghana and IMF staff are also working together to refine the inflation-forecasting model. Ghana’s transition to inflation targeting has shown the advantages and challenges of inflation targeting for a country aiming for middle-income and emerging-market status.

Note: This box was prepared by Marshall Mills and Hans Weisfeld.

Inflation targeting may serve as a benchmark for regimes along the monetary policy continuum (see Box 2.7). While the choice of policy regime depends on each country’s specific conditions, inflation targeting provides many insights about the adequate policy response to economic developments and may thus serve as guidance for policy decisions.

Box 2.7.Inflation Targeting: A Benchmark for Monetary Policy in Managed Floats

Inflation targeting, or elements of inflation targeting, can serve as benchmarks on which to assess current monetary policy frameworks in the region. This box describes key elements of inflation targeting.

Inflation targeting is a monetary policy framework that has five main elements (Mishkin, 2007): (i) the public announcement of short- to medium-term numerical targets for inflation; (ii) an institutional commitment to price stability as the primary goal of monetary policy, to which all other goals are subordinated; (iii) an information-inclusive strategy in which many variables, not just monetary aggregates and the exchange rate, are used to inform policy decisions; (iv) great transparency and communication of plans, objectives, and decisions of the monetary authorities; and (v) increased accountability of the central bank for attaining inflation objectives.

In practice, an essential aspect of inflation targeting is the need for decisive policy actions to meet inflation objectives. Because inflation is partially predetermined in the short run, authorities focus on the impact of current policy decisions on future inflation. Indeed, inflation-targeting regimes are often described as “inflation forecast targeting” (Svensson, 1997). Several implications follow. First, central banks must make the best use of current information to update their forecasts of future inflation. Second, authorities must anticipate the effects of current developments on future inflation and change their policy instruments accordingly. Authorities cannot allow for current developments to have a long-lasting effect on inflation expectations, because the credibility of the policy regime may be questioned. For this reason, policy actions must be sufficiently decisive to influence expectations and keep inflation stable.

Although it is sometimes asserted that certain institutional and technical preconditions are required to implement inflation targeting, the experience of many emerging market countries suggests otherwise (IMF, 2004b). Often-cited preconditions include full central bank independence, a well-developed technical infrastructure to forecast and model inflation, an economic structure under which domestic prices are not overly sensitive to commodity prices and exchange rates, and a healthy financial system (Eichengreen and others, 1999). However, most emerging market countries that adopted inflation targeting in the past 10 years did not meet several of these preconditions. Instead, the adoption of inflation targeting was subsequently associated with improvement in all these areas, which suggests that inflation targeting may encourage the central bank to take a proactive approach to making improvements.

Inflation targeting is nonetheless challenging. The ability to control inflation in the short to medium term requires that policy levers influence economic activity in a way that is adequately understood; a challenging task in low-income countries (see Box 2.2). However, these challenges are a defining feature of any monetary policy, and not just for inflation targeting. The main difference with other regimes is that inflation targeting brings this discussion to the forefront of the policy debate. Under money or exchange rate targeting, the goals of monetary policy are confined to meeting intermediate targets that are (in principle) more readily controllable by the authorities. But the relationship between intermediate targets and inflation is equally difficult to pin down, for the same reasons.

Inflation targeting has important corequisites. Because central banks enjoy substantial discretion in implementing policy under inflation targeting, inflation stability requires that the central bank be perceived as committed to price stability. Otherwise, adopting inflation targeting will not help cement the authorities’ reputation, and more conservative policies will be required to keep inflation in check. Credibility is thus an essential element of inflation targeting, although it can be acquired over time, following the formal adoption of inflation targeting. An additional corequisite is the need for fiscal sustainability. This is critical for the success of any monetary policy regime, but it may be more easily demonstrated in a regime with an explicit intermediate target.

This benchmark is especially relevant for the many countries that are already practicing a regime with very flexible money targeting. These countries may not have the full set of preconditions often considered necessary for inflation targeting, but in effect they have largely forgone the comfort (or perhaps fiction) of simple money targeting. Authorities could more effectively respond to shocks—stabilizing output and inflation while anchoring long-term expectations—if they adopted certain elements of inflation targeting: a focus on adjusting policy instruments to bring expected inflation toward the target; an emphasis on developing an explicit understanding of the monetary policy transmission mechanism; and a public communication strategy for explaining how policy will achieve the inflation target over the medium term.

The question of the appropriate response to rising food prices illustrates the potential advantages of adopting elements of inflation targeting. Food prices have increased substantially over the last year in many sub-Saharan African countries, far outpacing nonfood inflation (Chapter 1). To the extent that price developments reflect world prices or local weather conditions, it is optimal to allow a temporary spike in inflation while taking measures to prevent entrenchment of higher inflation expectations. A clear and plausible public explanation of how food inflation developments alter the inflation profile over the short run, and how the central bank is planning to meet inflation targets over a longer horizon, can do a great deal to anchor expectations during these episodes. The communication strategy recently demonstrated by the South African Reserve Bank regarding recent breaches of target is a good example of such a response.

Role of the exchange rate

The exchange rate can play a potentially useful role in support of inflation objectives in a floating regime. In a number of African countries, the absence of a commitment to a particular level and periodic, if infrequent, adjustments suggest that these might be characterized as regimes in which the exchange rate is being used as an operational target in the service of an inflation objective.14 This approach may make sense when the exchange rate is an important variable in determining inflation and inflation expectations, or if other targets (money, interest rates) are much less effective.

However, the exchange rate needs to adjust when circumstances change. Achieving credibility and transparency with this regime requires that the exchange rate target ultimately be subordinate to the inflation objective. In other words, the targeted level (or rate of change) of the exchange rate must be consistent with the inflation objective. This implies that the exchange rate target must also be consistent with the other policy levers, such as the announced growth of reserve money targets or the short-term interest rate, which have been chosen to steer the economy toward meeting the inflation objectives. Otherwise, tensions will arise between inflation and exchange rate objectives.

The experience of emerging market countries indicates that countries that targeted the exchange rate while simultaneously pursuing short-term inflation objectives eventually abandoned exchange rate targeting.15 In Chile, the use of exchange rate bands in the context of an inflationtargeting regime was abandoned in 1999. Israel’s framework also featured both exchange rate and inflation targets for several years, but exchange rate flexibility was officially adopted in 2005. In both countries, but especially Israel, exchange rate targeting initially helped anchor inflation expectations but it was gradually abandoned as tensions arose between multiple operational targets. These experiences suggest that exchange rate targeting, even if carried out for the purpose of achieving price stability, serves more as a transitional regime.

Moreover, pegging the exchange rate without admitting so greatly reduces policy transparency and may affect the credibility of monetary policy. Because it is difficult to distinguish a peg from a regime where the exchange rate is heavily managed to achieve the inflation objective, for example, the commitment to price stability may thus be hard to assess. Lack of transparency may be particularly undesirable in sub-Saharan African countries, where external shocks can be very large and large real exchange rate movements are required to preserve external and internal balance. In such an environment, lack of policy credibility may lead to inflation expectations no longer being anchored; inflation must fluctuate if the nominal exchange rate is kept constant.

Countries with less direct control of the exchange rate can still try to reduce short-term exchange rate volatility. The fact that real exchange rate volatility is considerably higher in floating than in fixed regimes (the “Mussa” puzzle; Mussa, 1986) could lend support to the pursuit of exchange rate stability as a secondary, subordinate objective, especially if exchange rate fluctuations are not driven by fundamentals (Jeanne and Rose, 2002).

Reducing excessive real exchange rate volatility may help mitigate some of its potential adverse effects on trade flows and investment, particularly since outside of South Africa there are no hedging instruments and derivative markets in the region.16

However, using monetary policy to stabilize exchange rates is challenging. The pursuit of secondary objectives—exchange rate smoothing, avoiding currency misalignments—risks confusing the public about the primary objective of policymakers. The risk of failure must also be considered: unsuccessful attempts to influence the exchange rate could undermine credibility. Another major policy risk is that central banks may not always be able to correctly identify episodes of exchange rate misalignment. Targeting the exchange rate or reducing exchange rate volatility could also cause higher volatility in interest rates and output and persistently higher inflation (Calvo, Reinhart, and Végh, 1995).

Monetary policy is limited in its ability to influence the real exchange rate in the medium term. Experience has shown that the use of monetary policy for purposes other than price stability ultimately leads to higher inflation, which is often difficult to bring down and can have negative consequences for long-run growth. For external competitiveness, empirical studies emphasize the potency of fiscal rather than monetary policy (for Africa, see Elbadawi, Kaltani, and Soto, 2007).

Adjusting to aid shocks is a special case of the difficulties associated with central bank targeting of the real exchange rate.17 If aid results in a corresponding increase in government spending, there may be pressures for the real exchange rate to appreciate. Attempting to contain such pressures may lead to higher inflation, interest rate volatility, or both without impeding the real exchange rate appreciation in the medium run. Ultimately, differing views on what is the appropriate real exchange rate (one based on central bank preferences or one driven by public expenditure) need to be reconciled.

Understanding the monetary policy transmission mechanism

As countries move along the monetary policy spectrum, the need to understand the channels through which monetary policy affects inflation becomes more obvious. Moving backward from final objectives, an understanding of the transmission mechanism helps policymakers decide by how much to change their operational target. A clear grasp of the transmission process also helps clarify the trade-offs between price and output stability. Finally, a clear exposition of the transmission policy to the public can enhance policy.18

In Africa, the transmission mechanism is uncertain and constantly changing (Box 2.2). The recent history of fiscal dominance in many countries still affects inflation expectations. After stabilization, the exchange rate channel is typically quite potent in Africa; credit channels are less so, although they are strengthening as the financial sector develops. While these channels will likely become prominent, it will take time before their precise underpinnings are well understood.

An unfinished agenda for sub-Saharan Africa is the development of modern analytic frameworks for understanding the monetary policy transmission mechanism. Such frameworks have proven useful to a large number of central banks in conducting monetary policy, particularly those targeting inflation. Relatively little such work has been done in low-income countries, though this is beginning to change (Box 2.8).

Box 2.8.Analytic Frameworks for Monetary Policy in Low-Income Countries

Structural analytical models have become an important tool for the analysis of monetary policy in central banks around the world. Because they embody explicit assumptions on the structure of the economy, they help organize a coherent view of the monetary transmission mechanism and assess the likely effects of policy decisions. Current models are derived from solid micro-foundations, and great emphasis is given to expectations in the transmission of policy (Berg, Karam, and Laxton, 2006; and Galí and Gertler, 2007).

Analytical models are better suited than traditional econometric models to analyze policy decisions. The latter are pinned down by past data and may be misleading when there are changes to the policy regime. In addition, although empirical evidence may help, policymakers’ judgment should play an important role in the model’s calibration. Analytical models can also help characterize potential risks to the macroeconomic outlook and the required policy response in alternative scenarios, which is difficult to do in reduced form or partially identified econometric models.

The benefits of analytical models may be particularly pertinent to African countries. Ongoing structural changes in sub-Saharan Africa blur the relationship between policy objectives and instruments, and historical data are both limited and of little use for policy analysis. In particular, the recent move away from fiscal dominance and toward greater policy independence highlights the need for frameworks that assess likely effects of policy decisions under the current regime. Moreover, uncertainty regarding the transmission mechanism can be dealt with explicitly by conducting sensitivity analysis—that is, exploring alternative parameter values to check for robustness.

Little work has been done on using and adapting analytical models to address policy-related questions in Africa, although this is starting to change. This partly reflects economic policies until recently, where efforts were directed at stabilizing inflation, and fiscal consolidation was the key feature. In this environment, the fine-tuning that is associated with monetary policy in emerging or developed countries is less of a concern. In addition, the use of money targets is less amenable to standard models, which both focus on and recommend direct targeting of inflation and the use of the short-term interest rate as the policy instrument.

New and ongoing work has focused on three related areas:

  • The optimal monetary policy response to increases in aid. Analytical models have been developed to look at the optimal degree of reserve accumulation and open market operations when public spending increases as a result of positive aid shocks (Adam and O’Connell, 2006; Berg and others, 2007; Adam and others, 2007; and Peiris and Saxegaard, 2007). Accumulating more reserves when public spending increases can have negative consequences for inflation and private sector growth; some degree of exchange rate flexibility can contribute to the economy’s adjustment to the aid and other shocks.

  • Analysis of current monetary policy in several countries in the region. Although the discussion of policy decisions in emerging markets often relies on analytical models (see Harjes and Ricci, 2005, for an application to South Africa), only recently has this approach been extended to other African countries: Nigeria (Steinberg, forthcoming) and Ghana (Mills and Weisfeld, forthcoming). Given the current state of the economy, analytical models have been used to assess the policy decisions needed to meet medium-term inflation objectives.

  • The optimal role of monetary aggregates in monetary policy. Current research attempts to reconcile the relatively flexible use of money targeting in countries in the region with analytical models that justify some role for money as an indicator variable (Berg and Portillo, forthcoming). This research finds that the relative desirability of “partial” money targeting depends on country-specific characteristics: magnitude of money demand shocks versus shocks to aggregate demand, interest rate elasticity of both money demand and aggregate demand equations, strength of the interest rate channel, and wealth of timely statistical information about the state of the economy.

Choice of operational targets

There is a natural association between moving toward more explicit inflation targeting and moving from monetary aggregates to interest rates as operational targets for policy. The move away from monetary aggregates makes it clear that some form of inflation targeting, rather than money, is the true nominal anchor. Countries like Mauritius and Ghana have explicitly abandoned the reference to money and rely on short-term interest rates to determine policy.

However, operational targets need not change at this time. Even if intermediate money targets play a smaller role in policy decisions, it may still be preferable to have reserve money as an operational target: given financial market imperfections, the short-term interest rate that is under the control of the authorities may not reflect or influence credit market conditions. This approach coincides with current money targeting practices in the region in many ways; in particular, the reserve money growth targets should be implemented with sufficient flexibility to allow target misses and target rebasing as necessary. However, the agenda for making policy more effective is to be more explicit about this flexibility, analyzing how to use it wisely and emphasizing communication and transparency.

Choice of instruments

Most managed floats in Africa have two separate policy instruments: foreign exchange intervention and domestic open market operations. This allows authorities some ability to independently influence the exchange rate and domestic monetary policy even with an open capital account, provided that domestic and foreign assets are imperfect substitutes—a reasonable assumption in the African context.

The scope of independence among instruments should not be overstated, however. First, countries in the region may not be able to continue using two separate instruments if the current benign environment—growing capital and aid inflows, pressures for real appreciation—worsens. In particular, foreign exchange operations may no longer be an option if there are persistent pressures for depreciation and international reserves are low. Moreover, over longer horizons monetary policy itself is unlikely to have a persistent effect on the real exchange rate and hence on the nominal exchange rate, except insofar as it works through inflation.


The time is ripe for evaluating key elements of monetary policy frameworks in many African floating exchange rate regimes, notably the role of monetary aggregates. The trend toward countercyclical monetary policy in the rest of the world is clear (Goodfriend, 2007). But many low-income countries have only recently emerged from stabilization episodes, and there has been insufficient application of modern monetary policy analysis. This needs to change in coming years, as countries look to reduce economic volatility and achieve inflation objectives. In practice, policymakers apply monetary targeting with appropriate flexibility. And African monetary policy regimes are in various stages of transition. However, further clarity about the framework could enhance policy effectiveness.

Monetary aggregates can play a useful role in the conduct of policy, particularly as a tripwire indicating major deviations from monetary and fiscal policies that are consistent with stabilization. This function is most important early in the process of establishing credibility and ending fiscal dominance.19

Monetary aggregates may be less effective in guiding and signaling the policy stance in countries that have successfully reduced inflation and are now considering fine-tuning monetary policy to stabilize the economy while anchoring long-run expectations. Flexibility in money targeting implies that the danger is not adherence to inappropriate targets but rather lack of clarity about the true policymaking process. This nontransparency can impair public communication and the quality of the internal policy dialogue.

The discretion implied by the acknowledgment that money targets do not strictly guide policy should be constrained by a clear commitment to moderate inflation. More generally, countries should consider adopting elements of an inflation-targeting approach—notably, accountability, transparency, and communication.

Of course, countries need to proceed at a pace appropriate to their own institutional capacity. Moreover, uncertainties about the transmission mechanism and the state of the economy, often still-nascent credibility, and the prevalence of supply shocks limit the scope for activist monetary policy, particularly in the near term. Countries must continue to develop domestic financial markets, improve in-house technical capacity, and build appropriate forecasting tools.20 However, the role of prerequisites to moving away from strict intermediate targeting can be overstated: any monetary policy can only be effective if monetary policy is largely free from pressures to finance fiscal deficits and there is some understanding of how monetary policy affects the economy. A major potential benefit of more transparent frameworks would be the emphasis on understanding how monetary policy works in these economies.

There remains an important, albeit different, role for monetary aggregates, in addition to their tripwire function. They may contain useful information about the state of the economy, particularly in Africa, where data on output and other important indicators are often scarce. Difficulties in controlling short-term interest rates may also justify partial targeting monetary aggregates. Ultimately, the optimal degree of adherence to monetary targets during the transition to greater flexibility is a country-specific question that depends on each country’s structural characteristics and pattern of shocks.

There may also be a role for targeting the exchange rate in managed floats. The exchange rate may be useful as an intermediate target, particularly when other elements of the transmission mechanism are very weak and the signal provided by the exchange rate is strong. Moreover, intervention to smooth excess volatility may make sense. However, there are risks, for instance, that the regime will become an unannounced peg; that it will create a one-way bet for capital inflows; and that it may prevent an optimal policy response to unexpected shocks. The risks from heavy exchange rate management are likely to increase substantially with higher capital mobility (Chapter 3). Critically, sustained efforts to use monetary policy to hit real exchange rate objectives are likely to be counterproductive.

Ultimately, the transition to greater flexibility and transparency in monetary and exchange rate policy must be supported by sound fiscal policy. While the recent history of fiscal dominance may still influence inflation expectations, credibility earned in a very favorable global environment will be tested when times turn worse. It is thus imperative that countries build a reputation of fiscal sustainability and the independence of monetary authorities from fiscal or political pressures. That will greatly enhance the effectiveness of monetary policy.

Note: This chapter was written by Andrew Berg and Rafael Portillo. Contributors included Anubha Dhasmana, Jan Kees Martijn, Marshall Mills, Shanaka J. Peiris, Manrique Saenz, and Hans Weisfeld.

This subsection is based partly on a survey of IMF staff country desk teams, who were asked to describe current practices using the typology of objectives, targets, and instruments.

We include in the managed float category all countries that, at least de jure, allow for some exchange rate flexibility, from tightly managed floats to fully floating regimes.

See Stone and Bhundia (2004); Clarida, Galí, and Gertler (1998); and Svensson (2001), among others. For a (small minority) contrary view, see Reynard (2007) and Seitz and Tödter (2001).

The IMF Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER) classifies de facto exchange rate regimes in an attempt to assess whether this is the case in practice; because the purpose here is not classification per se, and because of an interest in looking in depth at the nature of (sometimes heavily) managed floating regimes, this classification is not the focus here.

This is based on a regression explaining money growth in a panel of 19 African money targeters with and without an IMF-supported program, using quarterly data from 2001 through 2006.

De facto classifications are very useful in that they attempt to reveal the true policy regime, which may differ from the authorities’ officially announced arrangements. However, the variety of de facto regimes in practice is reflected in the fact that alternative published methodologies tend to yield very different de facto classifications (Frankel, 2004).

In some of the countries in the first group, the monthly change is equal to 1 percent, not 0, more than 30 percent of the time. This grouping is merely suggestive for several reasons, including the fact that regimes may change through time in a given country.

For authorities to be able to influence both the exchange rate and domestic monetary conditions, they must have two separate effective instruments. The next section further discusses policy instruments.

Unlike regimes with short-term interest rates as operating targets, the predictability of future policy decisions may affect the viability of current exchange rate targets, given the possibility of large gains from exchange rate arbitrage. This suggests that the authorities will face market pressure to change the targets once these are clearly expected, particularly when capital mobility is substantial. It also suggests that information regarding target decisions, including the current target, the frequency of target changes, and the rationale behind target changes, may not be clearly announced. All of these reasons point to a potential lack of transparency in managed floats with such exchange rate operating targets.

A possibly fruitful approach to understanding the role of the exchange rate—but one that is beyond the scope of this chapter—would be to estimate policy rules for exchange rate fluctuations.

See Rose (2000) on the large effects of currency unions on trade. Masson and Pattillo (2004) discuss at length the performance of currency unions in Africa and current prospects for further enlargements.

Carstens and Werner (1999); and Berg and others (2003) discuss the role of money targets during stabilization episodes in Mexico and selected emerging market countries, respectively.

Outside of Africa, this appears to be the case in Singapore, where the authorities adjust the value of the Singapore dollar in response to developments in the economy. Several authors (Parrado, 2004; McCallum, 2007) consider that Singapore’s regime is similar to an inflation-targeting regime, the main differences being the choice of operational target (the exchange rate in Singapore versus short-term interest rates in most inflation-targeting countries) and the degree of transparency in policy decisions.

As a practical matter, many African countries seem to be able to partly stabilize the exchange rate over short periods using exchange rate and monetary policies. However, whether exchange rate smoothing is possible in low-income or emerging countries with relatively open capital accounts and independent monetary policy is an open question. See BIS (2005) for a discussion of the cross-country evidence, and Kamil (2008) for an analysis of exchange rate intervention in Colombia.

The macroeconomic challenges of aid shocks have been discussed in a variety of recent papers and are only touched on in this chapter. See Berg and others (2007); Gupta, Powell, and Yang (2006); Adam and others (2007); Peiris and Saxegaard (2007); IMF (2007a); and IMF (2007b).

The continued role of monetary aggregates in conditionality under IMF-supported programs probably can best be understood in terms of this tripwire function. See Ghosh and others (2005).

For example, a clear lesson from several recent country analyses under the World Bank–IMF Financial Sector Assessment Program is that there is a need to address pervasive excess liquidity in the interbank market and develop better liquidity forecasts.

    Other Resources Citing This Publication