1 Monetary Policy in Sub-Saharan Africa

Andrew Berg, and Rafael Portillo
Published Date:
April 2018
  • ShareShare
Show Summary Details
Andrew Berg and Rafael Portillo 

1 Introduction

Central banks (CBs) in sub-Saharan Africa (SSA) have made great progress over the past two decades in stabilizing inflation, to single digits on average, in the context of greater central bank independence, support from fiscal-based stabilization efforts, and more sustained and stable growth. They have done so by relying on monetary policy arrangements centred, at least de jure, on money targets, often with some form of a de facto exchange rate peg.

In about half of SSA countries, a hard peg provides a nominal anchor. In those countries that are the main focus of this book, however, the exchange rate is now at least partly flexible. Especially in these countries, policymakers are beginning to ask more of monetary policy than the achievement of some basic degree of stabilization, and existing regimes have lacked clear and effective policy frame works. This has affected central banks’ ability to steer financial conditions, respond appropriately to shocks, and avoid policy misalignments. Fully aware of these limitations, many central banks are therefore in the process of modernizing their monetary policy frameworks.

In this chapter we provide an overview of the issues facing monetary policy makers as they modernize. For the most part, we draw on the rest of the chapters in this book, as well as on recent efforts at the IMF to develop a view on some of these issues—efforts in which we were involved.1

Our focus is on SSA countries excluding South Africa. Monetary policy challenges in the latter country warrant a separate treatment; fortunately, in general its challenges are those of many commodity-dependent emerging market economies, about which there is a voluminous literature. For the most part we concentrate on countries with some degree of exchange rate flexibility, though we also briefly discuss the main challenges facing central banks in countries with hard pegs.2

This chapter, and indeed the entire book, represents an effort to bridge the economic and political realities of monetary policymaking in SSA with the lessons from the broader monetary policy literature and experience. Part of the foundation of this bridge is built in Section 2, which takes a historical perspective, describing the evolution of the macroeconomic environment and monetary policy landscape over the past three decades. Section 3 first proposes a set of benchmark principles for effective monetary policy regimes. It then discusses some critical features of the SSA economic environment that have shaped existing regimes and which any application of these principles to SSA must confront. Section 4 reviews the monetary policy landscape in countries with flexible exchange rates, while Section 5 considers the modernization agenda. Section 6 briefly considers hard pegs, particularly the CFA zone, while Section 7 discusses a strategy for using models to study monetary policy issues in SSA. Section 8 concludes.

2 A Brief History of Monetary Policy in SSA

2.1 From Independence to the 1980s: The Breakdown of Overly Ambitious Monetary Policy Regimes

Central banks in Africa began to emerge in their modern form in the 1950s and 60s as the countries regained their independence from European colonial powers.3 Pre-independence monetary arrangements were tightly managed by a set of currency boards, mainly anchored to sterling, the French franc, and the South African Rand. In the post-colonial era, the Rand Monetary Area and the CFA franc zone structures remained intact, even as their members attained political independence, and both continue to operate today with almost the same institutional structures they inherited at independence.

In the face of growing opposition amongst emergent nationalist movements, the currency board arrangements with the British Pound were dismantled and gave way to a set of independent central bank institutions. Established at a time where the dominant intellectual climate in economics favoured strong and centralized development planning, the role of these fledgling central banks institutions was very different from today. Along with other visible manifestations of the state, such as a national army, an airline, and a seat at the UN, a national currency and a national central bank, independent from colonial legacy, were seen as a subsidiary tool of national development.

Their distinctive character began to emerge particularly after the collapse of the Bretton Woods system of fixed exchange rates in the early 1970s. Central banks found themselves administering heavily managed exchanged rates, often in situations of severe rationing, so that parallel markets in foreign exchange were widespread; setting administered interest rates, typically far below market-clearing values; directing the allocation of domestic credit between sectors, in many cases through a state-dominated and highly oligopolistic banking system whose operations were often limited to mobilizing private savings for on-lending to government and SOEs at highly repressed rates; and—crucially—providing direct monetary financing of the budget deficit.

By the early 1980s it was clear that many central banks were being asked to do too much and were failing to deliver on most if not all of these multiple objectives, including the core monetary objective of providing an effective nominal anchor for prices. Many countries across Africa faced difficult external circumstances, including low prices for primary export commodities and external and domestic conflict. However, the highly distorted macroeconomic and monetary policy regimes also exerted a serious drag on economic growth and welfare. High and variable inflation became pervasive. Perhaps even more destructive to the broader economy were the flourishing parallel foreign exchange markets that badly distorted incentives for investment and encouraged widespread rent-seeking across the continent.

Central to this failure was the pressure on central banks to finance fiscal deficits from their own balance sheets. With under-developed domestic asset markets and tight controls on capital flows, and widespread financial repression, the demand for money was relatively inelastic. This presented governments with the scope to mobilize substantial seigniorage revenues, an attractive option where traditional tax revenue mobilization capacities were limited. But as controls weakened, and the velocity of circulation rose, inflation began to rise sharply and fiscal balances worsened.

2.2 1990s: Fiscal-Based Stabilization Efforts

From the mid-1980s to the late 1990s countries began reform programmes, often with exchange rate unifications and movement toward more market-determined, flexible exchange rates, and dismantling of exchange and trade controls (Figure 1.1). As in other developing regions, the number of countries with de facto managed or floating exchange rates in SSA increased by about 50 per cent between 1980 and 1990 (Figure 1.2).4

Figure 1.1.Capital Account Openness (Chinn-Ito) Indexa (Period Averages)

a Excluding countries fixed exchange rate regimes. A higher number indicates a more open capital account.


Figure 1.2.Exchange Rate Classification (Sub-Saharan Africaa)

a excludes CFA Zone countries

Source: AREAER Database, IMF.

The initial conditions of these programmes (heavily managed or even de facto pegged exchange rates, pervasive capital controls, and fiscally driven monetary policy) explain the appeal of a monetary policy framework anchored on the control of money-financing of the fiscal deficit. The logic of the ‘monetary approach to the balance of payments’ applies.5 Typical IMF-Reserve money programming in Africa combined a diagnosis of the stabilization problem that located the fundamental macroeconomic weakness in a lack of fiscal control within an operational framework that targeted domestic credit from the central bank to government. Embedded within broader reform programs aimed at the liberalization of domestic prices, interest rates, and the exchange rate, and supported by substantial donor assistance and official debt relief, reserve money programmes of this kind played a major role throughout the 1990s and early 2000s in restoring macroeconomic stability across the continent.

Other key elements of the two-decade transition were sharp reductions in central bank financing of government and financial liberalizations that eliminated interest-rate controls and introduced competition into the banking sector. With the assistance of IMF-supported programmes, substantial debt relief, and a favourable external environment, domestic credit to government declined from an average of 13 per cent in 1985–95 to 8 per cent in 1995–2005, and has remained broadly at that level to date. The re-establishment of fiscal control provided support for money-based disinflation programs to bring down inflation to single digits (or near single) in the context of higher economic growth and higher inter national reserves by the late 1990s, in line with the experience in other developing countries (Table 1.1).

Table 1.1.Inflation in SSA: 1985–95, 1995–2005, 2005–12
InflationGrowthInternational ReservesInflationGrowthInternational ReservesInflationGrowthInternational Reserves
Standard Deviation18.
Mean (Developing countries)
Note: Excluding countries with exchange rate pegs according to 2013 AREAER.Source: World Economic Outlook Database. Annual data (y/y growth) is used to calculate inflation and GDP growth. International Reserves are in percentage of GDP.
Note: Excluding countries with exchange rate pegs according to 2013 AREAER.Source: World Economic Outlook Database. Annual data (y/y growth) is used to calculate inflation and GDP growth. International Reserves are in percentage of GDP.

Liberalization of direct controls over the commercial banking system helped alleviate the prolonged financial repression (Adam and O’Connell, 2005).6 Real interest rates turned positive in 1995–2005, averaging 5 per cent as compared with about -11 per cent in the previous decade. Interest rate spreads, around 11 per cent, have remained high but are comparable to other developing countries (Table 1.2). Bank deposits and private credit as a share of GDP steadily increased during the last three decades (Table 1.2). These developments also coincide with greater use of open market operations by central banks in the region, all of which have increased the role of market signals and the importance of managing expectations in the implementation of monetary policy.

Table 1.2.Bank-Deposits, Private Credit, and Spreads: SSA (Mean, in per cent)
SSASSASSADeveloping Economies
Bank deposits to GDP16.823.229.337.5
Private credit to GDP11.714.820.528.9
Interest rate spreads11.814.011.39.9
Note: Excluding countries with exchange rate pegs.Source: International Monetary Fund.
Note: Excluding countries with exchange rate pegs.Source: International Monetary Fund.

By the early 2000s, then, countries such as Ghana, Nigeria, Uganda, Kenya, Zambia, and Tanzania were beginning to enjoy sustained growth with low and stable inflation.7 Macroeconomic stability was increasingly accompanied by the deepening and development of domestic asset markets and, in some cases, by moves to liberalize the capital account in order to encourage greater private capital inflows, including into sovereign debt.

3 Challenges for Monetary Policy in SSA

The story so far is one of success. However, it is incomplete. The reduced role for the exchange rate as nominal anchor and increasingly developed financial markets revealed weaknesses with existing policy frameworks. In particular, the money targeting regimes did not provide effective frameworks for formulating and implementing policy. At the same time, ambitions grew for monetary policy to not just contribute less volatility but to play a greater countercyclical role. More generally, effective monetary policy, including smoother functioning of interbank markets and the provision of clearer interest rate signals, became part of the broader financial development agenda.

Before assessing the current state of affairs, it is useful—and more transparent—to present a benchmark for effective policy regimes. In the decades prior to the global financial crisis there was a revolution in the practice and thinking of monetary policy, which started in small advanced economies and then spread to other countries. The IMF (2015a) recently summarized these lessons into seven principles. First, central banks should have a clear mandate, set in the law, and the operational independence to pursue it. Second, price stability should be the primary objective of monetary policy, at least over the medium term. Third, central banks should have a numerical medium-term inflation objective to operationalize the price stability mandate and guide policy actions. Fourth, central banks should nonetheless take into account the implications for output and financial stability when making policy decisions. Fifth, central banks should have an effective operational framework, generally centred on the control of short-term interest rates. Sixth, delivering on price stability requires a forward-looking strategy that maps objectives into policy decisions. And finally, a central element of the monetary policy framework is clear communications, to help explain policy decisions and outcomes and provide guidance about the future.8

Two questions immediately arise when thinking about the application of these principles to monetary policy in SSA. First, have they been made obsolete by the lessons from the global financial crisis? And second, can they really be imported effectively to SSA countries with such different economies and monetary policy challenges from the countries where they were developed over the past twenty or so years?

The global financial crisis has, if anything, strengthened the value of these principles along certain important dimensions. For example, inflation-targeting (IT) emerging market countries performed better during the financial crisis than non-IT countries, including in responding more quickly to the global downturn and avoiding deflationary inflation expectations.9 The prospect of deflation in advanced economies has underscored the importance of medium-term numerical inflation targets to help anchor expectations. In addition, the crisis served to reinforce the importance of central bank communications, in particular the use of forward guidance as a monetary policy tool.

The crisis did, however, starkly underscore the limitations of price stability as the sole focus of central bank actions, and the importance of financial stability. Much of the broader post-crisis policy discussion has focused on how to incorporate tools for macro-prudential and how to integrate them with traditional monetary policy tools in service of financial stability.10 The implications of this debate, especially for Africa and low-income countries more generally, have yet to be fully fleshed out.11 However, we see the new focus on financial stability and macro-prudential tools as important refinements to pre-crisis arrangements, rather than a complete overhaul. Modernizing policy frameworks along the lines described earlier should remain the priority for SSA central banks, even as they pay greater attention to financial stability issues.

One notable change relative to the pre-crisis consensus has been greater experimentation with new instruments in advanced economies, mainly quantitative easing. We see little scope for the use of these tools in SSA CBs, given the lower likelihood of monetary policy being constrained by the zero lower bound on interest rates. This does not imply that SSA CBs limit themselves to a single instrument. CBs in the region, and other developing counties, do (and will continue to) rely on a variety of instruments. These include reserve requirements and sterilized foreign exchange interventions, which should be thought of as a separate tool from short-term interest rates, and used for different purposes. We discuss this issue in more detail in Sections 4 and 5.12

The question of the applicability of these principles to SSA economies is a more central one for our purposes. Serious thinking about monetary policy in SSA implies a reasonably accurate view of how these economies work, and in particular about the effects of monetary policy. The range of uncertainties among economists and policymakers is huge. Does monetary policy even matter for output or inflation, for example? Underlying these questions are even deeper ones about whether standard macroeconomics really applies to economies with such different economic structures.

This standard macroeconomics was developed first on the basis of decades or even centuries of experience of fairly stable institutions and consistent data series, and thousands of research papers, in countries such as the United States and the United Kingdom, and then more recently a still relatively large volume of research and experience in emerging markets. And even in these countries, a strong consensus is hard to achieve. Every major recession in advanced countries is accompanied by a torrent of discussion within the economics profession about how the basic models of economics are broken, and about the most basic facts of monetary policy. See, for example, the vigorous recent debate between prominent US economists about whether higher interest rates will raise or lower inflation, or whether inflation responds to output gaps any more, if it ever did.13

Low-income economies are certainly different. In a typical SSA country, the bulk of the population works in smallholder agriculture, the formal sector amounts to perhaps 10 per cent of GDP, and there is little in the way of manufacturing exports. So the research agenda is huge and the literature sparse. Subsequent chapters in this book look at some of these key differences and their empirical implications, and adapt modelling frameworks that have been successful in emerging markets to capture the key features of low-income countries and analyse their implications for monetary policy. Here we summarize some of the key points.

3.1 The Monetary Transmission Mechanism

Perhaps the most critical question is about the transmission mechanism of monetary policy. An extremely weak or unreliable transmission of monetary policy to the economy might limit the scope for monetary policy to serve as a key policy tool for macroeconomic stabilization.14 Or, perhaps transmission is much stronger from the exchange rate or monetary aggregates to inflation and output than is the case for interest rates, which might have implications for the design of a monetary policy framework.

There are many reasons to think that the transmission mechanism in low-income countries may be different and, in particular, relatively weak. The overall magnitude of the effects on aggregate demand and inflation from monetary policy decisions is likely to depend on the extent of financial deepening. African countries have shallow financial markets, so that changes in financial conditions brought about by monetary policy may directly affect a smaller share of the population. Furthermore, the nature of the policy itself decisively shapes the nature of trans mission, and the opacity of existing frameworks may be undermining the effectiveness of policy. Where exchange rates are heavily managed or the capital account closed, transmission through exchange rates is also likely to be attenuated.

Some policymakers and researchers conclude from this assessment that the transmission mechanism is weak or even non-existent. Mishra and Montiel (2013) argue that the impulse responses to monetary policy shocks derived from structural VARs, the tool of choice for identifying the effects of monetary policy shocks, are typically weak and statistically insignificant in low-income countries.

In our view, this evidence may result from difficulties in applying standard empirical approaches to LICs rather than a lack of underlying transmission. Chapter 6 shows that typical features of LIC data, including short sample lengths, measurement error, and frequent policy regime changes can greatly reduce the power of VARs to uncover the monetary transmission mechanism.

In addition, the policy regime itself strongly shapes transmission, rather than or in addition to deeper structural factors. Monetary policy relies on a clear understanding by financial market participants of central bank actions, both current and likely future (the expectations channel). Such a clear understanding is likely not to emerge under existing arrangements in SSA: the combination of money target misses, noisy short-term interest rates, and incipient communications make it difficult to assess policymakers’ intentions (more on this in Section 4). Under these conditions, the analysis in Chapter 9 reveals that monetary policy decisions have a smaller impact on longer-term rates, inflation, and output, compared to interest-rate-based frameworks, even when policy intentions are the same, and even when the underlying economic structure is supportive of monetary policy effectiveness. The corollary is that we should expect a strengthening in the monetary transmission as the policy framework becomes clearer.

Arguably, the strongest evidence that monetary policy ‘works’ in developed countries comes not from VARs but from the history of the Volker disinflation and the Great Depression. Armed with the experience of these episodes, decades of careful research have gone into producing empirical work that yields the ‘right’ signs. Even in the US, with its uniquely long, stable data series and policy regimes, economists experimented for many years before arriving at acceptable results, e.g. solving the ‘price puzzle’, that inflation seemed to rise after a monetary policy shock, and the ‘liquidity puzzle’, that interest rates tended to rise in response to an increase in the money supply.15 Inspired by this perspective, Chapter 5 looks at the effects of a dramatic tightening in monetary policy in the East African Community in 2011. It finds a well-functioning transmission mechanism, especially in those countries where the stance of monetary policy was communicated clearly, consistent with the previous argument. It also finds that the depth of financial markets is a less clear indicator of the strength of transmission than the clarity of the regime.

It may still be that transmission in LIC is generally weaker and more uncertain than in other countries. This in turn can suggest caution in trying to fine-tune monetary policy. However, this point can easily be overemphasized. First, deep uncertainty about the transmission mechanism is not unique to low-income countries but rather is a general characteristic, perhaps especially of countries implementing new policy frameworks, often in the face of rapid structural change or financial crises. And second, this does not in general justify inaction. Indeed, weak transmission may explain the much larger policy movements that are often observed in SSA countries.

Stepping back, the idea that policy action requires a precise and reliable quantitative understanding of transmission represents an excessively idealized view of the monetary policymaking process. There is a critical element of ‘tâtonnement’ for all countries, including low-income countries: assess the state of the economy and the outlook; adjust policy if it seems too tight or too loose; and repeat. For this process only some confidence about the sign of the effect of monetary is critical. And finally, even a weak transmission mechanism leaves monetary policy to play the role of nominal anchor, and doing so in a way that responds effectively to shocks and avoids generating its own argues for the application of the principles discussed above.

3.2 Supply Shocks and Macroeconomic Volatility

The economies of low-income countries are dominated by supply shocks (Chapters 4 and 11). Indeed, it is difficult to identify a Phillips curve-type relationship in the data because of the dominance of supply shocks, which tend to generate a negative correlation between the output gap and inflation in these countries, see Figure 1.3. These supply shocks critically shape the role of monetary policy.

Figure 1.3.Correlation (at Business Cycle Frequency) between Inflation and the Output Gap, Against Income per Capitaa

a Income per capita (2012) is normalized by income per capital for the US

Source: IMF, World Bank Development Indicators, Haver, OECD.

The so-called ‘divine coincidence’ of monetary policy is that, in the face of shocks to aggregate demand, the stabilization of inflation also serves to stabilize output.16 Supply shocks, in contrast, push inflation up at the same time as they reduce output, presenting a trade-off between output and inflation stabilization. Fortunately, the principles articulated above can help manage this trade-off. Indeed, the difficulties of managing this trade-off, notably in small open commodity-dependent countries such as New Zealand and Canada, were some of the main driving forces behind the evolution in monetary policymaking that is captured in the principles, most notably the emphasis on price stability over the medium-term.17

Many of these supply shocks call for adjustments to the real exchange rate. Developing countries with more flexible exchange rate regimes, including in Africa, tend to do a better job of shielding their economies from the effects of these shocks, thanks to the shock-absorbing role of the exchange rate.18 The challenge under floating regimes is to prevent the fluctuations in nominal exchange rates from spilling over into inflation, especially when a large nominal depreciation is required. In this regard, floats in SSA have a mixed record, with higher average inflation relative to pegs. Of course, a floating regime is not in itself a monetary policy regime, and so countries wishing to reap the benefits of greater exchange rate flexibility must develop well-formulated monetary policy frameworks to keep inflation anchored.

Shocks to international food and fuel prices pose an additional set of challenges. In the African context, food makes up a large share of the consumer basket, so that the direct impact of food price shock is larger. In addition, SSA countries are net food importers on average, and many are net oil importers, so that the inflationary impact from higher international prices could be compounded by the real and nominal depreciation required for external adjustment.19 These shocks have therefore been a source of inflationary pressures, especially during 2007–08 and 2010–11. One complication, however, is that the direct effect of these shocks often masks underlying monetary policy misalignments, which then amplifies the overall inflationary effect. This was the case in Kenya during 2011 (Chapter 15).

Domestic supply shocks are an even larger source of inflation volatility. This is because the agricultural sector is heavily exposed to weather-related shocks. One implication is that inflation is inevitably more volatile in low-income countries—SSA countries included—with the larger volatility reflecting supply-side changes to relative food prices (see Chapter 11). Much of this volatility is unlikely to disappear even as countries modernize their policy frameworks.

Capital flows are an additional source of external shocks. SSA countries are less integrated with global capital markets, which, all else being equal, suggests less exposure to shocks stemming from the capital account.20 However, the experience of Zambia during the global financial crisis, which is discussed in Chapter 17, shows SSA countries are not immune to capital flow reversals, and that the latter can have a large impact on domestic financial systems and the economy. In addition, SSA countries are becoming more integrated, as can be attested in the growing number of countries that have tapped international bond markets, many for the first time, in recent years.21 Greater exposure to these shocks is thus to be expected in the future.

3.3 Fiscal Policy as a Source of Volatility and Pressures on Monetary Policy

Fiscal dominance—where the need to finance the government deficit through money printing determines the rate of inflation—remains a fundamental challenge to monetary policy in only a few countries in SSA, after the progress described in Section 2 above. In a much larger group of countries, however, fiscal policy can greatly complicate the conduct of monetary policy. Central banks in Africa must contend with highly volatile and pro-cyclical fiscal policy. Sometimes the source of fiscal volatility is a high dependence on revenues from the commodity sector, and the lack of binding fiscal rules to ensure inter-temporal smoothing. In other cases fiscal pro-cyclicality stems from the political cycle. Certain features of African economies, for example the large share of the population that lives on their current income, amplify the effect of these shocks on aggregate demand.22

Even if fiscal dominance is a (not-so-) distant memory, the volatility and pro-cyclicality of fiscal policy creates other forms of fiscal pressures on monetary policy. One stems from the cost of monetary operations.23 This is a source of contention with the government, particularly in cases where the financial system is in a situation of structural liquidity surplus, for example due to sizeable interventions in FX markets, and a legacy of past quasi-fiscal operations have left the central bank with low or negative net worth. In this case sterilization operations, which are necessary to maintain an appropriate policy stance, can sharply reduce central bank profits (seigniorage). In principle, this can be resolved simply by recapitalizing the central bank, for example through the transfer of government bonds. But concerns that the Treasury will use the opportunity to look into the CB’s operating expenses, or the mistaken belief that the central bank should not make losses, can often result in a monetary policy stance that is more accommodative than optimal.

Another type of pressure occurs when the central bank does not tighten policy as aggressively as it would like to, out of concern for the effect of that policy on fiscal solvency. Pressures of this type are likely to materialize in regimes, such as IT, in which the central bank takes direct responsibility for short-term interest rates. Avoiding this type of pressure may be one possible reason why many central banks in SSA have yet to formally adopt interest-rate-based frameworks, and why those that do often implement changes to the policy stance without changing the (highly visible) official policy rate.

3.4 Management of External Revenues and the Coordination of Fiscal Policy and Central Bank Operations

Central banks in SSA play an important role in managing external government revenues, such as aid and commodity windfalls. As the government’s banker, the central bank helps manage the associated foreign exchange, and in its monetary policy-making function it manages the domestic money creation that results from these foreign exchange transactions. In principle, there is a benchmark of separation: foreign exchange from aid or commodity windfalls is sold into the market or to a fiscal entity (such as a sovereign wealth fund), while monetary policy is set through a policy interest rate, and any domestic money supply implications of the foreign exchange transactions are automatically sterilized. In practice, however, the central bank management of foreign exchange frequently becomes entangled with monetary policy.

To take one important example, Berg et al. (2007) document how, during aid surge episodes in several African countries with managed floats (Ghana, Mozambique, Tanzania, Uganda), concerns about real appreciation resulted in large accumulations of reserves. This policy response may have helped contain the appreciation pressures. But it also resulted in a peculiar situation in which the authorities tried to use the aid twice: once to increase government spending with the domestic currency counterpart to the aid inflows, and once to increase the stock of reserves with the dollars. The private sector was crowded out as a result, mainly through higher interest rates (when the accumulation was sterilized) and in some cases also through the inflation tax (when otherwise).24

The underlying general point is that central bank active management of foreign exchange reserves is a sort of quasi-fiscal policy that is closely related to other aspects of fiscal policy. This raises the issue of whether greater coordination of reserve policy with fiscal policy could help improve macroeconomic outcomes. It is difficult to see how this coordination can take place without affecting central bank independence, however. We return to the closely related issue of exchange rate management in Sections 4 and 5.

4 The Current Monetary Policy Landscape in Countries with Some Exchange Rate Flexibility

Many policy changes have been institutionalized in SSA through reforms cementing central bank independence and the adoption of new central bank charters. The majority of the central banks in the region have de jure (legislated) independence,25 and their de facto independence has been on average (0.26) very close to the developing countries’ average (0.25), using the measure in Lucotte (2009).26 Moreover, 70 per cent of SSA countries had accepted Article VIII of the IMF’s Articles of Agreement by the late 1990s (more than 90 per cent as of 2012), committing to refrain from imposing restrictions on payments and transfers for current account transactions and to refrain from discriminatory currency arrangements or multiple currency practices.

The de jure policy regime in place in most countries is best characterized as a hybrid regime (IMF, 2008, 2015a). An overview of the objectives and targets of monetary policy in the region reveals a set of managed floaters with a variety of conventional-looking objectives (price and exchange rate stability), but with money aggregates still present as both operational and intermediate targets (Table 1.3).

Table 1.3.De Jure Monetary Policy Frameworks in Sub-Saharan Africa
RegimesPolicy ObjectivesIntermediateOperational TargetMain Instruments
Pegs (23)Stability of the exchange rate regime (23) Price stability (23) Economic growth (12)Private sector credit (1)Exchange rate (23)Open market operations Foreign exchange sales
Money targeting (18)Price stability (all countries) External competitiveness (5) Exchange rate smoothing (12) Economic growth (9)Monetary aggregates (16)Reserve money (18)Open market operations (17) Foreign exchange sales (18)
Inflation targeting (3)Price stability (all countries) External competitiveness (1) Exchange rate smoothing (1)Interest rates (3)Open market operations (3) Foreign exchange sales (3)
Source: Regional Economic Outlook 2008; International Monetary Fund.
Source: Regional Economic Outlook 2008; International Monetary Fund.

With this brief overview, we now dig somewhat deeper to look at salient issues with the current state of monetary policy regimes in SSA, using the above principles as an organizing device.

4.1 Legal Frameworks and Operational Independence

The stabilization efforts post-1980s were supported by the adoption of new legal charters in many central banks in SSA. Assessments of central bank independence, however, show many SSA countries lagging behind richer countries. Ghana provides a case in point. A new Act adopted in 2002 established price stability as the central bank’s primary objective, granted operational independence, and created a monetary policy committee. The Act did not, however, prohibit the CB from lending to the government, nor did it provide MPC members with sufficient tenure protection. These two issues have come into focus during the recent period of high inflation in that country.

Adherence to existing legal frameworks has also been uneven. Though measures of de facto independence are more difficult to estimate, there is plenty of anecdotal evidence. In Zimbabwe the adoption of a new charter granting greater independence to the CB preceded the complete loss of monetary autonomy and the rise in inflation—and subsequent hyperinflation—in that country. Even in countries with more stable inflation, deviations from legal limits to direct central bank financing are common, which attests to the pressures that central banks continue to face.

4.2 Price Stability, the Medium-Term Inflation Target, and the Pursuit of Other Objectives

SSA CBs have bought into the idea that price stability is the primary goal of monetary policy, at least de jure. In many countries, however, the primacy of price stability remains to be established. Many CBs continue to pursue other objectives, for example supporting growth, financial deepening, or external competitiveness. This multiplicity of objectives and lack of clear hierarchy among them typically results in erratic policies, although to a smaller degree than in the past: the monetary stance is loosened, for example to support financial deepening, only to be tightened later once inflationary pressures appear.

This state of affairs is most visible in the central role that the exchange rate plays in policy frameworks of many SSA countries, including those with de jure exchange rate flexibility. Though some attention to the exchange rate is inevitable given its importance for inflation dynamics, in some countries exchange rate stability often takes precedence over price stability (Figure 1.4). The exchange rate serves as the de facto anchor, at least temporarily, and operations aimed at influencing the exchange rate end up determining the stance of policy, for example through the use of unsterilized interventions in the FX market. This risks removing the buffering role of the exchange rate and creating exchange rate misalignments, while the disconnect between the de jure and the de facto frameworks undermines the credibility, transparency, and effectiveness of monetary policy.27

Figure 1.4.De Facto Exchange Rate Classification in SSA

Note: Hard peg includes no separate legal tender, currency board, and conventional pegged arrangement; soft peg includes stabilized arrangement, crawling peg, crawl-like arrangement, pegged exchange rate within horizontal bands, and other managed arrangement; flexible includes floating and free floating.

Sources: AREAER and IMF staff analysis.

There are several related factors that account for the policy confusion. First, with the exception of IT countries, most central banks in the region lack an explicit, medium-term, inflation objective that can discipline policy and operationalize the pursuit of price stability.28 In its absence, policy is more likely to be driven by political pressures, recent events, or the pressing issue of the day. Second, even if the primacy of price stability is recognized, central banks typically lack a strategy for mapping objectives into policy decisions or for taking other objectives into account in a way that does not undermine price stability. Third, operational frameworks are not in line with international best practice, which obscures the actual stance of policy and facilitates deviations from policy intentions. We next discuss the last two points in more detail.

4.3 Operational Frameworks in SSA CBs

Reserve money targeting (RMT) remains the de jure operational framework of choice in SSA, in contrast with the now standard practice of setting operational targets on (and controlling) very short-term interest rates adopted by most advanced and emerging market central banks.29 This reflects in part the legacy of IMF-supported programmes, which emphasize targets on central bank balance sheet items as part of their conditionality, and which played an important role in the stabilization of inflation in SSA.

As discussed in Chapter 8, money targeting is implemented very flexibly, with frequent economically significant misses of money targets. These misses mainly seem to represent accommodation of money demand shocks, though some may involve policy shifts. Flexible implementation of money targeting is also evident in the process of adjustment after misses. In ‘textbook’ money targeting, where a constant growth rate of money serves as the ‘nominal anchor’, deviations from targets would be undone in subsequent quarters as the actual stock would be brought back to the predetermined target path. This does not seem to be what happens, however. Rather, the new targets themselves tend to accommodate, at least in part, deviations from previous targets. There is no sign that actual money growth itself moves so as to reduce earlier deviations from target. There is also little sign that inflation responds to these misses, at least in countries with inflation below the low teens.30

More recently, many CBs have introduced policy rates to signal the stance of policy, but deviations between policy and actual rates are common, and tensions between money targets and interest rate policy are inevitable. RMT can lead to highly volatile short-term interest rates, as the authorities respond partially and unpredictably to money demand shocks.31 In addition, tensions between money targets and desired interest rate outcomes frequently lead to complex regulations and interventions in short-term financial markets and a multiplicity of short-term interest rates. All this discourages financial market development (Table 1.4).32

Table 1.4.Characteristics of Deviations from Reserve Money Targets for Selected SSA Countries
CountryShare of observations with absolute deviations bigger than 1% (in per cent)2.5th percentile (in per cent of the target)87.5th percentile (in per cent of the target)
Source: IMF Staff Calculations.
Source: IMF Staff Calculations.

In addition, RMT makes the stance of policy noisy and difficult to interpret, both by financial market participants and the central bank itself; this topic is discussed in Chapter 9. Not all money demand shocks are accommodated, so that interest rates are a volatile and imperfect indicator of the current and expected stance of policy. Greater de facto flexibility vis-à-vis money targets reduces this volatility but at costs of greater discretion and opacity about the true operational framework. Of course, not all deviations from target represent accommodation of money demand, but it is very hard to tell in any particular situation. The effectiveness of the operational framework is hampered as a result.

An additional layer of complexity is brought about by recurrent interventions in FX markets, which are the main tool for managing the exchange rate in most SSA countries. There is often insufficient coordination between interventions in FX markets and other operations. As a result, interventions influence the stance of policy in unintended and undesired ways.

Given all this flexibility, the difficulty in inferring the stance of policy from the money targets or target misses, and the failure of money targeting itself to provide a nominal anchor, how are we to understand these policy regimes? The answer seems to be that these countries tend to practice an opaque version of ‘inflation targeting lite’, in which decisions about the setting and achievement of the money targets themselves depend on progress relative to inflation, output, exchange rate, and in many cases other objectives.33

One of the effects of the opacity of these regimes, in addition to poor trans mission of monetary policy, is that they break the important separation between policy design and policy implementation. In most CBs outside SSA, the former is typically determined by a monetary policy-making committee, following input from the staff of the forecasting, research, or economics teams, whereas the latter is done by the trading desk. Under RMT, it is typically the operations staff who decide whether to hit or miss targets partly for technical reasons, for example not to avoid disrupting money markets. And yet the decision to miss targets has implications for the stance of policy, even if there is no consultation with the monetary policy committee. This creates confusion about the division of labour and governance structure within CBs.

Some of this opacity may be desired by central banks aiming to avoid public responsibility for policy decisions (‘We are not setting interest rates so high; it is the markets. We are just following our monetary programme.’) In some particular instances this may be a second-best response to political pressures, whereby technocrats can hide behind the obscurity of the regime to conduct policy. However, international experience, and also that of a handful of countries in SSA that have made the most progress with regime reform, such as Uganda (Chapter 2), is that this is very much second best; the effectiveness and independence of policy is best served by adherence to the above principles, even in SSA.

4.4 Forward-Looking Strategy and Communications

As central banks in other regions have increasingly focused on the medium-term inflation outlook, a forward-looking strategy that guides policy decisions and communications has become the defining feature of policy. To a large extent, such a strategy is missing in many SSA CBs. This is not surprising, given the lack of numerical medium-term inflation objectives and the limitations of the operational framework. A policy framework aimed at stabilizing expected inflation is valuable for several reasons. Such a framework can lead to preemptive action that is less costly than falling ‘behind the curve’, as discussed in Chapter 19. Of course, it is difficult to have confidence in a particular inflation forecast, given all the uncertainties. Even so, aiming to stabilize the inflation forecast is a way of organizing many complex considerations into a simple narrative. The forward-looking policy process involves developing and communicating an understanding about why the economy is where it is now, in terms of shocks and imbalances, and then projecting how policy should react and how these will unwind.

In the case of SSA, the lack of clear strategy is most visible when thinking about how to respond to large external supply shocks, for example the international food and fuel commodity price surge of 2007–08. As discussed above, policy has long settled on the adage that central banks should accommodate first-round effects but prevent spillovers from these shocks into broader wage and price setting (second-round effects). However, it is far from clear what the above policy advice implies for money targets. Should these be missed, and if so, in which direction? And how is the missing of money targets meant to influence wage and price dynamics in response to these shocks?34

In practice, the framework itself becomes a handicap for articulating a clear policy response. For instance, in Zambia, concerns with money targets in the aftermath of the global financial crisis resulted in excessively tight monetary policy at a time when domestic banking systems were under stress. These policies were later reversed, but it can be argued that the policy framework amplified the initial impact of the crisis.35 The case of Zambia is discussed in more detail in Chapter 17.

An additional and related factor is insufficient internal analytical and fore-casting capacity in many central banks, which limits the staff’s ability to provide senior management with an assessment of the state of the economy, sound macroeconomic forecasts, and policy recommendations. Most SSA CBs have yet to develop in-house modern macro models that are standard in most advanced economies and emerging market CBs, though the issue of which models to use for these countries is an open question, as we discuss in Section 7. The absence of models adds to the lack of a clear quantitative view within central banks on how monetary policy is transmitted to the economy.

Communication policy is yet another area of modern monetary policy in which SSA central banks have much catching up to do. This is reflected in the low score most African CBs have in measures of central bank transparency (Figure 1.5).36 For example, very few CBs publish a medium-term inflation outlook or provide a forward-looking assessment as a basis for their policy decisions. Again, the lack of a clear communication strategy goes hand in hand with the lack of clear monetary policy frameworks. Perhaps not surprisingly, CBs that have adopted inflation targeting regimes, such as Ghana, have made important progress in clarifying their policy announcements.37

Figure 1.5.Measure of Central Bank Transparency

Note: The de jure transparency index was developed by Dincer and Eichengreen (2014). It ranges from 0–15, and is the sum of scores to questions ranging from political, economic, procedural, policy, and operational transparency.

Sources: Dincer and Eichengreen (2014) and IMF (2014d).

5 Modernizing Monetary Policy in Sub-Saharan Africa

SSA central banks are well aware of the limitations of their existing frameworks and are looking to improve along the various dimensions we have discussed. Ghana was an earlier adopter of IT, though the experience there has been mixed. Uganda was next in line; that country’s experience with IT is discussed in Chapter 2. Several other central banks, including Kenya, have explicitly discussed the possibility of adopting IT, even if they have yet to formally commit. Many other countries, while not explicitly considering the move to IT, are working to improve their operational framework by giving more prominence to policy interest rates, and by improving the design and use of open market operations and standing facilities. And many central banks are investing in their internal capacity, for example through technical assistance programs developed jointly with the IMF.38 SSA CBs are also increasingly exchanging views on these and other topics, through peer-to-peer events and regional fora.

Although there is no one-size-fits-all approach to monetary policy modernization, central banks can learn from the experience of many countries outside SSA, as well as from early movers within the region. One key lesson is that progress requires sufficient operational independence and a sufficiently clear central bank mandate for price stability, even if adherence to these two principles is always work in progress. Building and maintaining political commitment is therefore critical. The central bank has a leading role to play in building the necessary consensus for reform.

Another lesson is that central banks can make progress in a number of areas simultaneously. Progress can be self-reinforcing: the development of analytical capacity is more likely to impact policymaking if it is consistent with the way policy is designed and implemented, which requires clarity about the strategy and the operational framework. The adoption of an explicit numerical objective can provide an impetus to investing in analytical capacity and the communication strategy; while an effective operational framework can make central banks more comfortable about explicitly committing to an inflation objective. These synergies call for a comprehensive approach to reform.

A related issue is whether to explicitly adopt a new regime, namely ‘IT’, and if so whether to do so at a specific stage of the modernization process. What the international evidence corroborates, and the above discussion implies, is that countries do not need to satisfy a strict number of preconditions before they can adopt IT.39 If anything, the opposite is true. A clear framework is more conducive to reform. This should not be surprising in view of the observation above that current RMT frameworks already amount to an obscure form of IT ‘lite.’

Central banks are naturally conservative institutions, and some may be concerned with their ability to deliver on their inflation objectives, in which case both reforms and the adoption of new regimes may be gradual. And the evidence does suggest that the announcement of IT per se does not generally yield immediate benefits, for example in terms of better-anchored inflation expectations or a lower sacrifice ratio, i.e., the cost of reducing inflation in terms of output.40 However, there is substantial evidence from emerging markets that the benefits of IT accrue over time, as banks learn by doing and gradually gain credibility and as observers come to understand the regime, all this despite initial errors in the achievement of the inflation target.41

A critical aspect of modernizing monetary policy frameworks relates to the questions of how to apply forward-looking monetary policy effectively in the SSA context, along the lines of the principles briefly articulated in Section 3. Here, we briefly review two of the issues, corresponding to some of the most important challenges discussed in Section 3.42

5.1 Responding to supply shocks

Adherence to the above principles of monetary policy, particularly the emphasis on a clearly communicated forward-looking strategy to achieve the inflation objective over the medium term, should help smooth the adjustment to supply shocks. It is sometimes argued that core inflation—typically defined as inflation in the consumer price index, but excluding some key goods that are particularly subject to transitory supply shocks, notably food and fuel—may better reflect the sort of underlying pressures to which policy should respond. However, in SSA countries, food prices are such a large part of the CPI that explicitly targeting core inflation may not serve to anchor the public’s overall inflationary expectations.43 There is an important commonality between emphasizing current core and expected headline inflation: in both cases, there is less need to respond to transitory supply shocks, as long as expectations are well-anchored. However, not all non-core shocks are temporary (for example food prices can also respond to aggregate demand), so the explicit targeting of future headline is perhaps more robust, as well as keeping the focus on the index most people know.

It is a sort of a central bank (and IMF) folk theorem (with essentially no mention in the academic literature) that policymakers should let the ‘first-round’ effect of shocks pass through and react only to ‘second-round’ effects. As Chapter 10 describes, this policy can be difficult to implement in practice, when examined closely. It is difficult because even the first-round or impact effect of a shock (i) depends on the particular shock and the structure of the economy, including the country’s integration with international capital markets; and (ii) even for a given shock is a property of the entire system, depending among other things on the monetary policy reaction function. Thus, the first round depends on how policy reacts to the second round.

Even with the best monetary policy framework, the prevalence of supply shocks speaks to the limits to monetary policy. Monetary policy should not be asked to influence relative prices, for example of food relative to non-food, which can experience significant and persistent movements (Chapter 11). And given nominal rigidities and the desire to minimize output fluctuations, there are limits to the extent to which monetary policy can eliminate the temporary inflationary effects of supply shocks and, given the prevalence of these shocks, the volatility of inflation itself.

5.2 The Role of the Exchange Rate

One perhaps particularly thorny question for many SSA CBs with flexible exchange rates relates to the role of exchange rate management and FX interventions.

The international experience shows that countries that modernize their monetary policy frameworks also move towards more exchange rate flexibility. Indeed, that has already been the case in SSA. The de facto anchoring role of the exchange rate is diminished in favour of a more explicit focus on price stability. Foreign exchange interventions do not disappear, however, as can be seen in their wide spread use in many emerging market CBs with explicit inflation targeting regimes. These countries are attempting to influence the real exchange rate and external financial conditions facing their country, while maintaining their inflation objectives. Given these trends, it is very likely that some degree of exchange rate management will endure in SSA.

There has been growing interest in understanding how FX interventions fit in EM’s monetary policy frameworks, an issue we tackle in Chapter 13.44 The focus is on sterilized interventions, which ensure that the central bank retains control of short-term interest rates as the monetary policy instrument used for domestic stabilization. For sterilized interventions to serve as a separate instrument, they must operate through a different channel, in this case the portfolio balance channel.

Interventions in IT countries and others that adhere to the principles articulated above are guided by a different concept of exchange rate stability. Rather than stabilizing the exchange rate so that it serves as a nominal anchor, the objective is to respond to large and costly deviations of the exchange rate from its medium-term equilibrium value, for example because of risk-on risk-off episodes (associated with sizeable capital flows) or other temporary factors unrelated to fundamentals. Given the uncertainty associated with the assessment of these deviations, central banks tend to lean against the wind, that is to intervene when the exchange rate is appreciating or depreciating more rapidly than usual, but without targeting a specific level.

The possible benefits of an active FX intervention policy will have to be weighed against the costs. In practice, it is not easy to identify exchange rate misalignments; over-stabilization would reduce the exchange rate’s shock absorbing role. Frequent interventions can undermine the clarity and coherence of the regime and raise doubts about the primacy of the inflation objective, which can be especially dangerous when central banks are building credibility. Finding the right role for the exchange rate will continue to be an unsettled and evolving issue in many SSA countries. However, improving monetary policy itself along the lines discussed above requires, and in turn should encourage, exchange rate flexibility and more coherence between exchange rate intervention and the monetary policy regime.

6 Hard Pegs and Currency Unions in SSA

A sizable share of SSA economies are members of the CFA franc zone (CEMAC and UEMOA, 16 countries in total), with their currency pegged to the euro, while others are members of the Common Monetary Area (Namibia, Swaziland, Lesotho) and have their currencies pegged to the South African Rand.45 Other African countries are not part of monetary unions but have hard pegs, e.g., Cape Verde. We briefly review the main issues for these central banks, with a focus on the CFA zone.

Hard pegs to strong currencies offer countries a strong nominal anchor, which typically results in low inflation. The CFA zone is no exception, as countries in that region have lower average inflation than their SSA peers.46 In this sense, CBs in countries with hard pegs are importing the credibility of the monetary institutions they are pegging to (the ECB, in this case).

The cost is twofold. First, central banks lose much control over monetary policy. In the case of the CFA zone the two central banks may retain some monetary control due to limited capital mobility in that region. Even then, however, maintaining the peg, among other things by preserving a sufficient level of international reserves, still takes precedence over all other considerations. Other policies are therefore necessary to achieve macro stability (fiscal, or even macro-prudential).

Second, under a hard peg, adjustments to the equilibrium real exchange rate can only come about through changes in the price level, which implies that variations in inflation are to a large extent necessary for external adjustment and beyond the control of the monetary authorities (see Chapter 20 for a discussion of inflation dynamics in the CEMAC region). In addition, not all real exchange rate adjustments are equal. Achieving large real depreciations is very costly and difficult, given the need for prolonged deflation, and in most cases countries are forced to abandon or adjust their peg. The CFA zone was confronted with such a scenario in January 1994 when, after a prolonged period of real exchange over-valuation, the CFA franc was devalued by 50 per cent, the first and, so far, only nominal adjustment to parity since 1948. The resulting devaluation of the CFA franc is still remembered to this day. Preventing overvaluation of the currency is therefore of the utmost importance, which puts additional onus on a stable fiscal policy.

While the CMA and CFA franc zone pre-date independence, the idea of common regional currencies (and even a continental currency) has long been a goal of various regional groupings, including the Africa Union.47 Of these, the putative East African Monetary Union (consisting of Kenya, Uganda, Tanzania, Rwanda, Burundi, and, possibly, South Sudan) is at the most advanced stage. The enabling legislation—the equivalent of Europe’s Maastricht Treaty—was ratified in December 2013, with the currency union scheduled for 2024. Unlike the CFA or CMA arrangements, the ambition of EAMU is to create a monetary union among a community of equals with a common currency that is guaranteed by neither a regional nor an international hegemon. As with the Eurozone, the plan is that the common currency will float so that the nominal anchor will be provided by the credibility and commitment of the supranational central bank’s monetary policy. It remains to be seen if this monetary union will be realized.

7 A Modelling Strategy for Analysing Monetary Policy Issues in SSA

Thus far in this chapter, we have made a number of empirical, analytic, and policy claims without reference to an explicit analytic framework. This can be useful, but there is great need for models to undertake policy analysis in SSA CBs. In our view, these models must meet two criteria. First, they must reflect modern thinking on monetary policy, drawing on both state-of-the-art macro theory and current practice in central banks in advanced and emerging markets. Second, they must be tailored to address key low-income-country specific issues. Despite the importance of the topic, there has been very little work in the academic and policy literature tailored to low-income countries.

As can be seen in many of the chapters of this book, our general approach has been to exploit the large literature based on fairly standard New Keynesian economic models, appropriately modified to capture key issues and features for SSA countries. This immediately presents one fundamental prior consideration, which is whether models based on nominal rigidities are useful ways to think about monetary policy in SSA low-income countries. It might be argued, for example, that these countries are dominated by flex-price informal markets and few explicit indexation mechanisms. Gali (2015) cites two broad lines of evidence in support of this sort of model as applied to advanced economies: micro data-based analysis such as Bils and Klenow (2004) that demonstrate price rigidities, and VAR evidence that monetary policy affects output, such as Christiano, Eichenbaum, and Evans (1999). Unfortunately, such micro evidence has not been analysed in low-income countries, as far as we know, presumably largely because the data are hard or impossible to find. And as discussed in Section 3 above and in Chapter 6, VAR-based evidence on the impact of monetary policies is even less reliable than elsewhere.

However, we believe that this (the assumption of nominal rigidities) is the right place to start for several reasons. First, the broader evidence on transmission, discussed in Section 3 above, is consistent with the basic features of these sorts of models. Second, major nominal experiments such as the monetary policy tightening episode examined in Chapter 5, and the devaluation of the CFA franc in 1994, had evident real effects. The large CFA step devaluation mentioned earlier—an understudied episode from this perspective—led to an almost-equivalent and highly persistent movement in the real exchange rate, with clear real effects.48 For developing countries more broadly, the nature of the nominal exchange rate regime matters for the real effects of supply shocks, again consistent with this sort of framework.49 In terms of underlying mechanisms, even less research has been done. But a number of drivers of nominal rigidities seem plausible in SSA: information-based lags in price adjustment, as in Mankiw and Reis (2002), seem if anything more plausible in the relatively information-scarce SSA environment. Public wage and price setting are also relatively important, including with a likely direct influence on private formal-sector wages.50

In the rest of this book, we draw on two sorts of models. The first type, which for the most part are covered in the chapters in Part II of the book, are structural models that are designed to better understand the mechanisms and policy tradeoffs facing SSA countries. The models are structural in that they are fully derived from micro-foundations and are best focused on particular policy issues. One of the most significant sets of differences between SSA countries and the canonical model is the nature of the policy framework, and in this context of the monetary policy reaction function. As we have already discussed, many countries have targeted money; Chapter 8 attempts to understand this practice as a reaction to fragmented financial markets and poor high-frequency information on output. Chapter 9 analyses the implications of typical SSA money-targeting operational frameworks for the transmission of monetary policy. Chapter 13 makes the case that managed floats are best understood in a model with two reaction functions, a Taylor-type rule for the domestic interest rate and another for the exchange rate, with sterilized interventions as the instrument.

A second major distinction in SSA economies relates to the supply shocks discussed above. Chapter 10 comes to grips with the meaning of ‘first-round’ effects of commodity price shocks, while Chapter 11 looks at the implications of subsistence for inflation and monetary policy. Finally, the fiscal/monetary inter actions in the case of public sector windfalls are explored in a dynamic stochastic general equilibrium (DSGE) model in Chapter 12.

As mentioned earlier, many authors have emphasized lack of financial depth as a distinctive feature of developing countries for monetary policy.51Chapter 17 develops a DSGE model with a banking sector to analyse the impact of the global financial crisis on Zambia. It finds that monetary policy added unnecessarily to macroeconomic volatility, for reasons related to the money targeting framework. It also argues, however, that even well-designed and implemented policies may not be able to do much to resist the volatility associated with the sorts of shocks encountered by Zambia during the crisis.

Finally, a lack of credibility is a common feature of central banks in SSA, a feature they share with many countries without a long history of successful monetary policy. Chapter 19 presents a model of endogenous policy credibility to study the optimal path of disinflation and the attainment of an inflation target, and applies it to Ghana. It finds that the output inflation trade-off is more severe at earlier stages of credibility, and policy needs to be geared toward achieving its target, at the expense of output.

The second sort of model is a small semi-structural New-Keynesian model for applied monetary policy analysis and forecasting in central banks. The model is semi-structural in that the model is not fully derived from first principles (micro-foundations), even though each equation has a structural interpretation. The analyses presented in Part III focus on a workhorse model, emphasizing the interaction of the output gap, the exchange rate, and inflation (a Phillips curve), the implications of the stance of monetary policy and the level of the exchange rate for output (an IS curve), a monetary policy reaction function, and an exchange rate determination equation based on some form of purchasing power parity.52

This is an entirely standard model. Many features of SSA economies, such as a strong exchange rate effect on prices, or a weak transmission mechanism due to lack of financial depth and a fragmented banking system, can be handled through the calibration of the model parameters. Other features require more significant modifications. In Part III of the book we examine explicit treatment of food and fuel prices (Chapter 15), additional arguments in the reaction function (money targets in Chapter 16, exchange rate objectives in Chapter 18), and limited capital mobility. In many cases, the richer models analysed above can inform the use of these simpler models in more operational setting, such as the analysis of capital account and banking system risk shocks (Chapter 17).

Central banks in the region are gradually making use of these types of models. This is part of a broader effort to develop in-house forecasting and policy analysis systems (FPAS), drawing on best practices that have emerged from the experience of central banks in other regions.53 For the most part, the focus has been on simple models that focus exclusively on monetary policy.

8 Conclusion

Central banks in SSA have come a long way. They played a critical role, though perhaps subordinate to fiscal policy, in stabilizing macroeconomies and thereby helping set the stage for the growth resurgence that much of the continent has enjoyed since roughly the mid-1990s. The challenges, however, seem to be getting tougher. Perhaps foremost is the difficult global economic environment: will SSA countries be able to keep growth going in the face of shocks related to uncertain growth prospects in China and the developed world, swings in commodity prices, volatile global capital flows, and other aftershocks from the global financial crisis? Are monetary policy institutions strong enough? Have central banks achieved effective enough monetary policy frameworks to adjust to these shocks, keep expectations anchored, and resist political pressures? Much progress has been made and much more is underway. Will pressures expose weaknesses that spur further reforms or rather derail them?

Many of these challenges lie in the domain of fiscal policy and more broadly still in the resilience of a broad range of institutions both public and private. Most of the shocks are ‘real’, not nominal: real commodity prices, resource output, FDI flows, foreign demand, and fiscal policy. However, in our view the agenda for monetary policy that we have outlined here can play a critical supporting role.

Central banks can work to implement clear forward-looking policy regimes that respond coherently to the full range of shocks. This will help avoid macroeconomic and financial crises, allow the exchange rate flexibility to avoid persistent misalignments due to commodity price shocks, and keep inflation expectations anchored while avoiding unnecessary swings in interest rates, inflation, exchange rates, and output. All this can keep bad times from exploding into vicious circles of macroeconomic disarray and allow policymakers time to address the full range of challenges. The lessons of the global financial crisis for monetary policy regimes themselves are still being digested. But in our view all of this broader reform is better built on the solid foundations we have discussed here.


    Adam, C. (2011). On the Macroeconomic Management of Food Price Shocks in Low-Income Countries. Journal of African Economies, 20, 63–9.

    • Search Google Scholar
    • Export Citation

    Adam, C. andO’Connell, S. (2005). Monetary Policy and Aid Management in Sub Saharan Africa, mimeo. Oxford: University of Oxford; Swarthmore, PA: Swarthmore College.

    • Search Google Scholar
    • Export Citation

    Adam, C.,Berg, A.,PortilloR., andUnsal, F. (forthcoming). Monetary Policy and Central Banking in Sub-Saharan Africa. In P.Conti-Brown andR.Lastra (Eds.), Research Handbook on Central Banking. New York: Edward Elgar Publishing.

    • Search Google Scholar
    • Export Citation

    Adam, C.,O’Connell, S.,Buffie, E., andPattillo, C. (2009). Monetary Policy Rules for Managing Aid Surges in Africa. Review of Development Economics, 13 (3), 464–90.

    • Search Google Scholar
    • Export Citation

    Ahmad, A. andPentecost, E. (2010). Terms of Trade Shocks and Economic Performance Under Different Exchange Rate Regimes. Loughborough University Working Paper 2010–08.

    • Search Google Scholar
    • Export Citation

    Ball, L. (2011). The Performance of Alternative Monetary Regimes. In B.Friedman andM.Woodford (Eds.), Handbook of Monetary Economics, 3B, New York: North Holland Elsevier.

    • Search Google Scholar
    • Export Citation

    Batini, N. andLaxton, D. (2007). Under What Conditions Can Inflation Targeting be Adopted? The Experience of Emerging Markets. Central Banking, Analysis, and Economic Policies Book Series, 11, 467–506.

    • Search Google Scholar
    • Export Citation

    Berg, A.,Karam, P. D., andLaxton, D. (2006). A Practical Model-Based Approach to Monetary Policy Analysis-Overview. IMF Working Paper 06/80. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Berg, A.,J.Vlcek,L.Charry, andPortillo, R. (2013). The Monetary Transmission Mechan ism in the Tropics: A Narrative Approach. IMF Working Paper 13/197. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Berg, A.,O’Connell, S.,Pattillo, C.,Portillo, R., andUnsal, F. (2015). Monetary Policy Issues in Sub-Saharan Africa. In C.Monga andJ. Y.Lin (Eds.), The Oxford Handbook of Africa and Economics: Policies and Practices. Oxford University Press.

    • Search Google Scholar
    • Export Citation

    Berg, A.,Aiyar, S.,Hussain, M.,Roache, S.,Mirzoev, T., andMahone, A. (2007). The Macroeconomics of Scaling Up Aid: Lessons from Recent Experience. IMF Occasional Paper 253. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Bernanke, B.,Laubach, T.,Mishkin, F., andPosen, A. (1998). Inflation Targeting: Lessons from the International Experience. Princeton: Princeton University Press.

    • Search Google Scholar
    • Export Citation

    Bils, M. andKlenow, P. (2004) Some Evidence on the Importance of Sticky Prices. Journal of Political Economy, 112 (5), 947–84.

    Blanchard, O. andGali, J. (2007). Real Wage Rigidities and the New Keynesian Model. Journal of Money, Credit, and Banking, 39 (1), 35–66.

    • Search Google Scholar
    • Export Citation

    Blanchard, O.,Cerutti, E., andSummers, L. (2015). Inflation and Activity—Two Explorations and their Monetary Policy Implications. IMF Working Paper 15/230. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Broda, C. (2004). Terms of Trade and Exchange Rate Regimes in Developing Countries. Journal of International Economics, 63, 31–58.

    Buffie, E.,Adam, C.,O’Connell, S., andPattillo, C. (2008). Riding the Wave: Monetary Responses to Aid Surges in Low-Income Countries. European Economic Review, 52 (8), 1378–95.

    • Search Google Scholar
    • Export Citation

    Buffie, E.,O.Connell, S., andAdam, C. (2010). Fiscal Inertia, Donor Credibility, and the Monetary Management of Aid Surges. Journal of Development Economics, 93 (2), 287–98.

    • Search Google Scholar
    • Export Citation

    Christiano, L.,Eichenbaum, M., andEvans, M. (1999). Monetary Policy Shocks: What Have We Learned and To What End? In M.Woodford andJ.Taylor (Eds.) Handbook of Macroeconomics, 1A. North HollandElsevier.

    • Search Google Scholar
    • Export Citation

    Cukierman, A. (1992). Central Bank Strategy, Credibility, and Independence—Theory and Evidence. Cambridge, MA: MIT Press.

    Cukierman, A.,WebbS., andNeyapti, B. (1992). Measuring the Independence of Central Banks and its Effect on Policy Outcomes. The World Bank Economic Review, 6, 353–98.

    • Search Google Scholar
    • Export Citation

    de Carvalho FilhoI. E. (2011). 28 Months Later: How Inflation Targeters Outperformed Their Peers in the Great Recession. The B.E. Journal of Macroeconomics, 11(1), 1–46. De Gruyter.

    • Search Google Scholar
    • Export Citation

    Dincer, N. andEichengreen, B. (2014). Central Bank Transparency and Independence: Updates and New Measures. International Journal of Central Banking, 38 (3), 189–253.

    • Search Google Scholar
    • Export Citation

    Edwards, S., andLevi-Yeyati, E. (2005). Flexible Exchange Rates as Shock Absorbers. European Economic Review, 48 (9), 2079–105.

    Federico, P.,Vegh, C., andVuletin, G. (2014). Reserve Requirement Policy over the Business Cycle. NBER Working Paper 20612.

    Gali, J. (2015). Monetary Policy, Inflation and the Business Cycle, 2nd edn. Princeton University Press.

    Giuliano, P.,Mishra, P., andSpilimbergo, A. (2010). Democracy and reforms: Evidence from a new dataset. American Economic Journal, 5 (4), 179–204.

    • Search Google Scholar
    • Export Citation

    Goncalves, C. andCarvalho, A. (2009). Inflation Targeting Matters: Evidence from OECD Economies’ Sacrifice Ratios. Journal of Money, Credit, and Banking, 41 (1), 233–43.

    • Search Google Scholar
    • Export Citation

    Goncalves, C. andSalles, J. (2008). Inflation Targeting in Emerging Economies: What Do the Data Say?Journal of Development Economics, 85(1–2), 312–18.

    • Search Google Scholar
    • Export Citation

    Hernandez-Cata, E.,François, C.,Masson, P.,Bouvier.,Peroz, P.,Desruelle, D., andVamvakidis, A. (1998). The West African Economic and Monetary Union. IMF Occasional Paper 170. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Hoffmaister, A.,Roldos, J., andWickham, P. (1998). Macroeconomic Fluctuations in Sub-Saharan Africa. IMF Staff Papers, 45 (1), 132–60.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (1977). The Monetary Approach to the Balance of Payments.

    International Monetary Fund. (2008). Monetary and Exchange Rate Policies in Sub-Saharan Africa. Regional Economic Outlook, Sub-Saharan Africa. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2013a). The Interaction of Monetary and Macro-Prudential Policies. IMF Policy Paper. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2013b). Issuing International Sovereign Bonds: Opportunities and Challenges for Sub-Saharan Africa. Regional Economic Outlook, Sub-Saharan Africa. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2014). Staff Guidance Note on Macro-Prudential Policy—Considerations for Low Income Countries. IMF Policy Papers. Washington, DC: Inter national Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2015a). Evolving Monetary Policy Frameworks in Low-Income and other Developing Countries. IMF Policy Papers. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2015b). Monetary Policy and Financial Stability. IMF Policy Papers. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2015c). West African Economic and Monetary Union: Common Policies of Member Countries. IMF Country Report 15/100. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund. (2016). Exchange Rate Regimes in Sub-Saharan Africa: Experiences and Lessons. Regional Economic Outlook, Sub-Saharan Africa. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Jonas, J. andMishkin, F. (2003). Inflation Targeting in Transition Economies: Experience and Prospects. In B.Bernanke andM.Woodford (Eds.), The Inflation Targeting Debate. Chicago: NBER The University of Chicago Press.

    • Search Google Scholar
    • Export Citation

    Kessy, P.,O’Connell, S., andNyella, J. (2016). Monetary Policy in Tanzania: Accomplishments and the Road Ahead. In C.Adam,P.Collier, andB.Ndulu (Eds.), Tanzania: The Path to Prosperity. Oxford: Oxford University Press.

    • Search Google Scholar
    • Export Citation

    Lane, P. andMilesi-Ferretti, G. (2007). The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004. Journal of International Economics, 73 (2), 223–50.

    • Search Google Scholar
    • Export Citation

    Laxton, D.,Rose, D., andScott, A. (2009). Developing a Structured Forecasting and Policy Analysis System to Support Inflation-Forecast Targeting (IFT). IMF Working Paper 10/65. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Leeper, E. M., andD. B.Gordon (1992). In Search of the Liquidity Effect. Journal of Monetary Economics, 29 (3), 341–69.

    Lucotte, Y. (2009). The Influence of Central Bank Independence on Budget Deficits in Developing Countries: New Evidence from Panel Data Analysis. Laboratory of Economics in Orleans, University of Orleans.

    • Search Google Scholar
    • Export Citation

    Mankiw, N. G., andReis, R. (2002) Sticky information versus sticky prices: A proposal to replace the new Keynesian Phillips curve. Quarterly Journal of Economics, 117 (4), 1295–1328.

    • Search Google Scholar
    • Export Citation

    Masson, P. andPattillo, C. (2005). The Monetary Geography of Africa. Washington, DC: The Brookings Institution.

    Mishra, P. andMontiel, P. (2013). How Effective is Monetary Transmission in Developing Countries: A Survey of the Empirical Evidence. Economic Systems, 37(2), 187–216. Elsevier.

    • Search Google Scholar
    • Export Citation

    Mishra, P.,Montiel, P., andSpilimbergo, A. (2012). Monetary Transmission in Low-Income Countries: Effectiveness and Policy Implications. IMF Economic Review, 60, 270–302.

    • Search Google Scholar
    • Export Citation

    Ostry, J. D.,Ghosh, A. R., andChamon, M. (2012). Two Targets, Two Instruments: Monetary and Exchange Rate Policies in Emerging Market Economies. IMF Staff Discussion Note SDN/12/01. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Schmitt-Grohe, S. andUribe, M. (2010). Liquidity Traps: An Interest-Rate-Based Exit Strategy. NBER Working Paper 16514.

    Sims, C. A. (1992). Interpreting the Macroeconomic Time Series Facts: The Effects of Monetary Policy. European Economic Review36 (5), 975–1000.

    • Search Google Scholar
    • Export Citation

    Stone, R. andBhundia, A. (2004). A New Taxonomy of Monetary Regimes. IMF Working Paper 04/191. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    Summers, L. H. (1991). The Scientific Illusion in Empirical Macroeconomics. The Scandinavian Journal of Economics, 129–48.


See IMF (2015a). We also draw extensively on Berg et al. (2015) and Adam et al. (forthcoming).


We do not cover the important topic of the role of central banks in promoting financial stability. See Adam et al. (forthcoming).


This section draws heavily on Berg et al. (2015) and Adam et al. (forthcoming), which contain more comprehensive references.


Developing countries include countries classified as emerging markets and developing countries according to the IMF world economy outlook (WEO).


See, for example, IMF (1977).


For countries in SSA with available data (and excluding pegs), the financial reform index reported by Giuliano et al. (2010) more than doubled on average in the decade between 1985–90 and 1995–2000 (the countries are Ghana, Kenya, Madagascar, Mozambique, Nigeria, South Africa, Tanzania, and Uganda).


See for example Kessy et al. (2016).


Inflation targeting clearly embodies these principles, as it was the historical development of this regime that helped clarify these desirable properties of monetary policy. In policy debates in Africa, however, as elsewhere, the term ‘inflation targeting’ has at times been a source of controversy. Some have interpreted it as implying a strict and exclusive concern with inflation, deeming it inappropriate for low-income countries. Others feel that central banks can only adopt ‘inflation targeting’ after a long sequence of reforms. The focus on principles is an attempt to move the debate forward.


See De Carvalho Filho (2011).


See IMF (2013) and IMF (2015b), among many others.


On the topic of reserve requirements, see the discussion in Federico et al. (2014).


On the argument that higher interest rates can help raise inflation expectations when the economy is at the zero lower bound, see Schmitt-Grohe and Uribe (2010). On the slope of the Phillips curve, see Blanchard, Cerutti and Summers (2015) and references therein.


This point is argued forcefully in Bernanke et al. (1998).


See Broda (2004) and Edwards and Levi-Yeyati (2005) for evidence on the effect of terms of trade shocks in developing countries across exchange rate regimes, and Hoffmaister et al. (1998) and Ahmad and Pentecost (2010) for similar analyses for sub-Saharan African countries.


Measures of de facto financial openness, for example based on the sum of international assets and liabilities in per cent of GDP, also show SSA countries lagging. See Lane and Milesi Ferretti (2007).


Chapter 2 provides a discussion of these issues in the case of Uganda.


See Chapter 12, Adam et al. (2009), and Buffie et al. (2008, 2010) on the pros on cons of various policy responses in this context.


Central Bank Legislation Database, International Monetary Fund, 2012.


Indices of central bank independence combine assessments of tenure protection of central bank’s senior management, operational independence, clearly legally defined objectives for monetary policy, and limits to central bank lending to the government. The construction of the indices is based on the methodology outlined in Cukierman (1992) for de facto independence and Cukierman et al. (1992) for de jure independence.


IMF (2016) notes some movement back from de facto floats to de facto intermediate regimes, particularly in the face of supply shocks in commodity-dependent countries. It also observes relatively problematic macro performance in these countries, on the whole.


To the extent there is an inflation objective, it is more akin to a short-term inflation forecast, which is revised to account for short-term pressures and does not guide policy in a meaningful way.


Targets on reserve money are part of a broader monetary programming exercise in which targets are also set for broad money, which is considered an intermediate target of policy. With a few exceptions, however (e.g., Tanzania), targets on broad money play a smaller role in policy discussions in practice.


Because of this de facto flexibility, we see little impact of the introduction of electronic payments systems such M-Pesa for monetary policy implementation. The much more important implications for financial inclusion and regulation are outside the scope of this book.


Berg et al. (2013) (the working paper version of Chapter 5) describe the implications of strict money targeting for interest rate volatility in Uganda.


See Stone and Bhundia (2004). Chapter 8 argues that there is little case that monetary aggregates play a special direct role in the transmission of monetary policy. It explores a role for money aggregates in the face of weak real data and uninformative financial markets. Our subsequent thinking, partly captured in Chapter 9 and Chapter 16, as well as in this chapter, is less sanguine about this interpretation and puts more weight on the negative effects of money aggregate targeting on the information content of financial markets. Chapter 14 discusses the influence of our experience working with central banks in this evolution in our thinking.


These issues are treated in Chapters 8 and 9.


Zambia had also experienced unintended changes in its policy stance during 2005–06, again related to the monetary framework, as increases in money demand that were not accommodated resulted in higher interest rates and exchange rate overshooting.


See IMF (2015a). Countries that are investing in their analytical capacity include Ghana, Kenya, Mauritius, Mozambique, Rwanda, and Uganda.


See Ball (2011) for a critical review of the literature on this topic. See also Goncalves and Salles (2008) and Goncalves and Carvalho (2009).


The case of the Czech Republic is informative in this regard. See Jonas and Mishkin (2003).


A third and obvious issue are the various pressures stemming from fiscal policy. We do not discuss these issues further here to keep the discussion brief, though some of the implications for monetary policy are discussed in Chapter 12.


When the relative price of food is trending, targeting core is particularly problematic as it would imply that the overall CPI could drift away from the target. Chapter 15 provides a brief discussion of these issues.


It is important to note that neither the CMA nor the CFA are entirely conventional hard peg arrangements, since in both cases the hegemon (the Reserve Bank of South Africa in the former instance and the French Treasury—not the ECB—in the latter) provides support to the small member states.


See IMF, AFR Regional Economic Outlooks, for example IMF (2008).


A common currency for Africa was a stated objective of the Organization of African Unity when it was founded in 1963 and was endorsed at the founding of the successor body, the African Union, in 2002 (see Masson and Pattillo, 2005).


See Broda (2004) and other references in n. 18.


Chapter 11 discusses the evidence on the (relative lack of) nominal rigidities in the food sector and implications for monetary policy.


These efforts have benefitted from support by the IMF and the UK Department for International Development (DFID). More information can be found at:

    Other Resources Citing This Publication