Back

Fiscal Policy Harmonization in the Context of African Regional Integration

Preliminary Draft
This is work in progress. We appreciate your comment and could be addressed to:

geda.uneca@un.org orAlemayehu@excite.com

(Paper Prepared for Regional Economic Cooperation and Integration Division of the Economic Commission for Africa)

October 2001
Addis Ababa

Alemayehu Geda, PhD
The Kenya Institute for Public Policy Research & Analysis, and
Department of Economics, Addis Ababa University
Alemayehu@excite.com

Table of Contents Page

I. Introduction 33

II. Why Fiscal Policy Harmonization: The Rationale and Analytical Framework 44

2.1 The Rationale for Fiscal Policy Harmonization 44

2.1.1 Macro Spillover/Externalities 55

2.1.2 Tax Competition and Integration: Lesson from Fiscal Federalism 66

2.2 The Analytical Framework 77

2.2.1 Fiscal Policy, External Sector and Integration: The Accounting Framework 88

2.2.2 Fiscal Policy, External Sector and Integration: Some Analytical Issues 99

III. The African Context 1313

3.1 The Genesis of the African Fiscal (Public Deficit) Problem: The historical Context 1313

3.2 The Macro Convergence Criteria and the Recent Performance of African RECs 1818

3.2.1 Recent Macroeconomic Performances of African RECs (1995-2000) 1818

3.2.2 Recent Fiscal Performance of African RECs 2121

3.2.3 The Convergence Criteria of African RECs 2525

3.3 Some Major Outstanding Issues 2626

3.3.1 Revenue Loss 2626

3.3.2 Issues of Sovereignty and Institutional Aspect of Macro Policy Coordination 2828

3.3.3 Asymmetry of Shocks and Policy Harmonization 2929

IV. Conclusion 3030

Reference: 3232

______________________________________________________________________

Fiscal Policy Harmonization in the Context of African Regional Integration

I. Introduction

Understanding the macroeconomic context for regional integration is important. This is because, as argued by O'Connell (1997), trade related policies aimed at facilitating regional integration efforts could be undercut not only by import substitution strategies but also by the subordination of trade policy to fiscal and external balance (macroeconomic) considerations (see O'Connell 1997: 90). This underscores the need to harmonize macroeconomic policies so that it accords with trade policies aimed at strengthening regional integration.

Unlike trade liberalization issues, the importance of macroeconomic policy coordination on economic integration has received relatively little attention. But, as O'Connell (O'Connel 1997: 90) noted, `Among the most often cited constraints to greater intra-African trade is the inhospitable macro-economic environment associated with overvalued exchange rates and non-convertible currencies.' Elbadawi (1997) shows a supporting empirical evidence for this although another estimation (see Alemayehu and Haile 2001) using another proxy couldn't find similar results to that of Elbadawi (1997). As noted by Vos (2001), in general, the literature on macro policy coordination is geared around the choice of exchanger rate regimes mainly because this, next to trade barriers, directly links a country to the rest of the world and also it sets the degrees of freedom for national macroeconomic policies<sup>1</sup> (Vos, 2001). Clearly, in the context of regional integration, the issue of currency inconvertibility is still a major obstacle while the issue of overvalued currency is of less concern these days due to the widespread exchange rate liberalization polices carried out in many African countries. One should also add, the related obstacle that comes form currency instability, as recently witnessed in the Southern African region (Malawi, South Africa, Zambia and Zimbabwe, for instance) (Alemayehu and Haile 2001). Similar problem is observed in West Africa. Aryeeteye (2000), for instance, noted that `with emphasis on tariff reductions, [it] is unlikely to increase trade significantly if exchange rates are not properly aligned and the underlying macroeconomic framework is unstable' (see also Ndung'u 2000 for a similar argument in East Africa). Thus, in addition to harmonizing trade policies, coordination of macroeconomic policies, covering fiscal, monetary and operations of all financial institutions, is a necessary condition for a smooth implementation of economic integration.

Post independence Africa witnessed two major macroeconomic regimes: the pre and post Structural Adjustment periods, where the mid 1980s could be thought as the watershed. The first regime, notwithstanding the political importance of pan-Africanism, is characterized by the post-independence preoccupation of building a fairly independent national economy. In general the policy stance was informed by import substitution strategy. This strategy was not pro-open trade policy and to that degree has adversely affected regional integration efforts. Notwithstanding such general pattern, there are some countries such as Algeria, Ethiopia, Mozambique, Tanzania, etc which pursued a strictly controlled regime and associated macro policies.

The adoption of Structural Adjustment Programs (SAPs) starting in mid 1980s in almost all African countries has led to openly adopting a liberalization (open economy) policy. The identical nature of policy instruments prescribed by International Financial Institutions (IFIs) across countries in the continent implies a defacto macro policy harmonization, at least at the level of intent. The fact that these policies are pro-open trade suggests a possible positive relationship between the macro policy pursued and regional integration efforts (see below, however).

In sum, although the importance of regional economic groupings is crucial to survive in the increasingly integrating world economy, addressing major obstacles such as macro policy coordination is a daunting task. It is important that both African governments' and their development partners appreciate this challenge. In particular, the latter can play a significant role by focusing on regional support programs. Member countries need also to take integration as an important economic survival strategy aimed at combating marginalization from the global economy. The role of macroeconomic coordination in this respect is very crucial. Within this general macroeconomic policy coordination context comes fiscal policy harmonization. It can be argued that fiscal policy in Africa is very much linked to monetary and hence macro polices. However, to have a deeper understanding of the fiscal issues, we will focus on fiscal policy harmonization. The analysis will, however, draw from issue of macro policy harmonization in general and monetary policy harmonization in particular when such linkage is crucial for understanding fiscal policy issues.

The rest of the paper is organized as follows. In section two we will highlight the rationale behind macro policy harmonization in general and fiscal policy harmonization in particular in the context of regional integration effort. The section also outlines the analytical framework for understanding macroeconomic harmonization issues. In section three we will focus on issues of fiscal policy harmonization in African context. This is done by examining both the recent trend of major macroeconomic variables and the fiscal posture of African RECs. The section also highlights major fiscal harmonization issues that need to be further investigated. Section four concludes the paper.

II. Why Fiscal Policy Harmonization: The Rationale and Analytical Framework

2.1 The Rationale for Fiscal Policy Harmonization

The need for macroeconomic coordination is usually discussed in the context of industrialized countries. In the 1980s and 1990s major macro data of the industrial countries show that there were divergent pattern of current account position, development in real interest rate and exchange rates. These were associated with large and divergent changes in world fiscal policies (Frenkel and Razin, 1996). In fact the emphasis on monetary policy, which was in vague in the 1970s, has changed to fiscal issues in the 1980s. This is partly motivated by the EU experience, the US government deficit & its implication for trade deficit as well as its repercussion on the rest of the world. With EU, monetary union issues such as exchange rate target zones and the harmonization of tax regimes increasingly came to be important (see Turnovesky 1997). Similar efforts are also featured in major integration schemes such as the Mercosur (Vos 2001) although fiscal polices issues had not been widely studied in the context of African RECs (O'Connel 1997, Masson and Pattillo 2001). The importance of the macro context (including macro spillovers/externalities) and issues such as tax competition underscore the importance of examining fiscal policy harmonization at depth.

2.1.1 Macro Spillover/Externalities

The impetus for fiscal policy harmonization comes from the link between fiscal deficit, exchange rate and the external sector. The latter two in particular are the channels through which a country is linked with the rest of the world. Facilitating such link is crucial for successful regional integration. Since major macroeconomic balances in general and the fiscal posture in particular have a direct bearing on the external sector, its harmonization is imperative for regional integration efforts. As noted by Goldstein (1994), national policy action can have quantitatively significant spillover effects, or externalities. These need to be taken on decision-making process to reach global optimum. This can best be achieved by use of macro policy harmonization. Put differently, the justification for fiscal policy harmonization could be found from the fact that some developmental objectives can best be handled by centralized (or coordinated) organs than individual countries. This is because they might have externalities or there could be the possibilities of exploiting scaled economies. As the experience in EU shows fiscal policy harmonization at the level of RECs could focus on correcting distortions or exploiting externalities that cannot be corrected or exploited by national fiscal policies. Externalities such as tax competitions, or issues where REC level social returns exceed that at national level are cases in point. The latter two factors were the main impetus behind fiscal policy harmonization in EU (see Masson 2000).

Another impetus for fiscal policy harmonization comes form its link with monetary policy. Masson and Pattillo (2001) after reviewing the experience of monetary union in West Africa noted that instead of trying to meet a very short deadline for monetary union, he countries of the region need to focus on convergence on low inflation, sustainable fiscal polices and structural policies necessary for strong growth. They noted that, if the role of France in the monetary union of West Africa is not taken on board, the need for fiscal policy harmonization will be much more critical. For instance, pursuing open trade policy requires harmonization of exchange rate. The latter is linked, among other things, to inflation differential. One of the instrument to harmonize inflation differential is fiscal policy harmonization.

Yet another rationale for fiscal policy harmonization comes from the `optimum currency area' literature that is usually discussed in the context of a monetary union, which is one major objective of regional integration schemes. In this context fiscal policy harmonization is imperative because it is one of the important preconditions (apart form wage flexibility and labour mobility) for optimum currency area formulation (see De Grauwe 1994).

2.1.2 Tax Competition and Integration: Lesson from Fiscal Federalism

Fiscal federalism is created with the intention of combining the different advantages, which result from the magnitude and littleness of nation (Oates 1999). To realize these advantages we need to understand which basic functions of fiscal policy could best be handled by the central or local authorities. This is the subject matter of fiscal federalism (Oates 1999). To the extent that this principle can be applied at regional or continental level, some basic principles of fiscal federalism can be used as the source of ideas for drawing fiscal harmonization policies in regional integration schemes. One important principle in fiscal federalism is that the central government should have the basic responsibility for macroeconomic stabilization and income redistribution. In both cases the argument stems from fundamental constraints on lower level governments - local government simply have very limited means (Oates 1999). This principle can be used to assign fiscal policy assignments between RECs and individual countries. This has been for instance used in EU countries through the principle of `subsidiarity' (see below). The other two important theoretical aspects of fiscal federalism: the welfare gain from fiscal federalism and the use of fiscal instruments could also be used to draw broad principles for fiscal policy harmonization (see Oates 1999, Masson 2000, Cangiano and Mottu 1999, Tanzi and Zee 1999).

The EU experience shows that fiscal policy is coordinated through the multilateral surveillance and excesses deficit procedure of the Maastricht Treaty, as show in the Stability and Growth Pact. Canginao and Mottu (1998), having noticed that, argued that such EMU policy framework is closer to that of a federal than a pure monetary union. Since there is no central fiscal authority, however, it relies on coordination of fiscal polices. The important question is whether there is any thing that RECs can draw from the theory and experience of fiscal federalism.

The standard fiscal federalism theory concludes that two of the three basic functions of fiscal policy (ie. Redistribution and stabilization) should be conducted at the central level. The third function (allocation) can be assigned to different levels of government. In practice, however, in exiting federations all three are largely carried by central governments (Canginao and Mottu 1998, Oates 1999). The question is what is the basis for allocating functions between central and local authorities in fiscal federalism context and can we use that across countries in regional integration context. The EU experience shows that the answer to this is in affirmative. This is handled by the so called `Subsidiarity' principle² of the Maastricht Treaty. Thus, taking the three basic functions of fiscal policy, allocative efficiency in EU is largely pursed at the EU level (through establishment of single market, removal of fiscal frontiers, recognition and harmonization of standards, norms and procedures, harmonization of indirect taxes etc) while redistribution and stabilization functions are largely left to individual countries (Canginao and Mottu, 1998). African RECs can design their macro policy coordination efforts in general and fiscal policy harmonization in particular by drawing from this European experience, which basically adopted one of the principles of fiscal federalism.

Another dimension of the principle of fiscal federalism and subsidiarity is the question of tax harmonization. With increasing integration, and in particular with common currency or harmonized exchange rate regimes, tax competition is likely to increase at least for two reasons (Canginao and Mottu (1998): (a) tax inclusive prices would become more transparent and (b) with the loss of monetary and exchange rate instruments, the role of tax policy in attracting business and enhancing competitiveness would become prominent. The latter in particular was the trend in many of African countries following trade liberalization policies. This trend could entail sever fiscal imbalance. Moreover, divergence in tax system in the context of a regional integration scheme could have differentiated growth implication across member states (see Frenkel and Razin, 1996). Thus, there should be an attempt to refrain from harmful tax practices so as to avoid a `rush to the bottom' of the tax system which would prevent government from sustaining desirable tax polices and financing necessary expenditures (Canginao and Mottu, 1998: 20, Oates 1999, Masson 2000). In a fiscal federalism context Rivlin (1992, quoted by Oates 1999) noted that it is almost axiomatic that competition among states result in inadequate level of public service. Her remedy is a system of shared taxes under which the revenue from a new centralized valued-added tax would be shared among the states, avoiding unnecessary competition (see Oates, 1999: 1135). Thus, the issue of tax competition is one of the important reasons for fiscal policy harmonization.

Understanding about harmonization of fiscal policy requires some degree of evaluation of fiscal performance in the continent. Such evaluation needs to be conducted across countries and over time in a consistent macroeconomic framework. Given the interlink among fiscal, monetary and exchange rate policies in the increasingly globalzing economy, fiscal performance can not be analyzed in isolation - i.e. it should be an integral part of the macro policy framework. Moreover, macro policy (fiscal policy included) needs to be measured not only against policy objectives but also by taking long-term and short-term considerations (Dinh, 2001). This suggests the importance of having analytical framework to guide our discussion of fiscal policy harmonization issues.

2.2 The Analytical Framework

It has been argued before that fiscal policy harmonization in Africa is strictly linked to macroeconomic policy harmonization. Since our objective here is to analytically link macro (including fiscal) policies to issue of regional economic integration, the best approach is to see the framework through which basic macro variables are linked to export and imports of goods and services. This is done in at two stages. In the first stage we attempted to examine the link between fiscal policy and regional integration using the national accounting framework. In the second stage we moved beyond the national income accounting relationship to the economics of the fiscal policy and trade (integration) linkage using the standard internal and external balance approach and along the framework outlined in O'Connel (1997). Since O'Connel's approach is fairly broad, we will be restricting our analysis to the fiscal aspect. In fact we will develop and extend O'Connel's insight by drawing fundamentals of fiscal policy issues from the experience of European Monetary Union and some aspect of (US) Fiscal Federalism.

2.2.1 Fiscal Policy, External Sector and Integration: The Accounting Framework

The cornerstone of an open economy macroeconomics accounting framework is the identity that links the internal balance with the external balance. Thus, in relation to the United Nations System of National Accounts this is the link between national accounts and the balance of payment. However, one problem, which one might encounter in using such an accounting framework, is the lack of institutionally disaggregated detailed data. This could, to some degree, be resolved by resorting to various multinational data sources (see Alemayehu et al 1992, Alemayehu 2001 for details on this). A major macro problem in the case of Africa, and developing countries in general, is how to finance investment. This may be addressed by starting from the national income accounting identity (equation 1) and re-writing it to yield the accumulation balance (equations 2 and 3) as,

[1]

Where Y is income, C consumption, G government expenditure and X and M are imports and exports of goods and non-factor services, respectively. F is the external finance (see equations 3 for definition) could be given as,

[2]

Rearranging, we may arrive at an explicit relationship between investment and its financing given by equation 3,

[3]

Where: I is gross domestic investment, S national savings and F net capital inflows. The latter is defined as the net change in assets and liability position of the country, and is equal to the deficit of the current account of the balance of payments (i.e. the external balance), which is given as,

[4]

Where N is net factor payment and current transfer to abroad. Combining these, disaggregated into public (g) and private (p) sectors and rearranging, yields

[5]

Where: T-G is the fiscal deficit.

This yields the basic identity which links the domestic investment and savings gap with the current account deficit or surplus, and hence the resulting capital inflow or outflow³. It can be read from equation 5 that the fiscal deficit (and also the fiscal posture as can be read form T and G⁴) are strictly related to the external sector as give by M, X and N. To the extent that regional integration efforts in Africa do largely show through the external sector in general and trade in particular, the fiscal stance has a direct bearing on regional integration. From equations 5, without venturing on to the issue of causality⁵, we see a clear positive association between the level of fiscal deficit and the current account balance. Thus, fiscal policy aimed at addressing fiscal deficit has a direct bearing on external trade and economic integration⁶. In the next section we will extend this accounting relationship to an analytical one.

2.2.2 Fiscal Policy, External Sector and Integration: Some Analytical Issues

The analytics of fiscal policy in the context of integration could be built on the accounting framework outlined above. The standard approach is to use the theoretical insight from the macroeconomics of the internal and external balance. This framework not only gives us major theoretical conclusion from systemic point of view but also could help to organize our thinking about fiscal policy harmonization.

Let us first consider the internal balance and how fiscal policy comes in the analysis. We will be using the same symbols as in section 2.2.1 above. A superscript `f' shows full employment level wile `*' shows foreign (as opposed to local/domestic) variables. `E' and CA stand for nominal exchange rate and current account balance, respectively. Assuming P* and E are fixed, inflation will depend on aggregate demand pressure which is strictly linked to the fiscal variables. Internal balance requires that full employment holds (i.e. Aggregate demand equals aggregate supply at Y^f ). This is give by,

[6]

From this internal balance requirement (equation 6) we note that fiscal policy expansion (an increase in G and a decrease in T) could stimulate aggregate demand. Devaluation (a rise in E) leads to an increase in domestic output. Current account surplus (CA) is a decreasing function of disposable income (Y^f-T) and an increasing function of the real exchange grate, EP^*/P⁷. The internal balance (IB) schedule in Diagram 1 below shows the combination of exchange rate (E) and fiscal policy that holds output constant at Y^f - the internal balance.

Once we have the internal balance, the next question is to examine how the two policy instruments (exchange rate and fiscal polices) could be used as policy instruments that could affect the external balance. This is crucial in the context of this study. This is because this channel offers the possibility of exploring the analytical link between fiscal policy and economic integration as can be read from the external balance.

Coming to the external balance, if we assume that the government has a target level of current account balance (CA^target), achieving this target requires that,

[7]

We note here that, given P and P*, devaluation (_ E) makes domestic goods cheaper and improves the CA. Fiscal expansion, however, has the opposite effect on CA. A fall in T or a rise in G (fiscal expansion - a rise in absorption) increases the demand for foreign goods and could result in worsening of the CA. Thus, to maintain the CA at CA^target as the country devalues, the government must increase G or reduce T - hence the EB schedule shows the combination of exchange rate and fiscal policy stance at which external balance holds at stipulated level.

Diagram 1: The Salter-Swan Internal and External Balances Diagram

With many countries in African RECs moving to floating rates, we could relax the fixed exchange rate assumption that we have used above. Taking macroeconomic interdependence on board, CA balance in each country will depend on RER, and domestic and partner counties' disposable income. This could be given as,

[8]

From equation [8] we note the additional point that a rise in trading partner's disposable income leads to an increase in the current account balance of the domestic economy. The internal balance in home country is also affected by this interdependence and would take the form shown in equation [9].

[9]

The CA of the partner country is basically the mirror image (in reverse) of the home country. Thus, its CA balance, measured in its local currency, could be given as,

[10]

Similarly its internal balance can be given by equation [11]

[11]

From equation [11], by collecting Y and Y* in the left and right hand side of the equation, respectively; and for a given level of exchange rate and fiscal stance, we could derive the intersection of the schedules at which aggregate demand equals supply in both countries. It is noted that these two countries are very much interdependent. More interestingly, a change in the level of output of one of the countries has a multiplier effect on the other - the so-called export multiplier effect.

A number of analytical conclusions, related to macro policy issues across RECs, could be made from the above analysis. First, we note that the framework shows the interdependence of the two countries (or REC members). It can be shown that a change in major macro variables (such as fiscal expansion) has a negative effect on current account balance of one of the countries (home country) while positively affecting the other. If these countries are operating in the context of a particular REC, there is a need to harmonize their respective polices so as to arrive at optimal policy mix that maximizes the welfare both countries. Second, change in fiscal policy in one of the countries does change not only the internal and external balance schedules (and hence affect the level of output in the two countries) but also has effect on the real exchange rate, which in turn affects the level of output in the two countries. In other words, the analytical framework (see Diagram 1) shows that both expenditure changing (fiscal policy) and expenditure switching (exchange rate policy) policies are required to attain internal and external balance. This may have two important implications: (a) in the context of RECs, fiscal policy may not be analyzed in isolation to exchange rate and/or monetary policies, and (b) if there is a possibility of offsetting the effect of one policy instrument (say fiscal) by another (say monetary or exchange rate), it accentuates the need for policy harmonization across member of RECs so that a win-win solution is arrived at. Finally, the framework helps to see that tax on international trade, upon which many African country depend, needs to be analyzed not only as a source of revenue but also as a fiscal instrument (through T) that does affect domestic absorption and internal and external balance of the trading partners. Thus, when we examine the macroeconomic environment and the fiscal structuree of African RECs in the subsequent section, we need to bear such analytical links in mind.

III. The African Context

3.1 The Genesis of the African Fiscal (Public Deficit) Problem: The historical Context

It can be argued that the pre-independence economic history of African countries is characterized by a formation of a primary commodity exporting and external finance constrained economy (see Alemayehu 2001). The impact of the subsequent (after political independence) events of the boom in commodity prices, the oil price shocks of 1973-74 and 1978-79 and the evolution of African debt from the early 1970s onward would be difficult to understand unless an explicit link is made between the historically formed structure and the pattern of trade and finance in the period 1970 to date. This had a direct bearing on the fiscal posture of post-colonial African economies. This section briefly summarizes this phenomenon. This evolution of African trade and finance in the post independence period could be categorized under three periods.

The first period refers to the late 1960s and early 1970s. This period is characterized by the first oil shock and the rise in commodity prices. A sharp bust followed the commodity price boom in 1974, and again after 1977 for coffee and cocoa (see Figures 1 and 2). The response in most African countries is a rise in government expenditure in particular in infrastructure sector. When the commodity price fall governments were not only unable to cut expenditure but also in need of maintaining on going projects. This has shown itself in a growing fiscal deficit that cannot be fully domestically financed. Thus, this has been accompanied by increased borrowing owing to improved credit worthiness when prices of export commodities rise and due to belief in cyclical nature of prices when commodity prices decline. This could be read from the pattern of trade and finance of many African countries examined in detail in Alemayehu (1997)⁸. The major point that emerges from examining this period is that following the rise in commodity price and access to loan there was a rise in public expenditure⁹.

Figure 1: Price Index of Some Major Agricultural Export Commodities of Africa (1965=100)

Source: Alemayehu (1997)

Figure 2 Price Indices of Some Major Mineral Export Commodities of Africa (1965=100)

Source: Alemayehu (1997)

The above analysis shows that the period between the late 1960s to mid 1970s was characterized by a rise in the price of commodities on which African countries had specialized for historical reasons. It was also a period in which imports of capital and intermediate goods (mainly to develop infrastructure) increased. Foreign borrowing complemented this effort. It is at this particular juncture that almost all countries were hit by the first oil price shock. This shock was reflected on the public deficit, which mirrors the balance of payment deficit as shown in section two above. This shock was tackled, partly, by resorting to external financing and partly by deficit financing. Another way of viewing the latter phenomenon is to consider the additional external finance (which eventually turned into debt) requirements of African countries as a policy response to the external shocks they were facing (See Balassa 1983, 1984; Ezenwe 1993) and the apparent weakness of their fiscal posture. The question is whether such policy responses were rational. Should the shock be seen as a temporary one? Both on the part of African governments and creditors these shocks were believed to be temporary. Given this belief (that is the expectation of an eventual rise in commodity prices) and given the then prevailing low real interest rate (which was even negative, see Khan and Knight, 1983:2), it seems rational that both lenders and borrowers responded in the same way. As it turned out, the frustration of these expectations (secular decline in commodity price and rise in real world interest rate) put an enormous fiscal burden on African countries¹⁰.

True, there were some domestic policy problems in managing public expenditure in this period. However, the nature of public expenditure did not constitute a reckless spending as is usually implicitly portrayed in most African related literature. For instance if we take the two extreme cases in spending composition during this period: Nigeria and Zambia; in Nigeria after the first oil boom nearly 80% of public expenditure was on physical and social infrastructure. Capital expenditure was twice that of current expenditure. Public expenditure on trade, industry and mining rose from 7.3% in 1970-74 to 26% in 1975-80, transport from 21.3% to 22.2% in the two periods while general administration dropped from 22% to 13.6% (Alemayehu 2001, 1997).¹¹ Contrary to the case of Nigeria current expenditure in Zambia was nearly 75% of total expenditure in 1970-74 and this is largely attributed to the Zambianaization policy, which is dictated by the inherited colonial structure. Nonetheless, from 1972 (strengthened in 1974) the government attempted to curb current expenditure. For instance consumer durable import was reduced from 28% in 1974 to 18% in 1978. Similarly subsidies, with attending political costs, had been reduced in the early 1970s. In general by the mid 1970s public and private consumption had been substantially reduced from the high level of the 1970. This pattern was similar in many African countries (see Alemayehu 1997 for detail).

In sum, following the rise in commodity prices and access to loans there was a rise in public expenditure. Given the inherited colonial structure that necessitated spending on social and physical infrastructure to address the problem of the then neglected sections of the population; given prevailing hope in technology transfer through import substitution; and given the uncertainly about commodity prices, the expenditure was not reckless. In fact, in most African countries the relative share of functional expenditure hardly changed following the commodity boom of 1973-74 in general and 1976-77 for cocoa and coffee exporters (see Annex I for the empirical evidence). The capital expenditure did change, however, owing, as noted above, to the import substitution strategy pursued. The policy problem that emanated from failing to predict commodity price collapse and mange demand was a secondary one. The major problem is the specialization in a commodity whose price is cyclical in the short-run and declining over time. This argument should not be taken as endorsing some white-elephant investments carried in some African countries, however. Perhaps the major domestic policy problem associated with the rising expenditure was the way in which the import substitution (IS) strategy was conducted. While the IS strategy was a sound one, it was carried out in the context of a disarticulated domestic production and consumption structure. The latter is in particular vivid in the neglect of: the industrial and agricultural linkages; future demands for recurrent cost of intermediate inputs; and development of the human capital required. Moreover, policy makers made the mistake of taking IS and Export-led strategy as competitive rather than complementary¹².

This pattern was compounded by another development in the global financial markets. The oil price hikes not only forced oil importers to become more dependent on borrowing, they also created what is called the OPEC surplus -pax Arabica? (Bacha and Alejandro 1982). This surplus was circulated through the international banking system. The Euromarket became an important source of financing for a number of African countries, which had never borrowed before (Krumm 1985, Mistry 1988). The situation was reinforced by a second oil price shock (Kruger 1987, Salazar-Carrillo 1988 in Taiwo, 1991:39; and Ezenwe 1993). In a way this supply of foreign finance has eased the pressure on the growing fiscal deficit of African economies. However, it started to show in the accumulation of debt. The new funds borrowed were spent on mining companies and major public projects. But, in general, these loans were characterized by harder terms. When the second oil price shock came in the late 1970s, with commodity prices continuously deteriorating as shown in Figures 1 and 2, most countries were unable to absorb the shock (Krumm, 1985: 1-9). Thus, by the end of the 1970s the total external debt grew almost ten fold¹³.

The second period refers to the late 1970's and early 1980s.The end of the 1970s had witnessed the second oil price shock. ¹⁴ Major commodity prices continue to decline, prompted, inter alia, by the recession in the industrial countries. The early 1980s was also characterized by a hike in real interest rate in industrial countries, chiefly due to lax fiscal and tight monetary policy of the US.¹⁵ (see Khan and Knight, 1983:2). The hike in real interest rater aggravated the interest rate cost of nonconcessional and private debts that became increasingly important during this period (see Alemayehu 1997 for detail). This development prompted many African governments to continue borrowing (and get credit) on the assumption of a cyclical turn around in commodity prices. These new loans were used to finance enlarged oil bills and avoid sharp politically/socially disruptive cut backs in public expenditure (Mistry, 1988:7). The experiences of most African countries, discussed in detail in Alemayehu (1997), during this period generally confirm this pattern.

The third period refers to the late 1980s to the 1990s. This period, as that of late 1970s was generally characterized by continually declining commodity prices and the deterioration of terms of trade. For the period 1985-90, when a large number of African countries undertook adjustment programs, the deterioration in the barter terms of trade of nine major export commodities resulted in a 40% decline in average export revenue (compared to 1977-79 average), despite a 75% increase in export volume (Husain, 1994:168). As a result, African countries became more vulnerable to fiscal pressure and further indebtedness. Moreover, the capitalization of amortization and interest payment through the Paris and London clubs rescheduling had not only began to put a strain on the fiscal balance but also started pushing the debt stock upward (van der Hoeven, 1993 and Alemayehu 1997). This pattern is obvious from the reports of many countries examined in detail in Alemayehu (1997).

Given this general pattern from the mid 1980s to the 1990s, African economies were extremely indebted by the 1990s. Moreover, apart from investment in infrastructure (like the transport sector) which needed external finance for its maintenance, almost all countries had become dependent on external finance for securing imported intermediate inputs and ensuring the smooth functioning of their economy (See Ndulu 1986, Ngwenya and Bugembe 1987, Fantu 1991, Rattos 1992, Mbelle and Sterner 1991). Thus, throughout the two decades analyzed the value of import was persistently increasing in almost all countries¹⁶. This recurrent import demand problem, referred as import compression in African macro literature (see Ndulu 1986, 1991, Ratso 1992, 1994 and Alemayehu 2001) was compounded by actual running down of the capital stock, including infrastructure.

Thus, by late 1980s and early 1990s such historically structured African economies were vulnerable to events such as the industrialized economies recession, following the global monetary shock of 1979-81, which depressed commodity prices. This is also a time where the world economy witnessed (i) the emergence of high, positive real interest rate throughout the 1980s which increased the debt service burden of indebted countries, (ii) protectionism in the world market for agricultural products and low technology manufacturing which hampered diversification attempts and, finally, (iii) the prevalence of repeated official and private rescheduling, often at punitive terms (see Mistry, 1991:10-11 for detail). This crisis widened the role of multilateral finance despite being available at unacceptable terms - policy conditionality. One prominent features of such conditionality is a conservative fiscal policy (cut back on deficit). In return donors promised aid and soft loans¹⁷. Notwithstanding its detrimental social impact, most countries managed to stabilize their economy by austerity measures prescribed by IFIs. The collary of such conservative fiscal policy is that by the 1990s African countries found themselves not only being extremely indebted but also structurally unable to pay back their debt. Thus, the unsustainable fiscal posture appeared managed thanks to indebtedness.

3.2 The Macro Convergence Criteria and the Recent Performance of African RECs

Although the root cause of the fiscal deficit problems noted above may be addressed by radically changing the production and trade structure, combined with debt relief, fiscal austerity measures (notwithstanding their attendant social cost) had also their positive impact in Africa. Owing to implementation of structural adjustment programs (SAPs) across the continent, the pre-conditions for fiscal policy harmonization seem to have been sawn. This is because not only all countries are operating in a similar macro policy context but also adopted similar fiscal policies (such as target level of deficit, inflation and government revenue) as part of the SAPs. This doesn't seem to be changed by the recent introduction of the "Poverty Reduction Strategy Paper (PRSP)' whose fiscal vehicle is the `Medium Term Expenditure Framework (MTEF)'¹⁸. Now the question is whether such macro policies could be harmonized across RECs.

Attaining macroeconomic harmonization, using some convergence criteria, is not an easy task. RECs not only need to be committed to macro economic coordination but also be ready to finance the cost of convergence. Vos (2001) noted that the EU has effectively promoted economic convergence through its large compensatory fund, including the agricultural common fund and funds in the support of backward regions. Such fiscal transfers systems, as noted by Vos, are not in place in Latin American countries. Attempting to establish such fund in African RECs is much more challenging. Before arriving on such requirements, it is important to examine the feasibility of such macro coordination by examining the initial conditions. This is done in the next two sections.

3.2.1 Recent Macroeconomic Performances of African RECs (1995-2000)

Successful fiscal policy harmonization requires a fairly stable macroeconomic environment. Thus it is imperative to examine recent macroeconomic condition of African RECs so as to gauge the macroeconomic environment and the challenge of fiscal policy harmonization. Based on the analytical framework outlined in section 2.2.1, we have picked some major macro variables and examined their recent evolution. The analytical framework given in section II will give us how these variables are linked to the issue of fiscal policy harmonization. Detailed data on the basis of which the analysis below is made is given in Annex IV.

A): Common Market for Easter and Southern Africa (COMESA)

Real GDP growth across members of COMESA is disappointing in some of the countries such as Burundi, Comoros, Djibouti, DR Congo; the growth performance of the other members is promising, however. In particular, the growth in Egypt, Mauritius, Uganda and Ethiopia are worth emphasizing. However, in all cases the sustainability of the growth episode observed is vulnerable to external shock, including natural disaster. Inflation is a serious problem in COMESA. There is a wide fluctuation across members. There are countries where inflation is the worst (above 15%) such as Angola, Burundi, Madagascar, Malawi and Zimbabwe and good performers where it is contained below 10% (e.g. Comoros, Djibouti, Ethiopia, Uganda, Egypt, Mauritius). Low saving and investment rate also characterize COMESA members, which are invariably below 20% of the GDP. The only exception being Mauritius and, to some degree, Seychelles, Swaziland and Zimbabwe. Moreover, COMESA members are characterized by low monetization of the economy as can be read from the M2 to GDP ratio. All members are also characterized not only by low export intensity (export as percent of GDP) but also there is a wide gap between exports and imports. Again this indicator varies enormously across members. In some of the countries the import to GDP ratio is nearly twice to that of the export to GDP ratio. Again the only exceptions are Mauritius and to some degree Rwanda, Seychelles, Namibia and Swaziland. In these countries not only the ratios are very closes but also the level of trade intensity is very high (above 50% of GDP). However, in majority of the countries this problem shows itself in current account deficit and hence indebtedness. Thus, the debt to GDP ratio in the majority of the member is above 100 percent. In a couple of countries it is about 50% and is insignificant only in Eritrea. On the other hand it is interesting to note that there seems to be good progress in exchange rate management across COMESA, as can be inferred from the parallel market premium which is very encouraging.

The major conclusion that can be drawn from the major macro features of the COMESA member is that macroeconomic stability across members is not yet achieved. More importantly, major macro variables do not seem to be sustainable as they are characterized by wide fluctuation. It seem logical to conclude that fiscal policy harmonization in the existing macro environment in COMESA is a daunting task.

B. Economic Community of Western African States (ECOWAS)

GDP growth in ECOWAS countries was quite impressive. In almost all members this rate was above 3.5 percent and fairly stable. Inflation however is a problem across the REC. The inflation rate is very high reaching as high as 70 percent in some countries. The level of monetization of the economy, except in Cape Verde, is also very low. The share of saving and investment in GDP is very small. Moreover, the level of investment is invariably, except in Guinea, is about double the magnitude of saving. This, as shown in the analytical framework, implies a fiscal deficit that is reflected as an external balance problem. This can be read form the export and import intensity of the region. In all countries the export to GDP ratio is about 10 percentage points above the import to GDP ratio. In some of them it actually is twice the size of imports. This internal and external balance problem seems to show itself not only through high level of inflation (see section two above) but also through the very high debt to GDP ratio - which invariably is above 50 percent and for most countries above 80 percent. There is the positive record of low parallel exchange rate premium across the region. However, some countries such as Guinea-Bissau, Nigeria and Senegal seem to have some problem with the exchange rate market. In sum, notwithstanding the positive record on growth, other macro indicators of ECOWAS are not impressive. Fixing this problem need to be considered as part of the strategy to bring fiscal policy harmonization.

C. Arab Maghreb Union (AMU, North Africa)

Growth rate performance in AMU is dissimilar. In Mauritania and Tunisia, the performance is very good. In Algeria and Morocco, on the other hand, growth is quite erratic. Prices (inflation), however, are very stable (Algeria being the latest addition to this success). By standard of other RECs, the AMU members have high degree of monetization of the economy. Except in Mauritania, the saving investment gap is very narrow. In addition, the ratio of saving and investment to GDP is very high - at least by standard of other African RECs. Similarly, the trade intensity is very high. In line with our discussion in the analytical section, the gap between import and exports is very narrow - showing a healthy internal and external balance. Notwithstanding this healthy external balance, the level of debt is very high. This may reflect unhealthy history of external and internal balance. As that of other RECs, the premium of the parallel market rate, except in Algeria, is negligible. Compared to other RECs, the AMU has relatively better macro balance. This is positive initial condition which could facilitates fiscal policy harmonization efforts

D. Southern Africa Development Community (SADEC)

The pattern of growth across SADEC members is not uniform. In about six countries it is erratic, in two of them very bad and in another three it is very good. In the rest of the members (including South Africa which has a strong impact on the SADEC members) there is a declining trend. Similar variation is also observed on the score of inflation. In majority of the countries it is below 10 percent. However, there are worst performers such as Angola, Malawi, Mozambique and Zambia where the inflation rate has reached above 100 percent. The region is also characterized by very low level of monetization. Saving and investment in SADEC have distinct pattern, which reflects the level of development of member states. In the relatively well-developed countries (Mauritius, South Africa, Botswana, and to some degree in Zimbabwe) these ratios are not only high (although the S. Africa figure is low) but also very close to each other. Botswana is an exception where saving is greater than investment. In the relatively poorer countries (such as Tanzania, Malawi and Mozambique) the investment rate is nearly twice the saving rate. In the rest of the countries the investment rate is larger than the saving rate. Thus, in majority of the SADEC member countries there is a domestic resource gap, which is reflected as an unhealthy gap between the share of imports and exports in GDP. There are a few countries where the import and export shares are equal (again Botswana is an exception where its exports are larger than its imports). The direct reflection of this is that each member country has distinct debt profile. The majority of the members (except Botswana and South Africa) have a huge debt burden. The exchange rate policy pursued in the region seems to narrow the parallel market premium across SADEC. Notwithstanding the latter, macroeconomic policy harmonization in SADEC is a real challenge.

E. Customs and Economic Union of Central African States (UDEAC)

The growth rate performance in UDEC is quite erratic. However, the growth rates in Cameroon and Gabon seem relatively stable. The level of inflation, especially in the mid 1990s across UDEC was worrisome. There is, however, a sustained decline in inflation in recent past. This is also a region characterized by very low degree of monetization. With the exception of Cameroon and Gabon, the investment rate is larger than the saving rate among member countries. This is reflected in the relative size of exports and imports. The domestic resource gap is mirrored as external balance deficit in all but Gabon and Cameroon. In the latter two countries, exports are larger than imports (reflecting perhaps the impact of oil revenue in these countries). Again the parallel market premium is negligible in this region too.

F. Economic and Monetary Union of Western Africa (UEMAO, formerly CEAO)

The growth rate of GDP is UEMOE is quite impressive. In almost all countries the recent growth rate is above 4 percent. Inflation, which was a major problem in mid 1990s, is also stabilizing below 5 percent recently. The degree of monetization of the economy is very low, however. The macro performance in terms of saving and investment is not impressive either. In all cases, except in Cote d'Ivoire, the level of investment is nearly twice that of the saving rate. This shows not only unsustainable fiscal balance but also a problem in the external sector. This can be read from the comparison of import and exports. Except in Cote d'Ivoire, the import ratio is invariably higher than the export ratios by about 10 percentage points. In three countries it is actually twice that of imports. This is reflected on the huge debt burden of the member countries. In tandem with other RECs, the parallel market premium is negligible - reflecting some degree of success on exchange rate policy.

In sum, in this section we have attempted to briefly examine the evolution of major macroeconomic variables in the second half of the 1990s. The analysis is informed by the internal and external balance framework outlined in section two. The whole purpose of examining the macro context of REC members is to judge whether there is stable and sustainable macro economic environment, which is a fundamental aspect of macro (fiscal) policy harmonization. From the examination of the data organized across the RECs, the following main points seem to emerge. First, we noted that all the RECs are besieged by internal and external balance problems. To the extent that internal balance problems could partly be addressed by fiscal policy, and since this could have implication on the external balance, fiscal policy issues have a direct bearing on regional integration efforts. The more harmonized such polices are the more effective they will be. Second, the macroeconomic environment in all RECs does not seem to be stable. The instability comes not only from domestic policy problems but also, and perhaps more importantly, from external shocks. Successful fiscal policy harmonization requires creating stable macro environment. This may be handled well by policy coordination. Third, there is variation across RECs in terms of major macro problems they do face. For instance in COMESA growth is a problems while in ECOWAS inflation is much more important. The design of fiscal policy harmonization needs to take such difference on board. Fourth, all RECs are characterized by low level of monetization of the economy. This may point to the importance that RECs need to give to fiscal, as opposed to monetary policy, issues in the short to medium term. Finally, exchange rate policies pursed in much of African countries seem to bear fruit across RECs. Fiscal policy harmonization needs to build on this success. This needs to be complemented by solving the indebtedness problems, which is pervasive across all RECs.

3.2.2 Recent Fiscal Performance of African RECs

One of the convergence criteria that is usually found in many integration schemes in Africa is the ceiling on fiscal deficit and/or inflation targeting (the latter implicitly targets money supply and fiscal deficit). When such convergence criteria is drawn, as noted by Cangiano and Mottu (1998), it is imperative not to suppress the symptoms of an excess deficit bias without eliminating the cause which may be deeply rooted in countries institutional arrangements, such as their budget procedure. This is in particular true in many African countries where such institutional weakness is pervasive. Future research along this line could inform macro policy coordination efforts across RECs in Africa. In this section, owing to lack of detailed country level studies, we will attempt to pinpoint at the salient features of the fiscal posture of African RECs using macro level data.

A): Common Market for Easter and Southern Africa (COMESA)

Across all COMESA member countries expenditure (as the share of GDP) is larger than the revenue. As a result, all members registered a budget deficit (both with and without grants). The share of expenditure and revenue in GDP varies across members. Only in nine countries does the revenue to GDP ratio is above 20 percent. In contrast the share of government expenditure in GDP is on average above 25 percent in fourteen countries. This shows not only collection of revenue is problematic but also countries are living beyond their means. Thus, in all member countries foreign financing of the deficit is becoming the norm. This is reflected by the growing trend of the share of domestic and foreign debt as percent of GDP. The share of domestic and foreign debt in total GDP is not uniform across members. In eight out of twenty member countries domestic indebtedness is important to finance budget deficit. This has resulted in building up of domestic debt, which reached up to sixty percent of GDP in some countries such as Egypt and Seychelles.

The structure of revenue in COMESA shows that tax revenue is the main source of government revenue. In almost all countries it constitutes an average of over 75 percent of government revenue. In all COMESA member countries, except in Angola, indirect taxes are much more important than direct taxes. In fact there is 10 to 20 percentage points difference between indirect and direct taxes as the share of total revenue. This may point to the degree of emphasis that need to be given to indirect tax harmonization at early stage of fiscal policy harmonization scheme. The disaggregation of indirect taxes into domestic and international trade taxes shows that the latter is very significant in COMESA. This has repercussion for regional integration as countries may lose it with further integration (see section 3.3 below, however). The dependence of COMESA members on either domestic or international trade tax seem also to depend on the level of development of member countries. Relatively developed countries (such as Egypt and Kenya) seem to rely on domestic taxes while the others seem to depend much on international trade taxes. Country level further study about this issue may help to introduce safeguard clauses for some countries that could be incorporated in the design of fiscal policy harmonization.

B. Economic Community of Western African States (ECOWAS)

Like that of COMESA, a level of expenditure that is in excess of revenue characterizes EOWAS members. As a result in the second half of the 1990s, all members recorded a deficit (with and without grants). The only exception is Nigeria where in three of the five years examined it had a surplus - related to oil revenue windfalls. There was also a sizable gap between deficit including and excluding grants - indicating the significance of grants in financing deficit. This is not the case, however, in Cont d'Ivoire, Nigeria, Sierra Leone and Togo. Foreign financing of the deficit varies across members. In some countries it reached as high as 10 percent of GDP while in others it is negligible. In majority of the countries, however, it is below 5 percent of GDP. Although outstanding domestic debt is important in few countries, such as Cote d'Ivoire and Nigeria, outstanding foreign debt is a major problem in ECOWAS. In almost all countries the foreign debt to GDP ratio is above 100 percent.

Tax revenue is an important source of government revenue in ECOWAS. In five countries it ranges about 50 to 65 percent of total revenue. In the rest of the member states (except in Guinea-Bissau, where it ranges from 23 to 43 percent) it constitutes over 75 percent. Within the tax category, COMESA members extremely rely on indirect taxes. Indirect taxes, as percent of total revenue, are on the average three times larger than direct taxes. Examining the indirect taxes using its disaggregation into taxes on domestic goods and taxes on international trades shows the existence of two groups of countries. In about eight countries where we have data, tax on international trade is very important. However, in there countries it is not as important as taxes on domestic goods. It can also be seen from the available data that there is a rising trend of taxes on domestic goods in many of the ECOWAS member countries.

C. Arab Maghreb Union (AMU, North Africa)

In contrast to other RECs, the share of revenue and expenditure in GDP in AMU is extremely closes (about 30 percent) resulting in a reasonable deficit level that ranged from less than 1 percent to a maximum of about 5 percent. Grants do not seem to be important in AMU, as there is no fundamental difference between the two definitions of deficit (with and without grants). Outstanding domestic debt is not significant either. However, the external debt to GDP ratio is very high: between 60 to 85 percent for Morocco and Tunisia, and between 80 to 240 percent for Algeria and Mauritania.

Tax revenue is the most important source of government revenue in AMU. For those countries where we have complete data, indirect tax is relatively more important than direct taxes. The relative importance of indirect taxes on domestic goods and international trade varies across members. In Algeria and Morocco both categories contribute equally to government revenue. In Tunisia, taxes on domestic goods are much more important than taxes on international trade, while in Mauritania the reveres is true. This variation of the importance of different taxes across member has implication for designing appropriate fiscal harmonization schemes. Fiscal harmonization policies need to address such difference in initial conditions. If this is not done, the cost of fiscal harmonization may vary across members - impacting on their desire to join such arrangement differently.

D. Southern Africa Development Community (SADEC)

Except in Botswana, a very high level of expenditure characterizes all SADEC members, which is invariably larger than revenue. This has resulted in a high level of budget deficit. Except in Lesotho, Mauritius, South Africa and Swaziland, the other members have a budget deficit in excess of 10 percent of their GDP. Except in Lesotho, Malawi, Mozambique and Zambia, grants don't seem to be important in reducing the level of budget deficit in SADEC. In general, foreign financing of deficit is not important across SADEC. External indebtedness is a serious problem in majority of the member countries (except in Botswana, Mauritius and South Africa). Domestic indebtedness is important in Mauritius, Seychelles, South African and Zimbabwe.

Tax revenue in SADEC is the main source of government revenue, contributing on average above 80 percent to total revenue. In a couple of countries it has dropped to about 50 percent. However, in majority of the countries it is above 90 percent. In almost all countries for which we have complete data, except in Angola, the indirect tax is nearly twice that of direct taxes. Its contributions also ranges from 50 to 70 percent of total revenue. When we disaggregate indirect taxes into the domestic goods and international trade, there is no distinct pattern across SADEC members. For some countries tax on domestic trade is relatively important (Angola, Botswana, and South Africa are in this category). For the others tax on international trade is important (this list may include: Lesotho, Mauritius, Seychelles and Tanzania). There is also a third group where the contribution of the two types of taxes is nearly equal (Mozambique and Namibia). It may be sensible to infer that mineral-based economies and relatively developed economies increasingly rely on taxes on domestic goods (compared to tax on international trade).

E. Customs and Economic Union of Central African States (UDEAC)

Although, like other RECs, UDEC member countries are characterized by the level of expenditure, which is in excess of their revenue (and hence budget deficit), they are unique in that the level of both expenditure and revenue as the share of the GDP is very low by the standard of other RECs. The lowest revenue is in Chad (about 5 percent of GDP) and the highest in Gabon (about 30 percent of GDP). In three of the countries (Cameroon, Central African Republic and Chad) the role of grants in ameliorating the budget deficit problem is very important - it reduced the budget deficit nearly by half. In the rest it was not important. Although the level of domestic debt is very small in all countries (except in Gabon), external indebtedness is a major problem in all countries.

The bulk of government revenue in UDEC comes from tax revenue. Indirect taxes are very important in Cameroon, Central African Republic and Gabon. In Congo Republic and increasingly in Equatorial Guinea, direct taxes are becoming very important. Disaggregation of indirect taxes into domestic and internal trade taxes shows that only Equatorial Guinea, Gabon and to some degree Chad have a relatively higher share of taxes on international trade. For the rest, tax on domestic goods is relatively better.

F. Economic and Monetary Union of Western Africa (UEMAO, formerly CEAO)

In all member states of UEMAO, expenditure is in excess of revenue. As a result all members have a budget deficit in the last five years. Like that of UDEAC, the share of both revenue and expenditure in GDP is very low (revenue ranging 10 to 20 percent and expenditure 12 to 35 percent). Grants are very important across member states and helped to reduce the budget deficit nearly by half. Financing of deficit relies heavily on external resources. Perhaps relating to this, the level of external debt is very high in all countries. In contrast, outstanding domestic debt is important only in Cote d'Ivoire.

The data on the structure of government revenue shows that UEMAO members, except Guinea Bissau, rely extremely on tax revenue. Within this tax category, indirect taxes are extremely important. On the average the contribution of indirect taxes is three times larger than direct taxes. A disaggregate picture of indirect taxes shows that in all countries tax on international trade is much more important than taxes on domestic goods. The only exception is Burkina Faso where recent trend shows the increasing importance of taxes on domestic goods. The importance of international trade tax in UEMAO suggests the possible (relative) higher negative impact of tariff reduction schemes on its members.

To sum up, in this section we have attempted to examine the basic fiscal features of African RECs. This is important since it could give us a clue about the recent fiscal structure of African RECs. This in turn may help to have a good idea of the challenge of fiscal policy harmonization in the ongoing regional integration effort. From the examination of the fiscal data of the various RECs, the following points seem to stand out. First, we noted that tax revenue is extremely important across RECs. Within this category indirect taxes in general, and international trade taxes in particular are found to be important. Thus, fiscal policy harmonization needs to focus on such taxes and their importance (and hence associated costs of integration) across members of RECs. Second, the analysis shows that the fiscal structure of countries has an enormous variation not only across RECs but also with a particular REC. This has implication for designing appropriate fiscal policy, which suits each cases and introducing safeguard clauses when some countries are adversely affected by harmonization efforts. Third, there seem to be a positive relationship between the level of development of a country and domestic indebtedness. This may be because of private capital inflows or strong exchange rate market where foreign exchange is not a main constrain. In such countries domestic indebtedness may serve as a substitute for external indebtedness. Designing sustainable fiscal policy harmonization needs to take such differences on board. In relation to this, since all the RECs are characterized by extreme level of external indebtedness, debt relief, by offering relative good fiscal position, can enormously ease the effort at fiscal policy harmonization. Finally, although due to lack of disaggregated data we have not examined the structure of expenditure in detail, the aggregate level of expenditure across RECs is very high. Fiscal policy harmonization strategy for the RECs needs also to take this issue on board.

3.2.3 The Convergence Criteria of African RECs

Annex II shows the macro convergence criteria drawn by African RECs. Although all the RECs do not have convergence criteria, some of the major RECs have target level of major macro variables. This includes, inter alia, inflation (3 to 10 percent), limits on budget deficit (0 to 10 percent of GDP), a positive lending and deposit rate, as well as liberalization of current and capital account.

The comparison of the performance of RECs, as discussed above, shows that the current performance is far below the convergence criteria set by the RECs' themselves. Success is largely limited to tackling the problems of overvalued exchange rate, attaining the target level of revenue to GDP ratio and, in some countries, the level of inflation. In general, it is safe to conclude that: first, in majority of countries (and RECs) the convergence criteria set out is not met; second, in all RECs the macroeconomic stability required for fiscal policy harmonization is not met either; third, there is a large variation in the performance of members of RECs. In a particular REC, it is not difficult to see both good and poor performers. Thus, the African RECs are not only far from achieving fiscal policy harmonization but also are characterized by weak macroeconomic condition and fiscal structure. Thus, realizing fiscal policy harmonization is a challenge to the on going integration effort. To help address this challenge, section 3.3 below briefly examine some major outstanding issues related to fiscal policy harmonization in the context of African RECs.

3.3 Some Major Outstanding Issues

3.3.1 Revenue Loss

Reducing trade barriers in economies where tariff revenue is one of the most significant sources of government revenue complicates the inter-temporal trade off between the apparent short-term loss of revenue and the expected long-term benefits emanating from regional integration. This has been cited by many countries as one of the major problems of regional integration in Africa. In countries, which trade a lot within a given REC, government revenue loss due to integration could be large. This is because tax revenue from international trade constitutes the main source of tax revenue in many African countries. For others such as Ethiopia, which trade less with its REC members, static revenue loss due to opening its market to COMESA is less than 1 percent of total revenue (although shifting from EU to COMESA could mean a lot of loss in tax revenue). Although estimating revenue loss requires detail country level study, we have used two sources of information to shed light on the possible magnitude of revenue loss.

Table 1, based on Teshome (1997), is constructed by disaggregating tax revenue from international trade into its two source: intra-REC and intra-African trade on the one hand and trade with non-African countries on the other. By assuming full liberalization on the former, the estimated revenue loss (as percent of GDP) is arrived at. As can be read from the Table, revenue loss is less than 0.5 percent of GDP across region and time. Thus, revenue loss is not a serious problem (see Table 1 for detail)¹⁹.

Table 1: Tax Revenue Loss From Full Liberalization (as percent of regional GDP)

Source: Teshome Mulat, 1997: 169-170.

Note: figures in parenthesis are standard deviations of the yearly observations.

To show the variation of the revenue loss across countries, we have used a sample of countries in COMESA to illustrate the possible impact of integration on government revenue. Table 2 provides a static estimation of the magnitude of revenue loss if member countries abolish tariff among themselves. The Table needs to be taken cautiously as it doesn't entertain both the possibility of shifting to COMESA suppliers and an institutionalization of a common external tariff which would be lower than the rate currently in use by members on third countries.

Table 2: Estimated Revenue Loss From Further Integration in COMESA

(Percent of Total Revenue, Excluding Grants)

Note: The rates are computed as the product of taxes on international trade and the share of each country's trade in total COMESA trade, based on World Bank, African Database (2000).

Table 2 shows that the average revenue loss is extremely small (3 to 3.5 % of government revenue excluding grants). This average however masks the possible adverse effect on some countries such as Malawi, Rwanda, Tanzania, and Uganda. However, in the opinion of some experts at COMESA this loss is insignificant especially in the light of the increasing importance of value-added and income taxes as well as the already low level of tariffs especially in capital goods and raw materials. There may also be dynamic gain from growth spurred by integration. If the problem is pressing, the COMESA experts argue, it can be handled by focusing on specific countries that have critical problems, and designing a compensation mechanism for their loss (Alemayehu and Haile 2001).

In sum, there are three major points that we could make about the issue of revenue loss. First, the macro-data based estimation shows that despite governments' fear, revenue loss may not be a major problem. This is the reflection of the low level of Intra-Africa trade. Moreover, the revenue loss could increase over time - suggesting the less costly nature of doing it now than latter. Second, if all members shift a good part of their trade towards REC members, revenue loss could appear as a major problem. This doesn't seem to occur in the short to medium run, however. Finally, a possible short to medium run problems may come from weak competitive position of firms in some of the countries. This will have an effect not only on domestic revenue but also on country level industrial policy. Thus, fiscal policy harmonization needs to take these issues on board.

3.3.2 Issues of Sovereignty and Institutional Aspect of Macro Policy Coordination

There is tension between surrendering the sovereignty of policy making and the gain that could be obtained from macro policy coordination. It is therefore logical to expect such tension in African RECs. The EU dealt with this problem using the so-called "Subsiderity principle" as discussed in section two above. This principle can also be the basis for fiscal policy harmonization in Africa.

Regional level policy coordination will also require an institutional set-up for evaluating and monitoring the agreed upon policy measures such as the convergence criteria adopted by members of African RECs (see Annex II). This may require an elaborated institutional set up and an effective surveillance mechanism. The experience in G7 countries (see Goldestein 1994) shows that member countries submit to the IMF their short-term projection of major macro variables for the current and next calendar year. These variables include, GNP, real total demand, consumer price index (CPI) and the current account balance, expectation of fiscal and monetary development over the coming three years. This is supplemented by the Fund's forecast (which is usually based on its global model called MULTIMOD²⁰). Fiscal out turns for the current and previous year are also given to examine possible forecast errors. High frequency data on exchange and interest rates are also form part of the information package. This forms the background material for governors of central banks and finance ministers meeting which takes twice a year. One of the important feature of this meeting is that expert opinion on the fiscal outturn and related issues are considered much more important than the figures that formed part of the background paper. Thus, establishing such institutional structure is very important.

Once the institutional framework for harmonization of macro polices are established, the next task is to agree on major convergence criteria and ensure its implementation. This requires working on surveillance mechanism aimed at monitoring progress towards the agreed upon targets. This needs to be backed by enforcement mechanisms so that the convergence criteria are adhered too. Examining similar problems in EU, Masson (2000) noted three general models that could handle coordination of fiscal policy. First, countries could agree to harmonize their tax and expenditure polices. Enforcing such agreement might be difficult without some EU-wide institutional involvement²¹. Second, governments could agree upon a common program administered by relevant European institutions- some sort of fiscal federalism. Third, coordination could involve intergovernmental surveillance over national polices, but no binding constraints on the exercise of nation sovereignty. In the latter model, coordination would result from peer pressure. At point in time it is also possible to see the operation of the three models in different aspect of fiscal policy (Masson, 2000: 12). Masson correctly argued that which model eventually prevails will be determined in the political arena, and economic considerations will be only one, and perhaps not the most important, consideration. It is imperative to note that establishing the institutional framework for fiscal policy harmonization is a very difficult process for advanced European countries, let alone for countries in Africa. In fact, after examining the experience of ECOWAS, Masson and Pattillo (2001) concluded that fiscal policy harmonization, in particular designing and implementing a fiscal restraint, is a difficult task because: (a) countries may not agree in the definition of the most relevant fiscal deficit concept, (b) the monitoring of compliance rules may be circumvented through illusory fiscal adjustment and creative accounting and, finally (c) it is not clear that a sanction mechanism is a feasible way to deter violations of fiscal restraints.

The lesson that could be drawn from this experience is that establishing the institutional framework for macro (including fiscal) policy harmonization and maintaining the surveillance mechanism to monitor the convergence criteria drawn is a daunting task. It requires: the submission of certain degree of autonomy in domestic policy making, establishing an elaborated and transparent institutional mechanism (or harness the existing ones) both to design feasible convergence criteria and to use it as surveillance mechanism and finally a skilled labour force equipped with a research wing which could undertake rigorous economic analysis. Thus, to have effective macro policy coordination across countries, African RECs need to address these issues both at regional and continental level.

3.3.3 Asymmetry of Shocks and Policy Harmonization

Another major problem that is commonly mentioned in the context economic integration relates to policy harmonization that result from asymmetry of shocks. One such problem relates to external shocks. The fact that most African countries are characterized by trade in primary commodities implies that they are vulnerable to external shocks such cyclical commodity prices and terms of trade deterioration. In a typical REC such shocks may not be well correlated (could have asymmetric impact). This could be due to variation in production structure of members of RECs²². In such context, it is possible that member countries may follow different macro (including fiscal) policies to safeguard their economies from macro imbalances. If shocks are different, the policies employed to contend them could also differ. This creates harmonization problem - a divergence from the convergence criteria- at the level of REC or across RECs. In such context, as argued by Vos (2001), based on the experience of Merccosur, it is sensible to consider the short-term restrictions and accept differences in policy regimes as they currently exist and enter a gradual process of setting common targets for macroeconomic stability. Full macroeconomic coordination under such asymmetries reinforces the need for compensatory funds and sharing of fiscal adjustment costs (see Vos, 2001: 8-9).

Asymmetries may also emerge from the size of member countries in relation to the use of `Cohesion funds' designed to subsidize poor regions or `compensatory funds' of RECs (see Masson and Pattillo, 2001). As noted by Masson and Pattillo (2001), in the context of their study on ECOWAS, `given the size of Nigeria, relative to its neighbors, the operation of such a fund will be asymmetric. Transfers to the smaller countries if they get into difficulties could be sizable, but if Nigeria were to draw, it could quickly exhaust available resource of the fund.'

The above discussion shows that asymmetry of shocks and its policy implication could be a major constraint to fiscal (macro) policy harmonization both within and across REC. Thus, designing fiscal policy harmonization needs to take such variation in policy responses to external shocks on board. Outlining details of activities for the short and the long run, as well as designing an optimal way of using cohesion/compensatory funds could be helpful way of approaching this problem.

IV. Conclusion

In this paper an attempt to examine the issue of fiscal policy harmonization across African RECs is made. The analysis commenced by discussing the rationale for fiscal policy harmonization. Issues of macro spillover effects and tax competition are taken as some of main arguments for fiscal policy harmonization. Attempt has also been made to draw relevant lessons from the theory of fiscal federalism. This is followed by section 2.2 where we have attempted to establish the analytical framework that helps to inform the discussion in the rest of the paper. This is drawn from the macroeconomics of the internal and external balance.

The rest of the study focused on examining the genesis of the African fiscal problem, the recent macro performance of African RECs, as well as their fiscal posture. We noted that the inherited social and economic structure of post-independence African states, in particular, the dependence on primary commodity trade, which is characterized by cyclical and declining trend of prices, has greatly contributed to the weak fiscal position of African countries in the 1980s. This is compounded by expenditure management problem.

Although various specific points and conclusions are made and noted in the paper, the following major conclusions are worth emphasizing. First, although the SAP followed across the continent has resulted in relatively better fiscal posture and some degree of success in managing major macro variables, the analysis in this paper shows that both the macro environment and the current fiscal posture (including indebtedness) leaves much to be desired. As a result, the convergence criteria drawn are hardly met. This underscores the challenge of fiscal harmonization across African RECs.

Second, revenue loss is usually taken as major problem in integration effort of RECs. However, we noted that, first, the macro-data based estimation shows that revenue loss is not a major problem. This is the reflection of the low level of intra-Africa trade. Moreover, the revenue loss could increase over time - suggesting the less costly nature of doing it now than latter. Second, if all members shift a good part of their trade towards REC members, revenue loss could appear as a major problem. This doesn't seem to occur in the short to medium run, however. Finally, a possible short to medium run problems may come from weak competitive position of firms in some of the countries. This will have an effect not only on domestic revenue but also on country level industrial policy. Thus, fiscal policy harmonization needs to take these issues on board.

The third important point relates to the need to have an institutional framework for realizing fiscal policy harmonization. Establishing the institutional framework for macro policy harmonization and maintaining the surveillance mechanism to monitor the convergence criteria drawn is a daunting task. It requires: the submission of certain degree of autonomy in domestic policy making by member countries, establishing an elaborate and transparent institutional mechanism, and a skilled labour force. Thus, to have effective macro policy coordination across countries, African RECs need to address these issues both at regional and continental level.

Finally, the paper shows that there is an enormous variation across RECs and member countries of RECs in terms of the macroeconomic environment, the fiscal posture, the asymmetry of shocks as well as the policy response to such shocks. This underscores the need to design a fiscal policy harmonization that suits the specific context of each REC with adequate safeguard measure for weaker members.

Reference:

Addision, Tony and Leonce Ndikumana (2001) `Overcoming the Fiscal Crisis of the African State' UN-WIDER Discussion Paper No 2001/12, UN-WIDER, Helsinki.

Adam, S. and D.L Bevan (2001) `Fiscal Policy Design in Low-Income Countries' UN-WIDEDR Discussion Paper No. 2001/67, UN-WIDER, Helsinki.

Alemayehu, Geda (2001). Finance and Trade in Africa: Modeling Macroeconomic Response in a World Economy Context. London: Macmillan (forthcoming).

Alemayehu, Geda (1997) `The Historical Origins of African Debt Crisis' ISS Working Paper, Money, Finance and Development Series No. 62, The Hague, The Netherlands.

Alemayehu, Geda, E.V.K. FitzGerald, K. Saramad and R.Vos (1992), `Trends of Public and Private Capital Account Variables in Developing Countries', Institute of Social Studies, The Hague (mimeo).

Alemayehu, Geda and Haile Kibret (2001) `Regional Economic Integration in Africa: A Review of Problems and Prospects with A Case Study of COMESA' (forthcoming in Department of Economics, AAU Working Paper).

Aryeetey, E (2000) `Regional Integration in West Africa' (Paper Presented at A Policy Workshop organized by the OECD Development Center at the Graduate Institute of International Studies, Geneva, October 2000)

Bacha , E.L. and C.F. Diaz Alejandro (1982). International Financial Intermediation: A Long and Tropical View. New Jersey, International Finance Section, Department of Economics, Princeton University.

Balassa, Bela (1983) `Policy Response to External Shocks in Sub-Saharan African Countries', Journal of Policy Modeling, 5 (1):75-105.

Balassa, Bela (1984) `External Shocks and Policy Response in Sub-Saharan Africa, 1973-78', Finance and Development, 21 (1): 10-12

Cangiano, Macro and Eric Mottu (1998) `Will Fiscal Policy b e Effective Under EMU' IMF Working Paper, WP/98/176, IMF, Washington, D.C.

Chowdhury, Aris and Iyanatul Islam (1993). The Newly Industrializing Economies of East Asia. London: Routledge.

Cuddington, John T. (1996) `Analyzing the Sustainability of Fiscal Deficit in Developing Countries', Economics Department, Georgetown University.

Dinh Hinh T (2001) Fiscal Solvency and Sustainable in Economic Management, The World Bank, Southern Africa Region.

De Gauwe, Paul (1994). The Economics of Monetary Integration. Oxford: Oxford University Press.

Elbadawi, Ibrahim (1997) `The Impact of Regional Trade and Monetary Scheme on Intra-Sub-Saharan African Trade' in A.Oyejide, I. Elbadawi and P.Collier (eds). Regional Integration and Trade Liberalization in Sub-Saharan Africa, Volume I: Framework, Issues and Methodological Perspectives. London: Macmillan.

Ezenwe, Uka (1991) `The African Debt Crisis and the Challenge of Development', Intereconomics, 28 (January/February):35-43.

Fantu Cheru (1992) `Structural Adjustment, Primary Resource Trade and Sustainable Development in Sub-Saharan Africa' World Development, 20 (4): 497-512.

Frenkel J.A. and A.Razin (1996). Fiscal Policies and Growth in the World Economy. Cambridge, MA: The MIT Press.

Ezenwe, Uka (1991) `The African Debt Crisis and the Challenge of Development', Intereconomics, 28 (January/February):35-43.

Ghura, Dhaneshwar (1998) `Tax Revenue in Sub-Saharan Africa: Effects of Economic Policies and Corruption' IMF Working Paper, WP/98/135, IMF, Washington, D.C.

Goldstein, Morris (1994) `Improving Economic Policy Coordination: Evaluating some new and Some not-so-new Proposals' in P.B. Kenen, Francesco Papadia, Fabrizio Saccomanni (eds.). The International Monetary System. Cambridge, MA: Cambridge University. Press.

Haggard, Stephan (1990). Pathways from the Periphery: The Politics of Growth in the Newly Industrialized Countries. Ithaca: Cornell University Press.

Husain, Ishrat (1994) ` Structural. Adjustment and the Long-term Development of Sub-Saharan Africa' in van der Hoeven and F. van der Kraaij (1994). Structural Adjustment and Beyond in Sub-Saharan Africa. London: James Currey.

Khan, Mohsin S. And M.Knight (1983) `Sources of Payment Problems in LDCs', Finance and Development, 20(4): 2-5.

Krugman, Paul R. and Maurice Obstfeld (2000). International Economics: Theory and Policy, Fifth Edition. New York: Addison-Wesley.

Krumm, Kathie L. (1985) `The External Debt of Sub-Saharan Africa' World Bank Staff Working papers Number 741, Washington D.C: The World Bank.

Masson, Paul and Catherine Pattilo (2001) `Monetary Union in West Africa: An Agency of Restraint for Fiscal Policies? IMF Working Paper, WP/01/34, IMF, Washington, D.C.

Masson, Paul (2001) `Fiscal Policy and Growth in the Context of European Integration' IMF Working Paper, WP/00/133, IMF, Washington, D.C.

Mbelle, Ammon and Thomas Sterner (1991) `Foreign Exchange and Industrial Development: A Frontier Production Function Analysis of Two Tanzanian Industries', World Development, 19 (4): 341-347

Mistry, Percy S. (1988). African Debt: The Case for Relief for Sub-Saharan Africa. Oxford University, Queen Elizabeth House: Oxford International Associates.

Mistry, Percy S. (1991). African Debt Revisited: Procrastination or Progress? The Hague: FONDAD.

Mohammed, Baba G. (1989). Oil boom and Structure of Nigerian Economy 1970-87 (Unpublished MA thesis). The Hague: Institute of Social Studies

Ndulu, Benno J. (1986) `Investment, Output Growth and Capacity Utilization in an African Economy: The Case of Manufacturing Sector in Tanzania', Eastern Africa Economic Review, 2 (1): 14-31.

Ndulu, Benno J. (1991) `Growth and Adjustment in Sub-Saharan Africa' in Ajay Chhibber and Stanley Fischer (eds). Economic Reform in Sub-Saharan Africa. Washington D.C: The World Bank.

Ndung'u, Njuguna (2000) `Regional Integration Experience in Eastern African Regions' (Paper Presented at A Policy Workshop organized by the OECD Development Center at the Graduate Institute of International Studies, Geneva, October 2000)

Ngwenya, A.R. and P.K.Bugembe (1987) `The origins and Causes of the African Social and Economic Crisis', in International Conference on Africa: the Challenge of Economic Recovery and Accelerated Development (Abuja, Nigeria 15-19 June).

Oates, Wallace E. (1999) `An Essay on Fiscal Federalism', Journal of Economic Literature, XXXVII (September): 1120-1149.

Shah, Anwar (1997) `Fiscal Federalism and Economic Governance: For Better or for Worst? Paper Presented at International Conference on Decentralization, Intergovernmental Fiscal Relations and Macroeconomic Governance, Brazil.

O'Connell, Stephen (1997) `Macroeconomic Harmonization, Trade Reform and Regional Trade in Sub Saharan Africa' in Ademola Oyejide, Ibrahim Elbadawi and Paul Collier. Regional Integration and Trade Liberalization in Sub Saharan Africa: Framework, Issues and Methodological Perspective, Volume I. London: Macmillan.

Rattso, Jorn (1992) `Import Compression, Investment Determination and Macroeconomic Adjustment: A Macroeconomic Model Relevant for Sub-Saharan Africa', Development Economics Seminar Paper NR. 92-1/4a. The Hague: Institute of Social Studies.

Rattso, Jorn (1994) `Medium-run Adjustment under Import Compression: Macroeconomic Analysis Relevant for Sub-Saharan Africa' Journal of Development Economics, 45: 35-54.

Tanzi, Vito and Howell H. Zee (1998) `Consequence of the Economic and Monetary Union for the Coordination of Tax Systems in the European Union: Lessons from the US Experience', IMF Working Paper, WP/98/115, IMF, Washington, D.C.

Teshome Mulat (1997) `Trade Liberalization and Government Tax Revenue Loss in Africa', Journal of World Trade, 31(1): 161-174.

Teshome Mulat (1998) `Multilateralims and Africa's Regional Economic Communities', Journal of World Trade, 32(4): 115-138.

Tunrovsky, Stephen (1997). International Macroeconomic Dynamics. Cambridge, Ms: The MIT Press.

van der Hoeven, Rolph (1993) `External Dependence, Structural Adjustment and Development Aid in Sub-Saharan Africa', (Paper presented at an International Policy Workshop on International Capital Flows and Economic Adjustment in Developing Countries. The Hague: Institute of Social Studies).

Vos, Rob (2001) `Currency Volatility and Macroeconomic Policy Coordination in Mercosur' (Paper presented at a Conference on `Mercosour after 10 years', .Leiden University, May 3, 2001).

ANNEXES

ANNEX I: Structure of Public Expenditure Following Commodity Boom
ANNEX II: Convergence Criteria of African RECs.
ANNEX III: Recent Macroeconomic Performance of African RECs.
ANNEX IV: Recent Fiscal Posture of African RECs.

Annex I: Structure of Public Expenditure Following Commodity Boom in the 1970s in Selected African Countries

Total Public Expenditure as % of GDP

ANNEX II: Convergence Criteria of African RECs.

Note: n.i = not indicated

Source: Economic Commission for Africa (ECA), Regional Cooperation and Integra ion Division (RCID)

1 Standard macro policy analysis sets the problem of choosing the right exchange rate regime as a trilemma (`unholy trinity'). It says that policy makers have to decide on three things: the desired degree of autonomy in monetary policy, the degree of capital mobility and the degree of nominal exchange rate flexibility (see Vos, 2001:3).

2 The principle states that `in areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the member states and can, therefore , by reason of the scale or effect to the proposed action, be better achieve by the Community' (Article 3b of the Treaty of EU quoted by Canginao and Mottu 1998).

3 Having this general framework it is fairly straightforward to have a disaggregated picture. First, the stock of financial data can be disaggregated. That is, the stock data could be disaggregated according to whether creditors are bilateral, multilateral, concessional, non-concessional, or private, (i.e. banks, portfolio and other commercial suppliers). Hence, instead of using direct flows reported in IMF Balance of Payments statistics, the flow counterpart could be derived from the change in stocks. This allows consistency between the stock and the flow data. In relation to trade data, the total exports could also be disaggregated by the SITC classification chosen for the purpose of the study. Since this classification tallies with the one used by UNCTAD in its Annual Commodity Yearbook, this allows one not only to arrive at aggregated historic data by SITIC, but also to assign UNCTAD-based relevant prices for each SITIC classification. Consistency of total exports, between that reported in the national accounts and the disaggregated UNCTAD data, could be maintained by introducing a category of 'other exports' as an adjusting variable. A similar approach may also be employed for imports (see Alemayehu 2001 and Alemayehu et al 1992).

4 T and G can be disaggregated into different categories of public revenue and spending. We left that in this paper and focused on the aggregate figures. Key fiscal variables for each REC are discussed in the next section, however.

5 This can be formally tested using the Granger non-causality test using time series econometrics.

6 No causality is implied here. It is possible that causality may run form the right- hand side (the external sector ) to the left hand-side (fiscal deficit) (see below and O'Connel 1997).

7 Notice that monetary policy is not a policy tool in fixed exchange rate context since a change in money supply by buying and selling domestic assets will cause an offsetting change in foreign reserves, leaving the money supply unchanged. We have relaxed this assumption below.

8 This list includes, Zambia, Sierra Leone, Tanzania, Ghana, Zambia, Kenya, Malawi, Nigeria, Egypt (See Alemayehu 1997 for detailed information about the evolution of the pattern of trade and finance since 1970 in each of this counties which are picked from each regions, using ECA classification, in Africa).

9 Whether this rise in spending has been a policy mistake is a contestable one. The Bretton Wood Institutions and many analysts argued that was a policy mistake. I have argued elsewhere (Alemayehu 1997) that given the inherited colonial structure that necessitated spending on social and physical infrastructure, the rise in government expenditure is not a policy mistake as such, as seems to be depicted in the good part of the African debt literature. This spending is necessitated by fundamental problems, which were structural/historical, and the resulting policies are the reflection of this reality (see Alemayehu 1997, 2001).

10 The situation was a little different for oil exporters (See Alemayehu 1997 for details).

11 This investment was not without result either. Following this expenditure universal primary education was almost achieved, more health infrastructure was built, infant mortality rate declined more than by third. However, the public enterprises built were seriously affected by recession in the North, high import content (59-60%), lack of domestic demand, which adversely affected their capacity to be self-sufficient (see Mohammed 1989).

12 One common comment is that East Asian countries (such as Korea, Singapore and Hong Kong) that were under colonial rule have developed while Africa is not. Such comments are not credible because the historical parallel is completely different. Hong Kong and Singapore prospered as enterpots owing to direct British colonial interest. Moreover, they are city-states incomparable to African colonies. Probably the only comparable country is Korea and to some degree Taiwan. However, the Japanese colonialism (which was as harsh as the others) had an aim of creating heavy industry and self-sufficiency in its empire, and, hence, has done better than the colonizers in Africa. Some figures may substantiate this point. Taiwan and Korea experienced higher GDP growth than their colonizer (Japan) between 1911-1939; their infrastructure has also developed (Taiwan having 600 kilometers of rails and 3,553 kilometers of road where there were none before). By the end of the colonial period primary school enrolment in Taiwan stood at 71% and similar pattern is observed in Korea. Owing to geopolitical factors (the cold war) Korea, for instance, obtained US $6 billion grants from USA between 1946-78 compared to US $6.89 billion for the whole of Africa. US military delivery to the two countries in 1955-78 stood at US $9 billion, the combined figure for Latin America being US $3.2 billion- one can imagine what the economic impact of this might be (see Chowdhury and Islam 1993). In Korea alone aid financed nearly 70% of total imports and equaled 75% of total fixed capital formation (See Haggard 1990 which also provides the political economy of this event).

13 However, Taiwo (1991) using regression analysis based on data from eleven sub-Saharan African countries (1970-88) noted that the most important factor for the debt crisis was the relative (periphery to center) level of economic development (measured as the ratio of per capita income of LDC to industrial world) and to a lesser degree terms of trade, relative prices, real cost of borrowing and openness of the economy.

14 However, the collapse of oil price from its 1979 hike although relived the oil importing countries it adversely affected oil exporting economies of North Africa and some of the countries in West and Central African regions (mainly Nigeria).

15 Besides, the terms for African countries were harder even compared to South Asian countries. For instance in 1980 African countries on the average had to pay an average interest rate of 6.6% on loans with a maturity of 18 years. The comparable figures for South Asian countries were 3.1% and 30 years (van der Hoeven, 1993:1).

16 An interesting area of further study is to explore the impact of services (especially of insurance and shipping), which seriously affected a number of small countries in Africa.

17 Thus, another major development in the 1980s and early 1990s was the growth of multilateral debt, especially that owed to the World Bank and African Development Bank and to a lesser degree the IMF.

18 The PRSP and MTEF framework is already being applied in more than 18 African countries.

19 Using this method, the estimated tax revenue loss from trade globalization or full implementation of the WTO agreement (i.e., 100 percent loss of taxation from foreign trade) is about 2 percent of Africa's GDP (Teshome 1997).

20 See Alemayehu (2001) for critical review of global models, including MULTIMOD, and their relevance for Africa.

21 In the context of fiscal federalism this is usually achieved through share-cost programs in which the local governments have the incentive to go along with the federal government standards.

22 Masson and Pattillo (2001) comparing the standard deviation of the terms of trade for EU and ECOWAS noted that the latter is extremely vulnerable to such shocks. They also noted that despite all members of ECOWAS do export primary commodities, these shocks are not typically correlated owing to different composition of commodities exported and the production structure of members. This inevitably entails asymmetry of shocks with associated policy harmonization problem.