I. Introduction and Background
Following the devaluation of the CFA franc, trade within the West African Economic and Monetary Union (WAEMU) 1/ is expected to pick up markedly, which will increase the mutual dependence of the member countries. Accordingly, there seems to be a consensus that fiscal policy, which represents the last independent policy instrument in the hands of the member countries, needs to be coordinated more closely in order to reap the full benefits of the monetary union. Plans for a tighter coordination of fiscal policies were pursued in the WAEMU’s predecessors, the West African Monetary Union (WAMU) and the Economic Community of West Africa (CEAO) 2/ for some years. These attempts were largely unsuccessful, as is evidenced by the CEAO’s failure to establish a preferential trading zone. Since the early 1990s, the West African Central Bank (BCEAO), with the support of the IMF, World Bank, and European Community, has been working on harmonization strategy. The aim of these efforts was mainly to promote the political integration in the region and to achieve domestic policy goals through the peer pressure of the union. However, a case can be made that fiscal harmonization may be warranted not only on political, but also on economic, grounds. The paper focuses on this latter aspect. More specifically, we try to spell out the economic rationale and provide an economic analysis of fiscal policies, their (negative) externalities, and their implications for the seven WAEMU countries. Basically, we attempt to answer two questions: (i) why should the countries harmonize their fiscal policies, particularly in the aftermath of the recent devaluation of the CFA franc? and (ii) have the authorities adopted the optimal harmonization strategy?
We conclude that the devaluation makes the elimination of fiscal externalities even more desirable than before. Harmonizing domestic indirect taxes, customs duties, and fiscal operations goes in the right direction, although special attention should be devoted to a consistent and theoretically well-founded implementation. In the following discussion, we will first examine to what extent fiscal externalities between the WAEMU countries exist, and to what extent they have an effect on the countries’ welfare. We then provide some reasons why it is particularly important to address these fiscal externalities now. Finally, we evaluate the authorities’ harmonization measures in the light of the theoretical and empirical findings presented earlier.
This paper is confined to an analysis of fiscal policy coordination and its implications within the WAEMU only; the effects on the community’s trade with other countries are not examined. This narrows the scope of the paper somewhat, since fiscal and nonfiscal (including exchange rate adjustments) externalities stemming from trading partners like Nigeria, Ghana, The Gambia, and France have strongly affected WAEMU economies. In addition, we define fiscal policy quite narrowly as those government activities leading to fiscal externalities. As we will argue below, this concerns mainly tax and tariff policies, as well as the size of the budget deficit.
II. Fiscal Policies Before the WAEMU Treaty
In a monetary union, more than in any other setting, one country can influence its trading partner’s welfare by its fiscal policy. Such effects are usually referred to as fiscal externalities. Strictly speaking, this paper deals only with negative externalities, i.e., with ways in which one country can harm its partner countries.
The impact on the other members of the union is twofold. On the one hand, there is the well-known impact on the efficiency of international resource allocation and the volume of trade. This effect is indirect, since it leads to a welfare loss for the zone as a whole; its origins lie in distortions at the microeconomic level. On the other hand, there is the more direct beggar-thy-neighbor effect: fiscal instruments like the tax and tariff system (at the microeconomic level) or the overall budget deficit (at the macroeconomic level) can be designed in such a way that there is an immediate welfare gain at the cost of partner countries.
In a monetary union like the WAEMU, a group of countries deliberately gives up one policy instrument (monetary and exchange rate policies) in return for more monetary stability, no exchange rate risk, and transparency of their foreign trade transactions. A general conclusion from the literature on currency unions is that in such a setting the externalities of adverse fiscal policies (especially deficit policies) are exacerbated (De Grauwe, 1992). One would therefore expect policymakers to devote a great deal of attention to the coordination of fiscal policies in a monetary union. Until very recently this was not always the case in the WAEMU and its predecessors.
1. Fiscal externalities: The trade-distorting effects
Experience in Europe and Southeast Asia shows that regional trade increases welfare. 1/ The WAEMU has the potential to follow this example because the monetary union represents a unique chance for strong trade integration among the countries of the region; after all, there is no exchange rate risk and financial transaction costs are minimized. So far, however, the WAEMU has not been able to fully reap the benefits of having one currency. Figure 1 shows that trade within the zone was, until recently, rather weak, amounting to just slightly over 10 percent of total trade. Only landlocked countries like Mali, Burkina Faso, and Niger have traded more heavily with their WAEMU partner countries.
There are several reasons for this rather disappointing trade performance within the old WAMU: (i) production and consumption patterns are rather similar in the sub-Saharan region, and countries have not (yet) developed fields of comparative advantage in trade among themselves; (ii) until recently, the overvaluation of the CFA franc made imports from outside the zone relatively cheap; and (iii) various fiscal policies are impeding trade.
It is difficult to estimate the empirical relevance of each of these factors. Only one of them--the impact of the overvaluation of the CFA franc on regional growth--has been the subject of some studies (Clément, 1994; Nashashibi and Bazzoni, 1993; Kouwenaar, 1991), and its relevance is hardly disputed. The effects of diverging fiscal policies on trade within the WAEMU have not yet been empirically tested. Quantifying such distortions at the microeconomic level would require a general equilibrium analysis; data and time constraints make such an approach impossible. However, some simple theoretical considerations may help to eludicate whether fiscal policies played a role in reducing trade and welfare.
Numerous fiscal policy instruments, which in principle can have trade-distorting effects, can be identified in WAEMU member countries:
protective use of domestic indirect taxation;
import quotas and outright import bans;
subsidies and tax breaks for import-competing industries;
technical, safety, health, and other regulations; and
export taxes and subsidies.
While all these protectionist practices may have some significance in the WAEMU, we concentrate here only on the first two, because they are presumably the most important ones and are also explicitly addressed in the present harmonization strategy.
a. Customs duties
As is well known from traditional trade theory, a tariff represents a social cost to society. Administrative and resource displacement costs and the loss of a consumer surplus generally exceed the gain for domestic producers and the government (through customs revenues). By inducing countries to import less than their optimal level, the volume of trade is reduced compared with the free-trade equilibrium. Thus, all trading partners together are prevented from fully reaping the mutual advantages of trade.
Do customs have a detrimental effect on the countries in question? There are several indications that they do. First, customs duties are the most important source of revenue in the WAEMU (Nashashibi and Bazzoni, 1993) and their rate levels are high. Second, rates are extremely differentiated, causing numerous distortions at the microeconomic level.
A recent study conducted by a French technical assistance team from the Inspection Générale des Finances (IGF, 1994) found that customs duty laws are far from uniform and simple, because different rates and exemption rules are applied to almost every commodity and country of origin. Moreover, the rate levels and structures differ markedly between the countries of the zone. This can be seen from Table 1, which presents average effective tariff rates (that is, the ratio of receipts and taxable imports) for three commodity categories. Such differentiated tariff rates distort the optimal allocation of international resources because they drive a wedge between the respective marginal rates of substitution of (i) domestic and imported goods, (ii) goods taxed at low rates and those taxed at high rates, and (iii) imported goods from different countries. Moving the rate structure toward more simplicity and uniformity could not only facilitate tax administration, but also mitigate the distortions mentioned above. In their Fund-supported programs following the devaluation, some countries abolished special tariffs on selected goods. This is a move in the right direction. However, most tariff systems remain hybrid and highly trade-distortive, and further action seems warranted.
b. Domestic indirect taxation
Domestic indirect taxes can have distortive effects on trade just as tariffs do. Since the value-added tax (VAT) and excise taxes represent the second-largest source of revenue in the countries in question, their effect will not be negligible. The question of whether a given tax rate structure is trade-neutral or not depends on the international taxation principle in place, i.e., whether the country in question applies the destination principle (DP) or the origin principle (OP). The two’ tax principles have very different implications. If the DP fully applies, countries are in principle free to choose rate levels of VAT and excises; trade distortions arise only if the definition of tax categories and/or the rate structure differ between countries. If, in contrast, the OP applies, indirect tax rates need to be fully harmonized in order to avoid trade distortions.
Figure 1.Trade Among WAEMU Partner Countries
(In percent of total trade; 1992)
Citation: 1995, 25; 10.5089/9781451844207.001.A001
Source: IMF Economic Information System
|Category I - Bask necessities (food, medical equipment, books)|
|Category II - Intermediate inputs (machines, trucks, fertilizer)|
|Category III - All other commodities (especially final consumption goods)|
includes droits de douane, droits fiscales, and import VAT; petrol taxes are excluded.
does not include VAT; total including VAT (173 percent).
includes droits de douane, droits fiscales, and import VAT; petrol taxes are excluded.
does not include VAT; total including VAT (173 percent).
Legally, WAEMU countries apply the DP, using border tax adjustments (BTAs) to correct for differences in indirect tax rates. This means that exports are tax-exempt, while imports and comparable domestic goods are taxed at the same rate. In practice, however, customs administration is weak and many commodities pass the frontiers without the BTA procedure, so that the actual situation may lie somewhere between DP and OP. It therefore seems useful to examine the present system of VAT and excises in the WAEMU, assuming, alternatively, that the DP and the OP apply.
If the OP applies, VAT and excises act like taxes levied on the production level, i.e., the consumer’s choice is determined by the tax rate in the country of production. International rate differences drive a wedge between the ratio of prices of domestic and imported goods. This distorts consumer choices, because the decision to buy commodities will not be based solely on comparative advantage, but on the tax rates prevailing in the country of origin. Table 2 presents VAT rates in the WAEMU countries. One sees that for the reduced category, rates range from as low as 5 percent (in Togo) to 11.1 percent (in Côte d’Ivoire); for the normal category, the rates range from 14 percent (in Togo) to 20 percent (in Côte d’Ivoire and Senegal). Hence, if one assumes that the OP prevails in the WAEMU, the present rate differences between member countries are highly distortive.
If, on the other hand, the DP does apply, the present tax system may, on the whole, appear much less distortive. When the BTA mechanism is in fact operational, it makes sure that consumers pay the tax rates of their country of residence, no matter where the respective commodity originates. International resource allocation may, however, be distorted by the rate structure within a country. It can be shown that an indirect tax system is only trade neutral if certain conditions regarding the ratios (not necessarily the levels) of tax rates are met. 1/ For example, a two-rate VAT system is trade neutral if
|Nominal VAT rates (in percent)|
|Benin||Burkina Faso||Côte d’Ivoire||Mali||Niger||Seneg|
|VAT rate structure (ratio to normal VAT rate)|
|Benin||Burkina Faso||Côte d’Ivoire||Mali||Niger||Seneg|
|VAT rate structure (1 + VAT rate/1 + normal VAT rate)|
|Benin||Burkina Faso||Côte d’Ivoire||Mali||Niger||Seneg|
It is difficult to perform this same exercise for the excise tax rates presented in Table 3 because some countries levy these taxes on an ad valorem basis and some do not. However, Table 3 illustrates that for some countries the above condition is probably even less fulfilled for excises; this is particularly the case if there are a large number of excise taxes with different rates (as, for example, in Burkina Faso and Côte d’Ivoire). Moreover, Tables 2 and 3 report only nominal tax rates. Tax exemptions, which differ widely from country to country, are an additional source of distortions. It should be noted, though, that most countries have made an effort lately to move toward a smaller number of VAT and excise rates.
Additional distortions are introduced by the fact that in the WAEMU the DP applies to some products, while the OP (i.e., no BTA procedure) applies to others. This situation can be optimal only if all countries levy the same uniform indirect tax rate on all commodities. Quite obviously, this is not the case. Also, a mix of DP and OP leads to uncontrollable tax revenue shifts between countries’ and it creates distrust and uncertainty, undermining the solidarity between WAEMU member countries.
We conclude that the present system of indirect taxation in WAEMU is highly distortive, no matter which international taxation principle one assumes. To simplify matters, we will in the following discussion assume that the DP not only prevails in law, but also in practice. From the exposition above, however, it should be clear how important it is to enforce one international taxation principle. For various reasons, the DP seems the better choice, as is discussed in Section III.2.
2. Fiscal externalities: The beggar-thy-neighbor effects
It is fairly well established that diverging fiscal policies cause trade distortions of the kind described above. However, this effect is rather indirect and abstract, and policymakers usually do not devote too much attention to it. There is a much more obvious threat to solidarity in the WAEMU: the use of fiscal instruments to conduct beggar-thy-neighbor policies. This means that one individual WAEMU country achieves a welfare gain at the cost of one or more union members. We will argue that such practices are rather common; their removal represents an additional argument in favor of more policy coordination in the zone. To simplify matters, we will concentrate only on the most important fiscal externalities of this kind, namely customs duties on imports, indirect taxes, and the size of the budget deficit. Wage and price controls, export taxes, quotas, and other quantitative import restrictions are omitted, although they may play an important role.
(in percent, unless indicated otherwise)
|Benin||Burkina Faso||Côte d’Ivoire||Mali||Niger|
|Wine (ordinary)||16.0||25.0||77 CFAF/liter||20.0||25.0-52.0|
|Soft drinks||16.0||25.0||25 CFA/liter||10.0||…|
a. Customs duties
The above-described trade distortions occur through the use of import duties, regardless of whether the countries in question have an influence on the price of traded goods or not. However, the theory of optimal tariffs indicates that a country may improve its welfare, if it has some influence on its terms of trade (i.e., if it is a “large” country, in the sense that it has some monopsony power vis-à-vis its trading partners). In this case it may gain at the cost of its trading partners. Therefore, the first question to ask is whether WAEMU countries have an influence on the price of goods that are traded within the zone. Only in this case can there be a beggar-thy-neighbor effect through the use of tariffs.
Information on the nature of intra-WAEMU trade is scarce. To our knowledge, no comprehensive study of the volume and the nature of goods traded between the seven countries exists. However, the recent IGF report (1994) provides selected information on some countries. From this it seems that cement, fertilizers, pharmaceuticals, and salt are mainly provided by the region’s two main exporters, Senegal and Côte d’Ivoire, and face little competition from outside the zone. Beef demand in the WAEMU is largely covered by Mali, Niger, and Burkina Faso. Other regionally traded commodities are fish, fruits, and vegetables. Most of these goods will be exposed to little or no competition from outside the region, simply because of a lack of transporting and refrigerating capacities. This has been particularly true since the devaluation, which has made imports from outside the zone much less competitive. 1/ We conclude that for most goods traded at the regional level, importing WAEMU countries have had some power to influence their terms of trade. This does not, however, imply that the optimum tariff argument generally holds. A large share of intra-WAEMU trade presumably passes the borders through informal channels and is not affected by customs duties at all. Therefore, it should be noted that, strictly speaking, the following discussion applies only to formal trade. Although its empirical relevance may be disputable, the optimum tariff argument will, at least to some extent, be valid; i.e., customs duties applied to imports from WAEMU partners will have a detrimental effect on the exporting country’s welfare.
There are various other, perhaps less abstract, ways in which the tariff structure can be used as a protectionist instrument vis-à-vis WAEMU trading partners. Here are some examples:
Hybrid customs systems and the lack of a common external tariff for the zone create potential for arbitrage, the emergence of a parallel market, and smuggling. Countries with low effective tariff rates attract more imports, and hence revenues. Intermediate inputs, for example, have been imported into the zone via Dakar rather than via Abidjan, because the effective tariff rate in Senegal is almost 50 percent lower. Another example may be even more illustrative: according to a recent study by Philip Morris EFTA (1994), Mali lost CFAF 1.3 billion in customs receipts for 1993 because of illegal cigarette imports; the incentive for smuggling is particularly strong because taxes and customs duties in Mali are almost four times higher than in WAEMU partner country Togo. 1/
Some countries have very low effective tariff rates on imports of intermediate goods. This may result in a comparative advantage vis-à-vis their WAEMU trading partners, because their industries face a lower tariff burden on their inputs and hence are more competitive. Fertilizers, for example, are virtually untaxed in Niger, Togo, Benin, Senegal, and Côte d’Ivoire, but are taxed at the rate of 3.3 percent in Burkina Faso and 5.1 percent in Mali.
Some countries treat imports from WAEMU countries even less favorably than imports from the rest of the world. This is the case for basic necessities in Mali and Senegal, for example, where the effective tariff rate on WAEMU imports is higher than for other countries’ imports. This is particularly remarkable because, until recently, the countries of the CEAO imposed a preferential tariff, the taxe de coopération régionale (TCR), among themselves. The TCR replaced all regular customs duties and represented about half of the levy on similar imports from other origins. Table 1 suggests that the preferential treatment of WAEMU imports was rather weak and never amounted to the envisaged 50 percent reduction of the tariff burden. 2/
There is also a noticeable country-specific discrimination among WAEMU countries for certain intraregional imports. This can be exemplified in the case of Mali, for which the IGF’s study (1994) provides more detailed data. In Category I the effective tariff rates on WAEMU trade range from 11.9 percent for imports from Togo to 23.9 percent for imports from Senegal. If one looks at specific commodities, protectionist use of the tariff system becomes even more evident. Insecticides produced in Côte d’Ivoire, for example, are subject to an effective tariff rate of 17.1 percent, while insecticides originating in Senegal face an effective tariff rate of only 4.2 percent; the same is true for fertilizers.
We conclude that discrimination against intraregional imports from WAEMU countries was (and presumably still is) quite common. The CEAO and other multilateral and bilateral trade agreements were not very successful in preventing such practices.
b. Domestic indirect taxation
Differentiated consumption taxes can be used as a second-best substitute for tariffs. The underlying argument is analogous to the one developed for the optimum tariff, i.e., it holds only if a country has an influence on its terms of trade vis-à-vis its trading partner. As we have seen above, this condition is likely to hold for most intra-WAEMU trade. Consider then, for example, a (hypothetical) country that produces beer but no wine. A relatively higher specific tax on wine will lead to a shift of demand away from wine, changing the relative prices of exportables and importables and eventually the terms of trade. Welfare will increase at the expense of the wine exporting country.
Even a VAT levied according to the DP can be used to pursue a beggar-thy-netghbor policy. A country applying its reduced VAT rate to commodities that are mainly produced domestically realizes a terms of trade gain vis-à-vis its trading partners. This theoretical finding basically holds in practice in the WAEMU. However, systematic evidence of such practices is difficult to find because this would require detailed information on bilateral trade flows and domestic consumption and production patterns. Some examples will suffice:
Protectionist use of specific taxes is evident, if the tax code explicitly differentiates between imported and domestic products. For some products the import equalization tax rate is higher than the tax rate on comparable domestic goods; the difference between the two then becomes a covert import duty. Burkina Faso, for example, levied in 1992 a specific tax of CFAF 65 per liter on domestic beer and CFAF 500-5,000 per liter on imported beer. The impôt spécial sur certains produits (ISCP) in Mali is applied to imported cigarettes (many of which come from neighboring WAEMU countries) at a rate of 40 percent, while the ISCP for locally produced cigarettes is only 20 percent.
Basic necessities, which are mainly produced domestically, tend to be subject to the reduced rate of VAT or are tax exempt. Manufactured goods, which have a higher imported share, are often subject to the normal or “luxury” rate of the VAT. Mali, for example, taxes pharmaceuticals at a VAT rate of 10 percent. It does not have a significant production of its own and most of its pharmaceuticals are imported from Senegal. Hence, there is a terms of trade effect because consumers are induced to shift their demand to traditional (nontaxed) pharmaceuticals.
As can be seen from Table 3, Burkina Faso, Mali, and Senegal have an excise tax on kola seeds, the sole exporter of which is Côte d’Ivoire. Therefore, consumption of addictive substances in these countries is induced to shift toward beer or cigarettes, which are domestically produced. Thus, the excise tax structure is used to improve the terms of trade. The same holds true for beer versus wine taxes in Côte d’Ivoire and Togo; by applying higher rates on wine (which is imported) than on beer (which is domestically produced), these countries attempt to influence the consumption pattern of alcoholic beverages in their favor.
c. Expansionary fiscal policies
During the last few years Côte d’Ivoire pursued a much more expansionary fiscal policy than the other zone members did. In 1993, for example, the fiscal deficit (excluding official grants) was 16.4 percent of GDP in Côte d’Ivoire, while the WAMU average was 8.4 percent of GDP. In part these deficits were financed through strong borrowing from the BCEAO. Concerns about such policies may have induced policymakers to include zone-wide deficit benchmarks in the harmonization strategy. Here we address the question of whether excessive borrowing from the BCEAO by some countries implies negative fiscal externalities for the other WAEMU countries. There are several channels through which this could be the case. We present the theoretical possibilities 1/ and discuss their relevance for the WAMU in the past.
(1) The effect on capital markets and domestic credit
If there were a perfect capital market, excessive deficits in one country would lead to an increase of the interest rate in the entire monetary union; 2/ investment would diminish, resulting in lower growth rates. In the WAMU this was not the case; first, because interest rates were basically determined exogenously by the interest rate level in France, to which the zone was connected by a fixed exchange rate, full capital mobility, and foreign exchange convertibility; and second, because government bonds were not traded in a fully developed capital market.
(2) The effect on the price level
Excess credit to the government will theoretically lead to an increase of domestic credit and--ceteris paribus--to an uncontrolled expansion of money supply (accompanied by a loss of foreign reserves). The resulting inflationary pressures would then eventually spread to all members of the monetary union. As can be seen in Figure 2, credit to the economy declined as credit to the Government increased, thus preventing an expansion of the zone’s overall money supply during the past five years. There is some evidence, however, that money supply induced inflation did play a role in the mid-1980s, when domestic credit and inflation rates in the WAMU increased along with a higher share of government credit. There is a further effect on the zone’s price level: Government expenditure in the deficit country is generally higher than elsewhere, driving up prices first in the country itself and then--through inflated export prices--in the zone as a whole. The significance of this effect depends again on the amount of trade within the zone.
The link between diverging fiscal policies and inflation in WAEMU members was recently examined in a paper by Boccara and Devarajan (1993). In the short run the authors observe significant differences in inflation rates between countries; these are partly explained by diverging spending levels. In the long run, however, Boccara and Devarajan (1993, p. 34) conclude that “… if some countries have expansionary fiscal policy (and thus inflation) the others must take a more contractionary stance.” Such forced policy reaction can be interpreted as a fiscal externality.
(3) The effect on the foreign reserve position of the zone
Excessive government deficits usually have a negative effect on the foreign reserve position. Since BCEAO countries maintain a common account of their foreign reserves at the French Treasury (the operations account), this may affect all member countries. In the last few years, large countries with large deficits accumulated significant liabilities in the operations account. However, there was no foreign exchange bottleneck for these countries because smaller countries made up for these liabilities by a positive foreign reserve position and convertiblity was guaranteed by France.
(4) The effect on the profits of the BCEAO
Without the excessive deficit spending of some countries, the profits of the BCEAO and hence its dividends accruing to other member countries would have been larger, for two reasons. First, assets in the operations account (see above) could have been higher; these assets would have earned interest for the BCEAO as a whole. Second, countries like Côte d’Ivoire, by exceeding the avances statutaires (limits on BCEAO credit to the government), negatively affected the BCEAO’s profit. According to the BCEAO’s regulations, C6te d’Ivoire would have been obliged to pay penalty interest on these credits. Instead, it accumulated arrears, enlarging the (book value) credit to the Government even further. Both these sources of forgone interest earnings of the BCEAO represented an opportunity cost and hence a fiscal externality to all WAMU member countries.
Figure 2.Composition of BCEAO Domestic Credit
(5) The effect on the terms of trade vis-à-vis other zone members
An expansion of domestic demand by deficit spending will indirectly shift the terms of trade vis-à-vis other zone members in the country’s favor, if there is a home goods preference. Consumers generally spend a larger portion of their additional income on domestic goods; this leads to a shift in relative prices similar to the one described above for indirect taxes. This effect cannot be quantified without use of a computable general equilibrium model. The fiscal externality will be stronger for countries that depend heavily on trade with other WAEMU countries, that is, landlocked countries like Mali, Burkina Faso, and Niger will be more heavily affected.
It seems that the first three effects did not become effective (or at least not evident) in the WAMU. Why, then, did the observed contradictions between fiscal and monetary policies in the 1980s not cause larger disruptions? A possible explanation is that the overall balance was maintained through a number of short-term measures such as
debt rescheduling, and
the accumulation of external and internal arrears.
Such means of financing are, of course, not available indefinitely. The hidden fiscal pressures may have persisted and contributed to the need for a devaluation of the CFA franc. Since the zone had one common currency, this parity change had to be the same for all member countries. The rate of devaluation (50 percent in foreign currency terms) was tailored to the requirements of the countries with the highest losses in competitiveness (Côte d’Ivoire in the WAEMU, Cameroon in the Central African Monetary Union), which also happened to be the main deficit countries. Thus, it “… did not meet all the needs of every economy” (Touré, 1994). Alternatively one could say that the devaluation itself represented a fiscal externality of deficit countries vis-à-vis zone countries with more prudent policies.
III. Fiscal Policy Coordination after the WAEMU Treaty
1. The need for fiscal policy coordination after the devaluation
The analysis of the past has identified three major areas in which WAEMU countries harm other member countries and trade in the zone as a whole: (i) customs duties, (ii) domestic indirect taxation, and (iii) budget policies. These fiscal externalities are due to diverging fiscal policies and could be prevented if there were more cooperation between member countries.
The devaluation of the CFA franc in January 1994 may well have been the biggest macroeconomic shock the WAMU had faced since its creation, 35 years ago. There are three major reasons why this decisive adjustment measure makes a close coordination of tax, tariff, and budget policies more urgent than ever:
(i) Import substitution following the parity change will lead to more trade between member countries. There are already indications that intraregional trade of commodities like machines and equipments has picked up after the devaluation. Such goods tend to be traded through formal rather than informal channels, and they are therefore affected by taxes and tariffs. The more such trade links evolve, the stronger the fiscal externalities described above will be (this is particularly true for the terms of trade effects at the microeconomic level).
(ii) Solidarity in the WAEMU will be jeopardized, as fiscal externalities become much more visible. The reason for this is that the above-described means of exceptional financing are no longer available. In connection with the devaluation, most CFA countries have negotiated programs supported by arrangements with the IMF and World Bank. This financing might partially replace the bilateral external assistance these countries used to receive from France. Adjustment programs typically emphasize fiscal discipline, limit exceptional financing, and require repayment of external and internal arrears.
(iii) In the field of tax and customs harmonization there is now more room for maneuver. The devaluation has an effect like that of a uniform external tariff rate hike of 100 percent, markedly increasing the competitiveness of the zone as a whole. Taxes at the disposition of governments can now be used more as an instrument to raise revenues than to protect domestic production.
Moreover, the chance of implementing a coherent harmonization strategy is good, because the simultaneous existence of IMF programs In all WAEMU countries offers a unique window of opportunity to coordinate fiscal policies in the zone.
2. An evaluation of the recent steps toward fiscal coordination
In February 1990 the Ministers of Finance of the WAMU and the BCEAO initiated a stronger coordination of fiscal policies. A Steering Committee, including member countries and principal donors, was established; its recommendations were adopted in principle by the Council of Ministers in 1992. The WAEMU treaty of January 10, 1994 marks the beginning of a process to implement the Steering Committee’s guidelines in the fields of indirect taxation, customs duties and budgetary performance. However, numerous details remain to be resolved.
a. Customs duties
Proposals for a unified external tariff and a preferential tariff for intra-zone trade are under preparation. Studies were carried out by the French Inspection Generale de Finances (IGF), and some preliminary proposals were made at the Steering Committee Meeting in Dakar in May 1994, taking into account the effects of the devaluation. In its report transmitted to World Bank and IMF it concludes:
The introduction of a preferential tariff rate, which had been proposed earlier, should be postponed because (i) member countries with relatively strong trade links to the rest of the zone would otherwise lose revenues and (ii) the devaluation already has an effect like that of the imposition of a common external tariff of 100 percent.
A “reasonable objective” for common external tax rates should be 5 percent for basic necessities (Category I), 20 percent for intermediate goods (Category II), and 35 percent for all other goods (Category III) because this (i) corresponds to a tariff rate reduction of 5 percent in the main member countries, (ii) ensures the same nominal revenues as before the devaluation, (iii) contains price increases for basic necessities, and (iv) offers some room for maneuver with regard to further rate cuts.
On the whole, there is merit in the IGF’s main policy conclusions. They advocate simplicity and transparency, making customs administration easier and less prone to induce fraud. Also, the proposed rate structure would remedy some of the fiscal externalities described above, for example, gaining a competitive advantage by taxing intermediate goods at a lower rate than there is in WAEMU partner countries. However:
Some of the assumptions underlying the IGF’s simulation method are disputable. For example, the IGF model does not assume any change in intra-WAEMU trade after the devaluation; the underlying import elasticities are not consistent with elasticities used elsewhere; and it may not be possible to cut exemptions across the board by 10 percent.
The proposed rates would only preserve nominal tax revenues. This may not suffice in a post-devaluation situation in which prices and hence government expenditures are expected to increase by up to 40 percent.
More attention should be devoted to exemptions. As shown above, it is the effective tariff rates on imports from different member countries that make the tariff system so hybrid; simply harmonizing statutory rates and reducing exemptions across the board will not remedy this highly distortionary rate differentiation.
Before a final proposal regarding customs rates is formulated, revenue estimates should be thoroughly revised. In particular, it remains to be investigated whether, for the time being, the idea of a preferential tariff should be abandoned. At first sight this idea appears attractive. After all, the devaluation temporarily exerts an effect like that of a reduction of the internal zone tariff rate. Moreover, revenue considerations as well as simplicity seem to favor a straightforward three-rate system for all imports, no matter where they originate.
b. Domestic indirect taxation
Harmonization efforts in the field of domestic indirect taxation are being coordinated by the Fiscal Affairs Department of the IMF. The most recent proposals were presented at the Steering Committee Meeting in Dakar in May 1994 (IMF, 1994) and in principle accepted by the BCEAO. The revenue estimates included in the respective study are based on pre-devaluation data, and hence can serve only as a very preliminary yardstick. In the following discussion, we will therefore concentrate on the allocative efficiency aspects of the present harmonization strategy.
(1) The current harmonization strategy
The main proposals of the current harmonization strategy adopted by the authorities in coordination with the IMF are:
Value-added taxes: Reduce the number of VAT rates to two (15-20 percent, and 5-10 percent); allow zero-rating for exports only; unify the list of tax categories and exemptions (which are supposed to be kept to a minimum). In the long run a single VAT rate is envisaged.
Petroleum taxation: Abolish the present hybrid taxation schemes (often based on valeurs mercuriales); all petroleum products should be subject to VAT, regular customs duties, and excises.
Excises: Apply special rates to a list of 13 selected commodities only; rates may vary between 0 and 100 percent.
(2) Destination versus origin principle
Any proposal for an indirect tax harmonization strategy in the WAEMU faces a basic dilemma: On the one hand, fiscal externalities of the kind described above are to be minimized. On the other hand, fiscal sovereignty is to be preserved as much as possible--that is, countries should have enough flexibility to choose tax rates that meet their individual revenue needs.
In view of these seemingly contradictory goals, what should the optimal harmonization strategy be? The basic choice lies between the two international taxation principles. It is of the utmost importance for the WAEMU to fully commit itself to one of these principles, once this basic decision has been made. The present mixture between OP and DP is unsatisfactory, because it creates substantial fiscal externalities, no matter what tax structure and rates are chosen. Therefore, there seem to be two alternatives:
(i) Move to the OP altogether. This implies abolishing the cumbersome BTA procedure. Firms in one country would be allowed to deduct the VAT paid on inputs bought in another member country; excises would be levied in the country of origin and would now resemble production taxes. If the WAEMU decides to adopt the OP, VAT and excise tax rates will need to be fully harmonized, because otherwise fiscal externalities would prevail and numerous possibilities for arbitrage would arise. Moreover, a compensation scheme for revenue shifts between countries (similar to the VAT clearing system envisaged for the European Union after 1996) would have to be put in place. Countries would have to rely on direct taxes in order to meet their specific revenue needs (assuming factors are not perfectly mobile between countries).
(ii) Fully enforce the DP. This means improving customs administration and ensuring that the BTA procedure applies to all flows of goods and services. The advantage is that member countries would have some freedom to choose their VAT and excise duty rates. As described earlier, trade distortions would be minimized if the rate structure of VAT and (substitutable) excises were identical, while rate levels do not have to be the same. Note, however, that under such a scheme there is still room for beggar-thy-neighbor policies by taxing importables more heavily than exportables. These can only be avoided if countries choose a uniform VAT and excise rate.
The present harmonization strategy assumes that the DP will be maintained, at least in the medium term. 1/ This seems appropriate, since abolishing the BTA procedure and switching to the OP within the WAEMU may seem tempting, but the net gain is likely to be negative for three reasons. First, a complicated and costly clearing mechanism would have to be put in place, because countries will not be willing to accept revenue losses. Second, the reduction of administrative costs would be marginal--at least as long as customs duties within the zone are not abolished as well. Third, leaving countries with direct taxes as their only independent tax instrument is unrealistic, because these taxes constitute only about 20 percent of revenues. Therefore, the OP is not an attractive option, not even in the long run.
If the DP is to be maintained, enforcing an efficient BTA procedure is essential. Strengthening customs administrations may be costly, but so are the fiscal externalities caused by a mixture between DP and OP. A well-functioning DP that minimizes fraud, arbitrage, and revenue shifts will strengthen the solidarity in the WAEMU. Most importantly, the DP would leave countries with at least some freedom to choose their revenue level.
(3) The optimal VAT rate structure
The WAEMU has opted to apply the DP to its international trade transactions. Given this choice, do the proposed rate bands minimize fiscal externalities?
The simplified two-rate VAT system undoubtedly represents a significant move toward trade neutrality. But it goes a step too far. As shown above, under the DP, only rate structures, not necessarily rate levels, have to be the same across countries. Therefore, the harmonization strategy has to be very explicit about the desired structure of VAT, but not on the rate level. In this respect, however, the proposed rate bands are not very specific. Permitting two rates ranging from 5.3 percent to 10 percent and 15 percent to 20 percent allows the ratios between reduced and normal rates to vary as much as from 5.3/20 - 0.27 to 10/15 -0.66. If fiscal externalities are to be minimized, a better strategy would be to require that the ratio
be at least roughly the same in all countries.
Even more desirable (at least from the efficiency point of view) than the presently proposed two-rate system would be a single-rate system. As shown above, this would not only be easy to administer, but it would also minimize the potential for fiscal externalities. 1/ In fact, Benin, Burkina Faso, and Niger have already introduced single rate systems. It therefore seems only logical to place the introduction of a single-rate VAT on the agenda of VAT harmonization. If only some, but not all, countries decide to switch to a single rate system, fiscal externalities will persist.
In this context the importance of carefully defining VAT-exempt categories should be emphasized. Equity considerations and the question of zero-rating versus VAT exemptions are left aside in this study, which deals fiscal externalities only. It should be noted, however, that the above-mentioned rules about a symmetric rate structure between countries also apply to exemptions. Therefore, exactly the same commodities need to be VAT exempt in all WAEMU countries, if fiscal externalities are to be minimized.
(4) Excise Tax Rates
The current proposal allows countries to freely choose their excise tax levels. This is supposed to leave governments with a tax instrument to meet their individual revenue needs. There are two problems with large excise rate differences within and between countries: (i) such differences induce fraud, smuggling, and arbitrage, which lead to fiscal externalities and revenue shifts between countries; and (ii) even if border controls fully work, rate differentials within a country can be used to pursue beggar-thy-neighbor policies.
In addition, the proposed list of 13 excisable products is rather long and the economic justification for taxing these products is sometimes dubious (why, for example, should the construction industry be burdened by an excise tax on cement?). Moreover, fiscal externalities and the potential for fraud will be minimized only if the number of excise rates is small and their rates are similar. A positive example in this respect is Benin, where all beverages (whether alcoholic or not) are taxed at the same rate of 16 percent. A negative example in the WAEMU is Côte d’Ivoire, where at least ten different rates on beverages are applied. Note also that the transparency of excise taxation in the region could be greatly enhanced if all countries applied ad valorem rather than specific rates. 1/ This would require Côte d’Ivoire and Togo to switch from quantity- to value-based excise tax rates.
In sum, the proposed indirect tax harmonization strategy is too strict on VAT rates and too soft on excise rates. Three main elements should be placed on the agenda:
(i) strictly enforce the DP by strengthening the BTA procedure;
(ii) introduce uniform VAT rates in each country, but let the countries choose their rate level fairly freely;
(iii) reduce the number of excise taxes to a few “sin” and luxury taxes (perhaps five); define a narrower band for rates (perhaps between 10 percent and 30 percent); and introduce ad valorem excise tax rates in all member countries.
c. Macroeconomic indicators and surveillance
The first step toward multilateral surveillance of budgetary policies is a harmonization of the method of recording government financial operations. The IMF and the French Ministry of Cooperation took a lead in designing a unified presentation of federal budgets in WAEMU countries (the so-called Tableau des opérations financières de l’Etat. TOFE). A TOFE has been developed that attempts to encompass the same definitions for all countries. However, many problems remain to be resolved. At the Steering Committee Meeting in May 1994, little progress was reported on the side of national budgeting authorities. The existence of programs in all WAEMU countries represents an opportunity for putting identical government accounting plans into place. This is clearly a prerequisite for effective surveillance.
Developing and enforcing binding rules for macroeconomic performance seems to be the most difficult field of fiscal harmonization and progress has been slow. There seems to be a consensus that some rules have to be put into place to ensure budgetary convergence and prevent fiscal externalities of the kind described above (Section II.2.b.). Proposals have been made by the BCEAO, the French Ministry of Cooperation, the European Union, and the IMF.
It is difficult to fully assess the current status of the discussions. The most recent attempt to define and quantify a number of indicators of budgetary performance seems to be a note by the BCEAO, dated September 17, 1993. This list, which was not further questioned at the last Steering Committee Meeting in May 1994, includes the following five indicators, or performance criteria:
(i) government wage bill (less than 50 percent of fiscal revenues),
(ii) domestic government arrears (elimination),
(iii) external government arrears (elimination),
(iv) public investment financed by domestic resources (more than 20 percent of fiscal revenues); and
(v) primary balance (more than 15 percent of fiscal revenues).
For 1993, none of the seven WAEMU countries meets all of these criteria. Criterion (i) is fulfilled only by Mali, criteria (ii), (iii) and (iv) are not fulfilled by any country, and criterion (v) is fulfilled only by Benin. A recent paper prepared by the IMF’s African and Policy Development Review Departments (1994) reviews Fund-supported programs in six WAEMU countries in light of the proposed convergence indicators. It concludes that all countries except Niger could meet the convergence criteria by 1996.
The choice of these five performance criteria provokes some additional thought. Clearly, the number of benchmarks should be confined to a very few, but meaningful, budgetary indicators. The three criteria relating to the primary balance and the elimination of arrears seem good candidates, because they directly address the fiscal deficit. The rationale for the remaining two other criteria (i) and (iv) seems less clear, at least if their primary aim is to avoid fiscal externalities and promote solidarity in the union. These indicators seem to target domestic aims rather than harmonization within the WAEMU. Minimum public investment and savings levels are intended to promote growth and ensure intergenerational equity; and a limit on the government wage bill is only indirectly related to deficit spending.
An additional problem is enforcement. The experience with the avances statutaires suggests that binding rules alone will not suffice. A sanction mechanism is discussed in a note by the EC Commission (1993). It suggests a “carrot and stick approach,” including rewards (budgetary aid financed by donor countries) for good performers. The outcome of the discussions on enforcement is not yet clear.
IV. The Path Ahead
The general harmonization strategy adopted by the BCEAO is commendable. It will mitigate fiscal externalities as identified in the first part of the paper. Moreover, the peer pressure in the context of a common effort will help to address domestic fiscal policy problems such as: simplification of the tax and tariff system, and broadening of the revenue base; and fiscal deficits will be contained.
However, some details of the present harmonization strategy need to be reviewed, especially in light of the recent devaluation:
(i) for customs duties, rate levels (but not necessarily rate structures) will have to be adapted to the post-devaluation macroeconomic framework;
(ii) for the VAT, the harmonization strategy should be less specific with respect to rate levels, but more specific with respect to rate structures;
(iii) for excises, more attention should be paid to similar rate structures, and the range of permissive rates should be confined;
(iv) for indirect taxation in general, enforcing the destination principle (DP) should become a central issue; and
(v) for surveillance indicators, some more thought should be given to the usefulness on benchmarks for public wages and investment.
In the monetary field, the CFA franc zone benefited greatly from the discipline imposed by the monetary union. The WAEMU’s envisaged fiscal harmonization will give the zone the chance to fully reap the benefits of its unique monetary arrangements by extending the same discipline to its fiscal policies.
BoccaraBrownie and DevarajanShantayanan“Determinants of Inflation among Franc Zone Countries in Africa,”Policy Research Working Papers. WPS 1197The World BankSeptember1993.
ClémentJean“Striving for Stability: CFA Franc Realignment”Finance and Development 31/2 (June1994) pp. 10–14.
De GrauwePaulThe Economics of Monetary Integration. Oxford: Oxford University Press1992.
European CommissionLe Dispositif de Surveillance Multilatérale de l’UMOA. Note prepared for the BCEAOJanuary1993.
FehrHansRosenbergChristoph and WiegardWolfgangValue-added taxation in the EC after 1992European Economic Review 37 (1993) pp. 1483–1506.
Inspection Générale des FinancesRapport d’Enouete sur l’Harmonisation des Fiscalités Douanières dans l’Union Economique et Monétaire Ouest Africaine,March1994.
International Monetary Fund Policy Development and Review and African DepartmentsCommon Policy Issues of the CFA Franc Zone CountriesJune1994.
International Monetary Fund Fiscal Affairs DepartmentUnion Economique et Monetaire Ouest Africaine: Implications Budgetaires de l’Harmonisation des Systemes Fiscaux IndirectsMay1994.
KouwenaarArendPrice and Volume Effects of a Devaluation in Developing Countries.IMF Working Paper WP/91/1301991.
NashashibiKarim and BazzoniStefania“Exchange Rate Strategies and Fiscal Performance in Sub-Saharan Africa,” Staff Papers.International Monetary FundVol. 41 (No. 1March1994) pp. 76–122.
PhilipMorris EFTAAnalyse du Marché des Cigarettes au Mali: Proposition de Réformes Fiscales et Commerciales,unpublished studyLausanneApril1994.
TouréMamoudou“The New Structural Adjustment Strategy in the CFA Franc Zone Aims at Stimulating Growth and Reducing Poverty,”Interview in IMF Survey. March141994 p. 7.
TouréMamoudou“30 Ans d’Union Monétaire: Bilan Critique,”in: BCEAO L’Union Monétaire Ouest-Africaine: 30 ans de coopération monétaire. bilan et perspectiveVolume 11992 pp. 9–21.
I would like to thank Mr. François, Mr. Dhonte, Mr. Tahari, Mr. Katz, Mr. Clément, Mr. Cossé, and all the economists of the Sahel Division I for helpful discussions and comments. All errors are mine.
The West African Monetary Union (WAMU) was founded in 1960; it links seven countries (all former French colonies) by a single currency, the CFA franc, which is pegged to the French franc. In addition, the West African Economic and Monetary Union (WAEMU) was established on January 10, 1994. Both organizations comprise the same countries, presently Benin, Burkina Faso, Côte d’Ivoire, Mali, Niger, Senegal, and Togo.
The CEAO membership included Burkina Faso, Côte d’Ivoire, Mali, Mauritania, Niger, and Senegal.
This is not necessarily true on a priori grounds. In theory, the net effect of fiscal harmonization may be negative, since regional trade creation may lead to trade diversion for individual countries. Empirical studies using applied general equilibrium models suggest, however, that the net welfare gain will generally be positive. See, for example, Fehr, Rosenberg, Wiegard (1993).
Trade neutrality is defined as a situation in which the marginal conditions for a Pareto-efficient free trade allocation are not violated. For a short formal exposition see Fehr, Rosenberg, and Wiegard (1993), pp. 1494-1497.
A prime example in this context is beef. Until 1994 traditional producers in the region--like Mali, Burkina Faso, or Niger--had difficulty in exporting their livestock to Côte d’Ivoire, because of fierce competition from subsidized beef imports originating in the European Community. Since the devaluation, regional exports have picked up markedly.
This problem is even more pronounced with respect to cigarette imports from Guinea; since this country is not member of the WAEMU, it Is not subject of this study, which concentrates on fiscal externalities within the zone only.
Note, however, that Table 1 reports effective tariff rates, which also take into account the effect of exemptions and fraud. In contrast, the CEAO’s 50 percent rule applies to nominal tariff rates.
An extensive literature on the effects of deficits in monetary unions has evolved around the European Monetary Union. The following exposition draws on some of these writings, notably De Grauwe (1992) and Wyplosz (1991). The need for macroeconomic--i.e., budgetary policy--coordination in the WAEMU is very clearly laid out in Touré (1992).
This does not necessarily mean that interest rates have to be exactly equal in all countries of the monetary union. Bond traders will attach risk premiums to bonds originating in countries with a low credit rating.
In contrast to the IMF’s harmonization strategy, the BCEAO seems to envisage an ultimate switch to the OP as its ultimate goal. See, for example, BCEAO (1993, p. 8).
A fiscal externality remains, because even in a one-rate VAT system there are VAT exemptions, which in principle allow for beggar-thy-neighbor policies of the kind described above. The possibility of using this instrument is limited, however, if exempt categories are defined identically in all countries.
Another common argument in favor of ad valorem rates is that they do not have to be adjusted in the case of inflation.