Trade Negotiations Insights • Volume 9 • Number 6 • July 2010

Fiscal impact of the Economic Partnership Agreement in West Africa

by David Laborde

Discuss this articleShare your views with other visitors, and read what they have to say

Will trade liberalisation bear an adjustment cost for ACP countries? If so, how much and who will pay for it? What are the strategies for reducing or even eliminating these costs? These are all sensitive questions facing ACP and European negotiators in the context of the negotiations of the Economic Partnership Agreements (EPAs).

Qualitatively and quantitatively defining the “net fiscal impact” of an EPA: a complex undertaking

In West Africa - one of the most important regions in demographic and economic terms among the ACP - European and African negotiators decided to inform their discussions with a regional Computable General Equilibrium model led by a joint committee. Indeed, answering the above questions required a common tool to quantify certain key issues in the negotiation, especially the notion of net fiscal impact. The loss of customs revenues has been a major concern for West Africa’s governments, which are faced with tense budgetary situations and economies with an extensive informal sector, presenting a challenge for replacing border taxes with a domestic tax system. The EU committed itself to accompany and support the ACP partners during the liberalisation phase in order to protect against the negative effects resulting from trade reform, which is why it is necessary to assess and determine the forms of this support. Nonetheless, three important points must be borne in mind. First, this is an unprecedented commitment. While the EU has negotiated numerous free trade agreements, until the EPAs it has not committed to pay the adjustment costs of its partners.[i] Second, the very concept of “net fiscal impact” is new, and must be defined before any quantification can be undertaken. Finally, one should note the originality of the approach, which revolves around the notion of partnership between the EU and ECOWAS, as the parameters and hypotheses of the model have been determined jointly by the two parties, in order to avoid politicising the exercise.

Trade liberalisation and its fiscal effects

The fiscal effects of trade liberalisation are numerous. First, the abolition of customs duties leads to a loss of the custom revenues derived from European imports. Moreover, the replacement of imports coming from non-EU countries (e.g. the United States) subject to multilateral customs duty by non-taxed European products will lead to additional revenue losses. At the analytical level, it is important to make a distinction between theoretical revenues (nominal rate multiplied by the value of imports) and the level of tax actually collected. The latter is sometimes significantly lower (less than 60% for some countries), because of tax avoidance, corruption or legal tax exemptions (i.e., sector-based initiatives, tariff suspension, imports by government or international agencies). Moreover, customs duties are not the only type of tax levied at the border: excise duties or value-added tax (VAT), for example, are also applied on imports. If those taxes are not directly affected by liberalisation, their tax base is. For instance, the volume of (and therefore revenue from) imports will increase, but the value of imports, including the value of custom tariffs, which is the basis of calculation for the VAT, can be reduced. In the case of a perfect substitution of taxed local products by imported products, the net impact will be negative. However, if the imports replace the production of the informal sector, which previously evaded the tax system, the impact will be positive. The combination of the different mechanisms will therefore depend on the consumers’ sensitivity to prices and on the efficiency of the initial tax collection.[ii] By changing the growth path at the macroeconomic level, liberalisation will also have an impact on income derived from VAT.

Indirect taxation is not the only channel to be considered. Income tax - both corporate and individual - will also be modified by changes in the levels of profit, wages and employment. Moreover, changes in profitability will also affect public - or quasi-public - companies and have consequences on the consolidated budget of public authorities.

Adoption of a Computable General Equilibrium Model

To be able to take all these effects into account, particularly within the scope of commercial liberalisation spread over 15 to 20 years, requires a model as comprehensive as possible. In this respect the regional Computable General Equilibrium model seems to be the best choice: it can determine the “net” impact on the government’s total tax income, considering all the effects previously discussed.

This definition of the “net” fiscal impact can seem incomplete, however, because it does not address the level of public spending that will also react to trade liberalisation. Nevertheless, the spending aspect entails difficult questions. For example, how will spending on social services and public investments be affected? For this reason, we have decided to focus on the “revenue” side of the equation.

Definition of the analytical framework

Once the net fiscal impact has been defined, an important question remains: what should be the point of comparison? At the beginning of the EPA negotiations, the African negotiators insisted that the old Cotonou preferences should to be used as the reference point (i.e., the cost or benefit of the EPA should be compared to the “current” situation). On the other hand, European negotiators insisted on the fact that the reference point should be the Generalized System of Preferences (GSP), since after 2007 the unilateral preferences could not exist anymore and this alternative was thus irrelevant.

The second option obviously leads to a more optimistic assessment of the EPA. Since 2008, the European position is de facto the most relevant, even if some observers think that it does not follow the initial commitment’s spirit. Yet this does not necessarily mean that the debate on this question is closed. In fact, we now need to know if the actual base for the negotiation of a regional EPA should be the GSP applied to exports to the EU, while also maintaining the actual pricing structures of Western African countries, or if the temporary agreements signed, and the associated concessions, must be used as a reference for Ghana and the Ivory Coast. In the latter case, the regional EPA usually results in a reduction of the liberalisation granted at a bilateral level and thus in a fiscal gain. Finally, the question of ECOWAS’ common external tariff arises: fixed at an intermediate level between the UEMOA (West African Economic and Monetary Union) level and the Nigerian level, for example, it would increase the tax loss of French-speaking countries and reduce that of Nigeria. We also have to keep in mind that the tax impacts will depend on the degree and the sequencing of the liberalisation. Greater market opening will result in higher losses from foregone custom duties. The very composition of the list of sensitive products plays an important part, and also presents a difficult choice for the African negotiators: should they protect the products that generate tax revenue or products considered critical to the rural or industrial development?

Preliminary results: a mixed picture

Since the negotiation process is not over yet, it would be a premature to present specific numbers on the fiscal impact of the EU-West Africa EPA. However, basing our judgment on the existing outlines of the agreement, we can give a couple of rough estimates. Overall, customs revenues should be reduced by 30% at the most[iii] for the whole ECOWAS, and between 20% and 40% for the different countries, which represents 10% of total indirect tax revenue.

Tackling the challenge: possible strategies

The evaluation of the net fiscal impact is just one step of the negotiation process. It is also advisable to consider solutions to the problem. First, one could consider the option of “fiscal neutralisation” through a transfer from the EU under the form of budget support. In this respect, it is important to note that the amounts to be invested are less important than the net fiscal impact previously assessed. In fact, if the transfer programme is implemented at the same time as liberalisation is undertaken, public spending will not have to decrease and economic growth will not slow down (which generated additional losses). Secondly, national fiscal policy can be reformed and adapted. For example, some customs duties can be replaced by excise duties, the VAT can be increased or direct taxes modified to get back a part of the lost customs taxes. Indeed, the suppression of customs duty results in a decrease in domestic prices of imported goods, which benefits consumers and firms but triggers a loss for the state. In response, governments can maintain tax pressure and in doing so, retrieve the amounts earned by other economic agents. If this strategy is the most logical in the long term, it is also difficult to implement in countries that have set up numerous tax reforms and are confronted by a large informal sector that escapes official taxation. Any increase of the tax rate would then only encourage a switch to the informal sector, worsening the situation even more. Thirdly, the development policy associated with the EPAs, the PAPED (EPA Development Program[iv]), could sufficiently reinforce economic growth and generate additional tax income that will cover the direct losses due to trade liberalisation. This scenario is more optimistic, but its success will depend on the exact characteristics of the PAPED.

Conclusion

Western Africa offers a very relevant example of the use of economic models in free trade negotiations. Even when incomplete, these tools can help quantify some of the questions that are of critical interest to political decision-makers and help clarify the consequences of various options. Well used, they offer a coherent framework for the negotiating parties, which force participants to explain their offers and expectations. In the context of estimating the fiscal impact of the EPAs, the selected approach allows the parties to tackle the problem head-on and to immediately define a tailored and negotiated strategy, instead of denying the reality or allowing utopian fantasies obscure the debate.

Author: David Laborde is an economist at the International Food Policy Reseach Institute (IFPRI - Washington DC) and has collaborated with the ECOWAS-EC Regional Preparatory Task Force in the context of ECOWAS EPA negotiations since 2007.

[i] It is necessary to note, however, that in the past, the EU did help West African countries to deal with the revenue loss related to their own regional intégration.

[ii] It is important to keep in mind that in certain cases, the efficiency of this collection can be endogenous to the deregulation process, if the latter comes with a tax reform or investments in customs administration, etc.

[iii] We take the upper bound of the opening rate, with an 80% threshold.

[iv] See on this subject our issue of Trade Negotiations Insights dedicated to the PAPED/EPAD : Trade Negotiations Insights, Vol9, No5, June 2010. See also : ECDPM. 2010. The EU Commitment to Deliver Aid for Trade in West Africa and Support the EPA Development Programme (PAPED). (ECDPM Discussion Paper 96). Maastricht: ECDPM