The African, Caribbean and Pacific (ACP) Group of States are exploring the opportunities of a free trade agreement. Most ACP members are African countries, which are currently negotiating the Continental Free Trade Area (CFTA).

With the exception of North Africa, an ACP agreement would extend an African CFTA to include the Caribbean and Pacific members of the ACP group of countries. In this analysis we show that while trade can be significantly improved by removing all remaining tariffs, this can be undermined if five to ten per cent of tariff lines are chosen to be exempt. Furthermore, the gains from addressing non-tariff measures offer greater scope to increase trade.

Both an African free trade agreement and extending it to include the Caribbean and Pacific countries will generate positive gains. Negotiators should focus on reducing non-tariff barriers and harmonizing regulatory measures that outweigh the impact of tariffs.

---

Introduction

The African, Caribbean and Pacific Group of States (ACP) are exploring the opportunities of a free trade agreement. The group of 79 developing countries was established to negotiate and implement cooperation treaties with the European Union but formed its own political identity in 1975. The EU has been negotiating Economic Partnership Agreements (EPAs) with five regions in Africa, the Caribbean and the Pacific to replace the existing preferential rules to make the ACP-EU relations compatible with WTO rules. Many ACP countries also have free trade agreements with other ACP countries as well as with third parties such as PACER in the case of Pacific countries.

The 54 member States of the African Union have agreed to establish the Continental Free Trade Area (CFTA) with the aim of substantially reducing the barriers to trade between the member countries. The 48 African ACP members account for 85 per cent of ACP trade. Negotiations have focused on the speed at which tariffs are removed and the extent to which various products may be regarded as sensitive and exempt from reductions.

Ten to fifteen per cent of tariff lines have been suggested as an appropriate proportion of sensitive products, depending on the development status of the country. However, negotiators are unlikely to agree to this because for some countries ten per cent of tariff lines cover all imports. At the other extreme, exemption of one per cent of tariff lines would seem to be overly ambitious, and is also unlikely to be the basis for agreement.

The authors use a well-known computable general equilibrium (CGE) model to quantify the impacts on trade and welfare for liberalization scenarios in Africa (CFTA) and the entire ACP group. They begin with full tariff eliminations for all products in the CTFA and then compare the outcome to a CFTA with exemptions for five per cent of products. These exemptions reduce the increase in trade by more than 60 per cent.

They then show that the potential benefits of tariffs reductions are by far outweighed by eliminating non-tariff barriers and promoting the convergence of regulatory measures, an area also considered in the CFTA negotiations. This suggests that negotiators should focus on non-tariff measures rather than exemptions for sensitive products. Extending the CFTA to include Caribbean and Pacific countries would lead to relatively small additional gains, although addressing non-tariff measures provides greater scope for increases in trade.

Six scenarios are run. Three involve the CFTA alone. The first is complete removal of tariffs (scenario CFTA). The second allows for exemptions for five per cent of sensitive products (scenario CFTAx). The third quantifies the impact of reducing non-tariff barriers without tariff reduction (scenario CFTAn). The remaining three scenarios replicate the first three but extend the analysis to the whole of the ACP region, that is, including the Caribbean and Pacific countries (scenarios ACP, ACPx and ACPn).

African countries as a group gain $3.6 billion per annum from the complete elimination of applied tariffs but even a modest exemption of five per cent of products reduces these gains to $1.5 billion. Reducing the trade distorting effect of NTMs leads to gains of $20 billion. For the Caribbean and Pacific countries, they gain $620 million from entering a free trade agreement with Africa, but once again the gains are reduced by two thirds if exemptions are permitted.

Removing tariffs and addressing non-tariff measures within Africa makes good sense. Extending the agreement to across the Pacific or Caribbean increases the gains by a modest amount.

Existing barriers to trade

Tariffs

Average MFN tariffs in ACP countries vary a great deal, from zero to 33 per cent. North and Southern African tariffs are rather low while West, Central and Eastern Africa tend to have higher tariffs. A customs union exists in West Africa but external tariffs are high. Average tariffs of Caribbean and Pacific countries vary as well. Many have simple average tariffs of around 10 per cent with slightly lower trade weighted tariffs. Agriculture tariffs tend to be twice as high as industrial tariffs.

Trade is also fragmented, with many countries having little or no trade with several other African countries, or only importing 200-300 items. This is important for the selection of sensitive products. Negotiators have expressed an interest in exempting from tariff reduction ten per cent of the tariff lines. For some countries, such as Zambia, this could cover all intra-African imports.

Non-tariff measures

UNCTAD has made a determined effort to list non-tariff measures in various countries and to estimate ad valorem equivalents from those listings. However, as yet, it does not have ad valorem equivalents (AVEs) of non-tariff measures for individual ACP countries or regions. Instead, we use estimates for Africa as a whole estimated by Cadot et al. (2015). We assume the NTMs estimated for Africa are applicable to each country. The absence of country specific data means each country has the same value for a given product or sector.

For the Caribbean we use NTM estimates taken from South America and for Pacific countries we use Asia estimates. Data for Latin America and Asia are somewhat similar. The SPS and TBT AVEs are modelled as productivity shocks. The others are treated as tariffs.

Regulatory NTMs have benefits, for example in limiting the spread of infectious diseases and pests, and therefore it is unrealistic to remove them completely. Some measures may be too entrenched to be negotiated away. That leads to a challenging decision about what proportion of NTM-related trade costs may be removed. Knebel and Peters (forthcoming) estimate that about a quarter of the trade distorting effect of technical NTMs can be removed through regulatory cooperation. In our simulation, we therefore assume that a quarter of the existing cost effects of SPS measures and TBT can be reduced. Outright non-tariff barriers, such as quotas and price controls, are fully eliminated.