The Greater Arab Free Trade Area: An ex-post appraisal within an imperfect competition framework

The aim of this research project is to provide some new and original insight concerning the GAFTA welfare and trade impact, 10 years after the implementation of this agreement. This project starts with the description and a critical analysis of economic integration and trade in the Arab world in the past decades. Particular emphasis is put on the provisions included in the GAFTA agreement. Its limitations are also discussed. Recent patterns in regional integration are also compared with the analysis of trade flows in Arab countries, especially since the implementation of the GAFTA agreement in 1998.

The next parts of the project are based on a twofold approach which relies on new theoretical developments in regional economic integration. The first approach involves a theoretical model of regional integration, followed by inquiries implemented in selected GAFTA countries and selected industries. This approach makes it possible to highlight several possible welfare effects of economic integration in the Arab region. It does not only include the gains related to the perfect competition framework (exploitation of comparative advantage, more efficient use of factors of production) but also the additional gains due to imperfect competition (terms of trade improvement, reduction in trade costs, existence of scale economies, greater product varieties for consumers) as well as dynamic effects (increase in foreign direct investment, growth effects) and the impact of economic distortions (taxes/subsidies).

This qualitative analysis is complemented by an empirical model (representing the second approach) which aims to quantify the trade effects of GAFTA. This model is an original combination of gravity models and supply-demand export models. Its main contribution is to simultaneously include gravity variables as well as export supply variables, especially scales economies and product differentiation. This model is subsequently estimated in order to calculate the effect of GAFTA on intra-regional trade, by using several appropriate estimators, of which Hausman and Taylor (which tackle endogeneity problems), GMM in dynamic models as well as transformed fixed and random effect models (for addressing multiple heterogeneity concerns).

The main results of this study are the following:

1.    Although the first attempt for regional integration dates back to the 1950s, the GAFTA agreement is certainly the most outreaching one. Indeed, tariffs have been fully eliminated on 1.1.2005; currently, it covers 17 countries in the Arab region; it relies on a negative list approach; it includes agricultural products as well as an additional regional agreement concerning trade liberalization of services signed in 2003 in addition to research and technological cooperation.

2.    However, the GAFTA agreement shows some limitations. First, although tariffs have been removed, some GAFTA members have introduced new trade barriers, which can be taxes or other NTBs. Secondly, the GAFTA agreement remains a perfect example of “shallow integration”. It suffers a number of problems, including the absence of full fledge dispute settlement mechanism (although there are efforts to have one), the inability to reach a detailed rules of origin scheme , a weak system of harmonized standards, the lack of harmonization of competition rules as well as the lack of protection of intellectual property rights. In addition, there is no provision for labour movement. Finally, there is a lack of supra-national institutions or a strong leading Arab country to solve the problem of disputed matters. In other words, mainly all aspects of “deep” integration are absent from GAFTA.

3.    Intra-GAFTA trade has significantly increased since GAFTA implementation in 1997 (+15% at a yearly average since 1997). This increase is greater than world exports (8%) and than extra-GAFTA exports (14%).

4.    As a proportion of total trade, intra-regional trade increased from 9.8% in 1998, to 11.2% in 2005. When excluding oil products, this share rose from 13.5% to 18.0% over the same period. This intra-regional trade share is comparable to some other regional grouping such as COMESA or ASEAN. However, it remains much lower than intra-regional trade in the EU or the APEC.

5.    More precisely, there are differences across countries and commodities. For example, some countries have strongly increased their intra-regional trade (Egypt, Jordan, Lebanon, Syria, Tunisia) whereas some other have experienced a stability or even a decline (mainly Gulf countries). At industry-level, the most important increase in intra-regional trade concerns food, manufactured products as well as machinery and transport equipment. Conversely, crude material, oil and fats have not enjoyed such an increase.

6.    Arab countries have succeeded to some extent in diversifying the products exported or imported. As a matter of fact, eight Arab countries have exported more than 200 products in 2005 (at 3-digit group level), against three countries only in 1995.

7.    The extended theoretical model of regional integration in imperfect competition (presented in Part 2) makes it possible to state that welfare effects due to regional integration can be decomposed in different channels:

a.    Perfect competition effects (trade volume, trade costs)

b.    Terms of trade effects

c.    Imperfect competition effects (production, scale economies, product varieties)

d.    Dynamic effects (investment, growth, FDI)

e.    Economic distortion effects (wages, domestic taxes)

8.    An application of this model to GAFTA countries through an appropriate inquiry reveals that:

a.    GAFTA has a positive effect on the volume of intra-regional trade. There are however some differences across countries and industries. As a matter of fact, almost all countries seem to have enjoyed positive trade effects, with the possible exception of Lebanon, for which the firms investigated complained about differences in energy prices due to subsidies in the other GAFTA countries. This has created an unfair competition situation where Lebanese firms are disadvantaged in the GAFTA regional market. Turning to industry-specific effects, the food  industry and chemicals have taken advantage of GAFTA, whereas textile and clothing have not enjoyed intra-regional trade liberalization so much, for several reasons (increased NTB, dumping, absence of differences in production costs and consumer tastes across countries, etc…).

b.    The reduction in NTBs has a neutral effect, since this reduction provided by the GAFTA agreement has been supplemented by the erection of new NTBs by some GAFTA members.

c.    Imperfect competition effects (production effect, scale economies, product varieties) are only slightly positive. This result contrasts to the very positive effects recorded for North-North regional integration, especially the EU. Several reasons can explain this difference: the persistence of NTBs which impede strong production effects and scale economies, the lack of product differentiation which impedes product variety effects, the lack of taste differences. As a result, trade is mainly inter-industrial with small imperfect competition effects. Finally, the lack of deep integration is a brake for creating a real single market where production effects and scale economy can really occur.

d.    Distortion effects have a significant impact, especially differences in taxes/subsidies across countries. Some countries take advantage of subsidizing their own production and exports (especially Saudi Arabia, the United Arab Emirates and Egypt) at the expense of the countries with the lowest subsidies (Lebanon).

e.    Terms of trade effects and dynamic effects have not been determined. This is mainly due to the fact that the firms interviewed cannot identify the complex link between economic integration and its indirect effects on prices, investment, FDI or growth.

9.    In part 3, an original trade model based on new developments in gravity models as well as export-demand model is proposed. It makes it possible to identify the following trade determinants

a.    The traditional gravity variables (GDP, distance, common language)

b.    Trade costs variables (border effects, regional economic integration)

c.    Imperfect competition variables (scale economies, product varieties)

d.    Expectations

e.    Hysteresis due to sunk costs

10.    An application of this model to GAFTA countries through a set of appropriate econometric estimators (Hausman and Taylor, Arellano, Bond and Bover, Transformed fixed and random effects models, etc…) makes it possible to quantify the impact of the above variables on intra-regional trade in GAFTA countries. This leads to the following results:

a.    Standard perfect competition trade effects significantly affect trade (GDP and distance).

b.    The trade effect of the GAFTA agreement is positive. In particular, the model exhibits a significant trade creation. Small trade diversion is highlighted for imports but not for exports. Overall, the net trade creation is positive. It is estimated to be about 26% of GAFTA trade.

c.    However, most countries exhibit current trade levels which are below their fitted levels, as showed by the calculation of export potentials. This suggests that the GAFTA agreement has not made it possible to increase regional trade above its “normal” level, especially in Morocco, Tunisia, but also Egypt, Jordan and Syria.

d.    Imperfect competition effects are small. In particular, although scale economies are significant in GAFTA countries, they hardly increase trade flows. These results correlate those already found qualitatively with the inquiry. Again, the main explanation may be found in market structures, where products are poorly differentiated, consumer tastes are similar and trade is mainly inter-industrial. In addition, the absence of deep integration impedes GAFTA countries to take advantage of existing scale economies, since the remaining NTBs makes it difficult to exploit the economies by producing for a large unified market.

11.    The main policy implications which can be drawn from the results are the following. If the objective is to enhance the trade and welfare effects of regional integration in the GAFTA region, several policies can be undertaken:

a.    All the loopholes in the current agreement should be fully addressed and further step toward deep integration must be achieved: In particular, progress must be made in favour of the adoption of clear and detailed rules of origin, the actual removal of new NTBs and trade frictions among GAFTA members, the adoption of common standards, the free movement of entrepreneurs, the protection of intellectual property, etc…Such a deep integration will not only increase direct trade effects of regional integration, but also increase indirect effects (scale economies, and dynamic effects) through the establishment of solid foundations toward a more integrated area. In this regard, it is worth mentioning that liberalization of trade in services on a GATS+ approach will surely have a positive impact on deepening integration among GAFTA members.

b.    Another mean to enhance GAFTA integration could be achieved through the cumulation of rules of origin among some of the GAFTA members in their other regional agreements as Agadir. The utilization of such cumulation schemes is likely to force GAFTA countries to cooperate and is likely to result in better allocation of resources.

c.    There is a need to design a system which ensures that domestic distortions do not yield negative spillovers on GAFTA members. The case of different systems of energy pricing in GAFTA members has proved to have negative effects, especially for Lebanon. Hence, at least rules governing subsidies should be fully articulated and efficiently implemented within GAFTA.

d.    GAFTA members should start cooperating on enhancing regional trade and investments in sectors that have proved to have benefited so far from GAFTA as food and some chemicals industries. Moreover, the NTBs that are affecting intra-regional trade in other sectors as textiles should be seriously tackled.

e.    There is a need to start a serious program on building a comprehensive database and information system on intraregional trade and investment opportunities. In addition, since there is still a lack of knowledge of the GAFTA agreement and its provisions in many firms, more information should be provided concerning regional economic integration in the Arab world.

f.    From a political point of view, it is also crucial that GAFTA countries can rely on a closer political cooperation as well as on common institutions that can make possible to control trade liberalisation in the region and solve trade disputes.

g.    More generally, conditions for economic growth should be developed, such as the reform of the states, the development of cross-regional infrastructures, such as railway and highways, progress toward more trade and FDI liberalisation not only within the GAFTA area but also with the other partners, etc..