Assessing the Impact of Preferential Trade Agreements A

Assessing the Impact of Preferential Trade Agreements: A Disaggregated Approach Economics Brown Bag Series Jad Chaaban 16/03/2007

Let me start by a story • A Lebanese policy maker asks you, a professional economist, for advice on the following: – Should we sign a free trade agreement with Europe? – If yes, how can we evaluate the impact this agreement would have on our consumers and producers?

In fact, the EU has had a long history of agreements with Lebanon • A cooperation agreement between the European Community and Lebanon was signed in 1978 • The cooperation agreement did not come into force • In 1995, Lebanon joined the Barcelona process and Euro-Mediterranean partnership • In 2002, Lebanon signed an interim association agreement with the European Union • On May 1, 2004, Lebanon joined the European Neighborhood Policy talks

What to do? First, you go home, and review the literature on trade agreements • This agreement is a typical Preferential Trade Agreement (PTA) – consumers benefit from lower import prices – cheaper imported inputs for producers – Yet stiffer competition; the government looses revenue • Trade diversion and trade creation are at stake: other trade imports (Arab countries) might be hurt as a result of trade diversion • Bhagwati and Panagariya, 1996; Winters, 2004

Then you look at the overall context of Lebanon’s relation with the EU • Lebanon is facing fierce competition in the region • The EU has a long history of economic agreements with Mediterranean Partner Countries MPC • The Association Agreement entered into force in only 4 countries • Early signatures and economic fundamentals have played a role (Baker, 2004) • Growth competitiveness central in explaining divergence among partners

Diverse features of MPC economies (2004) (Baker, 2004)

Status of EU-Med Cooperation (Baker, 2004)

Trade between EU and MPC (billion EUR) (Eurostat)

Then you see what other free trade agreements were already in place • Lebanon has one multilateral free trade agreement that GAFTA (Greater Arab Free Trade Area) • several other bilateral trade agreements with individual Arab countries. • Yet the share of total imports under these agreements evolved from 3. 69% in 1997 to 4. 34% in 2002 • low but expected to be boosted upward after the Association Agreement with the European Union, Lebanon’s main trading partner, comes into effect in 2007.

Evolution of Imports under PTAs 1997 -2002

Then, you narrow down the question • What is the impact of the EU PTA on the Lebanese import sector? • Will the EU have trade creation to an extent that hurts other partner countries’ exports to Lebanon?

To answer this, the tools are available • The typical approach here is to estimate a “gravity” equation which represents bilateral trade flows as a function of – income, population, distance between trading partners and membership in a common regional arrangement – Frankel, 1997, Soloaga and Winters, 1998 and Cernat, 2001); François and Reinert (1997).

Gravity Model

Gravity Model • Typical Determinants of volume of trade 1. Economic size 2. Distance 3. Cultural affinity 4. Geography 5. Multinational corporations 6. International institutions 7. Political borders

Gravity Model • Related to Newton Law • Estimates of the effect of distance from the gravity model predict that a 1% increase in the distance between countries is associated with a decrease in the volume of trade of 0. 7% to 1% • Generalized gravity model: • Adding other controls:

Perfect, you go to the policy maker with estimates, but…he is unhappy! • The policy maker wants to know the impact on imports IN EACH SECTOR • Because he/she believes that the Agreement will have different effects depending on sectors (textile, agriculture, industry…) • Can the gravity model answer this? No, because handles only aggregate figures

What can you do? • In this paper we propose a methodology that allows: – Sector-specific evaluation of trade creation and trade diversion – Handles very disaggregated data: Using for the first time data from a national Customs database, highly disaggregated transaction-based – Theory: Based on Consumer Choice Theory – Model: Simple to formulate – Estimation: Empirical specification simple to estimate

Model • Demand system that satisfies basic economic assumptions on consumer behavior. • Allows for direct and straightforward inference on consumer reaction to prices. • Consistent with aggregation: the final demand for a given good obtained by direct aggregation of individual consumer demands. • Deaton and Muellbauer (1980) • Notation: • Subscripts i and j denote distinct imported goods (food, textiles, machinery…). Subscripts h and k denote sources (regions from which the goods are imported: European Union, North America, …).

Detailed assessment of the impact of EU agreement • Evaluate the effects of the Lebanon-European Union PTA on the Lebanese import sector • An exceptional, highly disaggregated, dataset on Lebanese import transactions for 1997 -2002 • Results show that trade creation occurs for most import sections from the EU, while trade diversion only occurs in a small number of sections for Arab and regional countries and North and Latin America. • The Rest of the World stands to loose the most from the EU’s PTA with Lebanon, with decreasing import shares in almost all import sections.