Sapienza University Chair in REUTRADE 10 th May 2018 The Impact of Multilateralism and Regionalism for Developing Countries Riccardo D Angeli Antilla Fürst Alice Giro Cecile Hermse Eli Kalusheva Federico Maggiulli Sveva Morgigni Alice Politano
Table of Contents Part I - Introduction Trade Liberalization & International Trade Agreements: Multilateralism (GATT/WTO) and Regionalism & Generalized Systems of Preferences. Part II - The Impact of multilateral trading system on trade - Rose s perspective IIa - Key paper: AER 2004 Presentation of Rose s analysis (GM s explanation) & findings IIb - Challenges to Rose s argument - overview of the main Criticism to AER 2004 Part III - The Impact of multilateral trading system on trade - Nenci s perspective Presentation of Nenci s analysis and findings Part IV - Multilateralism and Regionalism in Developing Countries - are they effective? Criticism & Recommendations/Final Observations & Conclusion
Part I - Introduction 1. Overview of Trade Liberalization Economists disagree about a lot, but not everything. Almost all of us think that international trade should be free. (Rose, 2004). Trade Liberalization is a good thing.
1.1 Evolution of Trade Liberalization 23 January 1860 - The Cobden-Chevalier Treaty: the first modern trade agreement in history The MFN clause: the clause requires countries to grant the same treatment to all their trading partners. Great increase in world trade the network of Cobden-Chevalier Treaties Early 1900: WWI & Great Depression phase of protectionism Second half 1900: new phase of trade liberalization - creation of GATT (1947) and then of WTO (1995)
Average World Tariffs 1860-2000
2. Overview on International Trade Agreements When countries seek to liberalize their trade, they enter into trade agreements, which are pacts aimed at reducing or eliminating trade barriers between states. There are two main types of Free Trade Agreements: Multilateral Agreements Regional Agreements
2.1 Multilateral Trade Agreements MTA are pacts which are negotiated among large groups of countries and are usually difficult to negotiate. The General Agreement on Tariffs and Trade (GATT) Established in 1947 after the end of WWII, to reduce protectionist measures. MFN Clause > MFN status extended to every member state > dramatic widening of the international trade network (founding principle: nondiscrimination). Signed by 23 countries in Geneva on October 1947. Since its foundation, nine rounds (Doha Round still not concluded yet) 5 decades of work: involvement of more members and negotiation of more issues. Moreover, great increase in the role of Developing Countries.
The World Trade Organization (WTO) The WTO replaced the GATT in 1995 International organization that regulates international trade It is basically tasked with the opening of markets It has 164 members WTO held eleven International Conferences since 1996. The GATT first and the WTO later have dramatically reduced tariffs in the last 7 decades.
2.2 Regional Trade Agreements Regionalism = preferential trade agreements among a subset of nations. Based on: identifiable geographical region, geographical proximity, and an organization with a common sense of identity and purpose (economic, political, security/military, etc.) among the member states. Also called Preferential Trade Agreements (PTAs), because they grant preferential treatment only to the signatory countries, while maintaining barriers against other countries.
Types of Regional Trade Agreements Type Acronym Description Preferential Trade Agreement Free Trade Area Customary Union Common Market PTA FTA CU CM An agreement on the part of a set of countries to reduce but not eliminate trade restrictions among themselves An agreement that eliminates trade restrictions among a set of countries An agreement to eliminate trade restrictions and to adopt a common external tariff An agreement to eliminate trade restrictions, adopt a common external tariff and allow the free movement of labor and physical capital among member countries
2.3 Generalized System of Preferences Generalized System of Preferences (GSP): a preferential tariff system which provides for a formal system of exemption from the more general rules of the WTO, specifically from the MFN clause. It is granted to developing countries - beneficiary countries - by certain developed countries - donor countries. It is not negotiated: the preferential treatment is non-reciprocal. Aim: lower tariffs for the least developed countries encourage exports contribute to poverty eradication (support economic growth and job creation in the beneficiary country) GPSs promote sustainable development and good governance.
Rose s Research Article: Do We Really Know That the WTO Increases Trade? Panel of data from 1948 to 1999: since the GATT was established in 1947 and the WTO in 1998 the data only considers the years where the agreements were in place. It is based on 175 countries. Standard gravity model used to estimate the empirical effects of membership in the GATT/WTO on trade. These empirical effects are the come out of a statistical research based upon variables estimated according to what Rose defines as membership in the GATT/WTO: the liberalization of trade.
The Gravity Model The functional form used for estimate empirically the gravity model tog international trade flows was first applied by Jan Tinbergen in 1962. It is inspired by Isaac Newton s law of universal gravitation.
How come the gravity model suits Rose s research? It is flexible: it can be implemented with several extra conditioning variables (dummies) in which the distance between countries branches out, and hence affects trade. The variables are mostly proxies: they describe the behavior of a specific phenomenon that can t be directly observed. GDP is a proxy for measures of standard of living or quality of life. In the gravity model the GDP stands for the mass of a country. The variables are weighted with their specific nuisance coefficients.
X ijt : the average value of real bilateral trade between i and j at time t, by averaging the four possible measures: exports from i to j, exports from j to i, imports from i to j and imports from j to i. Y : real GDP. Binary dummies: equals 1 if the condition is verified and 0 otherwise. T t : comprehensive set of time fixed effects. ε ijt : the omitted other influences on bilateral trade.
Rose s dummies They represent other forms of control over the real bilateral trade between country i and country j. Landl stands for the number of landlocked countries in the pair Island stands for the number of island countries in the pair Lang stands for common language Cont stands for land border shared ComCol stands for colonies after 1945 under the same colonizer CurCol stands for one currently being a colony of the other ComNat stands for both being currently part of the same nation CU stands for same currency FTA stands for both belonging to the same regional trade agreement GSP stands for one being a GSP beneficiary of the other their value is either 0, 1 or 2 they are binary variables : their value is either 1 or 0.
The dummies that interest us the most in this study: Binary variable which equals 1 if both i and j are GATT/WTO members at time t. γ₁ measures the effect on international trade if both countries are members. If trade is created while they are members, the value will be positive. There are no specifications based on the level of development of countries i and j in these specific dummies (but we can find it in the GDP per capita variable) Binary variable which equals 1 if either i and j is a GATT/WTO member at time t. γ₂ measures the trade effect if one country is a member and the other is not. If trade is diverted from nonmembers, the value will be negative.
Default specification: augmented gravity model estimated with OLS, year fixed effects, and robust standard errors. The results are those expected landlocked countries trade less Country-specific fixed effects added to the benchmark equation Even if the first coefficient is positive, it is small compared to other effects in the column. Small negative coefficients. Neither is statistically different from zero at conventional significance levels. These specifications were unable to suggest that membership in the GATT/WTO has had a positive and strong effect on international trade.
Further specifications made by Rose Estimation on individual years at five-year intervals, from 1950 to 1995: GATT membership had significant effects in the first considered year but through time, with the increasing number of members these fade away and eventually became negative. The only negotiations rounds that had an economical large effect were the first two (Annecy and Torquay) but even these were statistically insignificant. Samples divided by country groupings Quadratic gravity terms added as nuisance variables. Coefficients for each regional trade agreement. The aggregate trade from either country to the rest of the world added. Re-estimation using five-year averages instead of annual ones. Dynamic considerations: a set of dummy variables which equal 1 if one of the two countries entered the GATT/WTO 5, 10, 15 or 20 years
Results Mainly the same through the different specifications: γ₁ and γ₂ are found to be economically small and statistically insignificant. GATT/WTO membership has an effect on trade only if the other variables change, therefore taking into consideration standard gravity effects erases any large effect of it on bilateral trade. There is little empirical evidence that the membership has a substantial and statistical positive effect on trade. Its role on promoting world trade is mostly limited. Nevertheless, this result doesn t mean that GATT and WTO have not increased trade. Rose s study uses data until 1999 so he doesn t exclude that WTO effects will be different, and he sets suggestions for future researches on the topic (differentiate exports and imports, decomposing trade by industry, shades on the membership meaning)
Criticisms to Rose s Work
Rose s Work - Rose, Andrew K. (2004b) Do WTO Members have More Liberal Trade Policy? Journal of International Economics. Test whether GATT/WTO membership had effects on trade policy - Rose, Andrew K. (2005c) Which International Institutions Promote International Trade? Review of International Economics. comparison between membership in GATT/WTO and in IMF and OECD Conclusion: GATT/WTO had no significant effect on trade growth or trade policy Rose s work importance opened a huge economic and political debate about the effects of an institutionalized system for world trade.
Main Criticisms Many were inspired to analyze the field and arrived at different conclusions. Rose s work subject to many criticisms - mainly four: INAPPROPRIATE PROXY FOR VARIABLE OF INTEREST INAPPROPRIATE DATA POOLING INAPPROPRIATE HANDLING OF FIXED EFFECTS SAMPLE SELECTION BIAS
INAPPROPRIATE PROXY FOR VARIABLE OF INTEREST Challenge to the quality of the trade flow measurement in Rose s work the WTO is not limited to the opening of markets. Biased analysis of the decrease in barriers, overlooks many important factors: - countries entering the GATT/WTO are not required to immediately lower their barriers on accession, there are transition periods; - many countries already enjoyed some form of preferential treatment because of MFN clause; - many countries liberalized before entering the GATT/WTO specifically to facilitate accession; - many developing countries export goods which are not subject to severe protectionist measures (fuels and minerals); - before the WTO in 1995, Developing Countries where not required to significantly lower their trade barriers
INAPPROPRIATE DATA POOLING 1. No consideration of asymmetries between countries and sectors 2. Sharp distinction between participants and non participants in the GATT/WTO. Rose only considers formal members (de jure members) as participants in the GATT/WTO, overlooking a large group of countries that are involved with the WTO, who imposes on them rights and obligations (the de facto (informal) GATT/WTO participants). Nonmember participants: formal Members colonies, provisional applicants, new sovereign states Rose: non-participants, Biased estimates, based on wrong comparisons.
INAPPROPRIATE HANDLING OF FIXED EFFECTS Most of Rose s regressions do not include country-specific fixed effects in the gravity equations jeopardizing results accuracy Rose is not considering the impact of the multilateral resistance effect, i.e. the barriers to trade that each country faces with all its trading partners. Controlling for relative trade costs is crucial for a well-specified gravity model. Clarification of the rationale for including multilateral trade-resistance (MTR) terms: two countries surrounded by other large trading economies, say Belgium and the Netherlands bordered by France and Germany respectively as well as by each other, will trade less between themselves than if they were surrounded by oceans
SAMPLE SELECTION BIAS Sample selection bias : choosing non-random data for statistical analysis a flaw in the sample selection process, where a subset of the data is systematically excluded due to a particular attribute mining the statistical significance of the test, or produce distorted result the sample obtained is not representative. Rose s case: selection bias = omission of zero trade observations. Rose overlooked the extensive margin of trade, i.e. whether two countries have economic exchanges at all not considering the zeros leads to a loss of information on new trading relationships generated by the GATT/WTO when country pairs initially did not trade, but started to trade after one or both of them joined the GATT/WTO.
Silvia NENCI: Tariff liberalization and the growth of world trade: a comparative historical analysis for the evaluation of the multilateral system 2009
Nenci s study - Criticism to Rose on three points: 1) lack of comparison between different periods 2) lack of distinction between institutionalized system and informal system 3) lack of distinction between industrialized and developing countries - Objectives of research: 1) verify the relationship between tariff barriers and world trade growth from comparative and historical analysis 2) derive indications on effectiveness of current multilateral trade system in stimulating world trade in the long term
Trade & Tariffs Post WWII period: - INTERACTION: participation in international trade EXPORTS - INTEGRATION: reduction of tariff barriers: TARIFFS
Trade & Tariffs /2
Nenci s empirical study AGGREGATE ANALYSIS: - measure the impact of average world tariffs on world trade on an aggregate level - based on three periods: trade liberalization on a bilateral basis (1871-1939), the restoration and non-restoration on tariff protections (1901-1939), and the liberalization on a multilateral level of 23 countries (1951-1986) - trade interpreted as exports PANEL ANALYSIS: - difference between industrialized and developing countries - focus on post WWII period - trade interpreted as imports
Aggregate analysis x = world exports y = world income p = measure of relative competitiveness t = average world tariffs
Aggregate analysis outcomes
Aggregate analysis outcomes /2 Systems of trade like the one generated by the Cobden-Chevalier or the one implemented by GATT, underline the scarce significant of multilateral trade agreements in boosting trade. In the post-world War II period, the effect of tariff changes on world trade is much less significant than in the previous period, probably in function of the loss of importance of tariff barriers and the emergence of so called non-tariff barriers. Nevertheless, the relationship between tariff changes and trade growth is confirmed from the long-run equilibrium.
Panel analysis M = imports Y = domestic income P = relative export and import goods prices T = tariffs
Panel analysis outcomes
Panel analysis outcomes /2 There is a greater significance of the relationship between trade liberalization and tariff barriers in the case of industrialized countries in comparison to developing countries. The Tokyo Round appears significant for both countries. This can be explained due to the effect that developing countries liberalized very little in the postwar period: they adopted free rider behavior, meaning they could take advantage of the liberalization without obligations of reciprocity nor benefiting directly from the interaction effect. The relation between tariffs and trade, despite being confirmed at the empirical level, appears to be less important and significant in today s trading operations.
Conclusions 1. There is a long-run relation between tariffs and trade, but it appears to be less significant in the period post WWII. 1. The WTO s reason of being is to be sought in its ability to establish a favorable climate for international trade at world level that operates through the establishment of certain rules and standards of conduct for the major players: Nenci names it the GATT effect. 1. The decrease of importance of tariff reduction is even stronger in the case of developing countries. 1. The success of the GATT seems to be more closely linked to the interaction effect, here understood as participation in international trade, than to a specific integration effect, understood as a reduction of tariff barriers. 1. An effective world trade organization is essential especially for developing countries, because the alternative presented by bilateral and regional agreements would increase trade diversion and discrimination.
The Doha Round The Doha mandate specifies that the needs of developing countries and least developed countries should be taken into account. Although the Doha Round has been labelled a development round, its limitations with respect to providing enough policy space for development have been evident. WTO members at the launch of the Doha Round in November 2001 committed themselves to the objective of providing duty free and quota free market access to least developed countries products. The experience of the Doha Round has raised questions about the ability and willingness of governments to advance the multilateral agenda.
Regional Trade Agreements Slow progress in the Doha negotiations has spurred developed countries, in particular, to pursue bilateral and regional trade agreements. Advocates of RTAs: - economies of scale - competition - attraction of foreign direct investment - the only option However, this proliferation of regional agreements has created a spaghetti bowl of crisscrossing arrangements.
Trade Creation VS Trade Diversion Preferential trade opening allows some domestic production to be replaced by imports from more efficient firms located in preferencereceiving countries, leading to welfare gains TRADE CREATION At the same time, preferential trade agreements may reduce imports from more efficient non-member countries, implying a welfare loss TRADE DIVERSION
Dispute Settlement Mechanism Regional growth had also had some impact on the unity and strength of international dispute settlement and remedies systems. There is the imminent danger that as countries pursue deeper integration within RTAs, dispute settlement provisions contained in the RTAs could build jurisprudence conflicting with that of the WTO. However, countries generally choose to use the WTO to resolve issues that may also fall within the scope of their RTA.
Repercussion on WTO Governance The embrace of RTAs by virtually every trading nation carries systemic implications for the multilateral trading system, most notably through increased discrimination and complexity in trade relations. It has been established that RTAs are inherently inconsistent with the objectives of the multilateral trading system of facilitating nondiscriminatory trade among all nations, because they preserve exclusive preferences among their members
The Link Between WTO and RTAs I believe that to the extent that PTAs are motivated by a desire for deeper integration rather than market segmentation, there could be a role for the WTO to promote greater coherence among non-competing but divergent regulatory regimes. There is no doubt that we need to build towards a more stable and healthier trading environment, where alternative trade policy approaches are mutually supportive and balance equitably the needs of all nations. Former WTO Director-General Pascal Lamy
Relation between MTS and RTAs Building blocks or stumbling blocks?
What role do regional agreements play in the multi-level process of international trade? Towards a model of best practice in regional agreements
MULTILATERALIZE REGIONALISM : BEYOND ARTICLE XXIV? - Distinguishing various categories of policies - Basic guidelines for RTAs - See legal avenues towards multilateralising regionalism