Table of Contents
2 Previous Research
3 Contents of both Agreements
3.2 The EU-Mexico FTA
4 Theoretical Framework of FTAs
4.1 Static Analysis
4.2 Dynamic Analysis
4.3 Political Economy
5 Development of Trade
5.1 Trade over GDP and Trade Shares
5.2 Commodity Groups
6 Econometric Analysis
6.1 Methodology: Time Series Econometrics
6.2 The Model
6.3 Regression Results - Export Demand Function
6.4 Regression Results - Import Demand Function
Appendix 1 - Mexico’s total exports to and imports from NAFTA
Appendix 2 - Mexico’s total exports to and imports from the EU
Appendix 3 - NAFTA’s import and export shares
Appendix 4 - EU’s import and export shares
Appendix 5 - SITC classification
Appendix 6 - Regression results for variables in levels
Bachelor’s Thesis in Economics
Title: Mexico’s free trade agreements – NAFTA and the EU-Mexico FTA in comparison
Author: Jan Prothmann
Date: December 2004
Subject terms: FTA, NAFTA, EU-Mexico FTA, time series analysis
Mexico has concluded various free trade agreements over the last decade, with NAFTA and the EU-Mexico FTA being the most important ones. In this paper, the effect of both agreements on bilateral trade is presented through descriptive statistics on the one hand, and by applying a time series analysis in the form of an export and import demand function on the other. The results of the graphical analysis indicate that the coming into force of NAFTA has led to a rise in Mexico’s trade with that bloc, even though the econometric analysis does not support the assertion that the conclusion of the agreement is responsible for that.
The EU-Mexico FTA has not led to a significant rise in Mexico’s exports to the EU, mainly due to transport costs. These costs are also the dominant factor in explaining exports from Mexico to its NAFTA partners, and therefore lead to the conclusion that the change in transport costs over time is more important than tariff reductions granted by the parties. Interestingly, Mexico’s imports from these two blocs are mainly determined by Mexico’s GDP and the exchange rate, rather than by transport costs.
Throughout economic history, various partnership agreements between countries can be found, all of which had the aim to gain from preferential trading possibilities. It might appear logical, without applying economic theory, that when trade barriers are reduced, trade promoting effects to different extents can be the result, at least between the concerned parties. Economic integration projects which all had the objective to motivate their trade relations were, for instance, England’s political commonwealth and free trade zone with its colonies, the German Zollverein, and the Austrian Customs Union with Modena, Parma, Hungary etc., just to mention a few. The objectives to form and take part in preferential free trade agreements have not changed till today - the basis is still to facilitate trade and experience economic benefits in general.
Mexico is not different in that sense, since it has also engaged in various preferential trade agreements with, up to date, 42 different countries on three different continents. Free trade agreements exist with Chile (1992), with Columbia and Venezuela (G3, 1995), as well as with Bolivia (1995), Costa Rica (1995), Nicaragua (1998), Israel (2000) and with El Salvador and Guatemala (Triangula del Norte, 2001). In multilateral forums such as the WTO, APEC (Asia Pacific Economic Cooperation), LAIA (Latin American Integration Association), FTAA (FreeTrade Area of the Americas) and OECD Mexico also holds strong positions.
The most important free trade agreement (FTA) for Mexico, according to various authors, is NAFTA, the North American Free Trade Agreement between the USA, Canada, and Mexico. This enabled Mexico to gain access to the world’s biggest market, namely the US market, and gave a signal to the world business community that Mexican government’s aims are directed towards the removal of trade restrictions and committed to freer markets (Ramirez, 2003). The agreement came into force January 1st, 1994, three years after the first negotiations were started. It can be seen as one of Mexico’s final steps in becoming an export-led economy, laying aside the import substitution model which was supported till the 1980s.
Another more recent preferential trade agreement which Mexico signed is the EU-Mexico FTA. In theory, it is considered to be Mexico’s second most important trade agreement, since it opened the possibility for Mexico to receive preferential market access to the world’s second largest market, the European market. The basis for this agreement can be found in 1995, when the Solemn Joint Statement between both parties was signed, in order to deepen and strengthen economical, political and co-operation relations, including the liberalization of trade. In 1997, Mexico and the EU signed the so called “Global Agreement”, consisting of the pillars Economic Partnership, Political Co-ordination and Co-operation. Finally, the EU-Mexico FTA for goods came into force on July 1st, 2000, and services are included since March 1st, 2001.
Both agreements named above undoubtedly had, and still have, a considerable effect on Mexico’s economy when it comes to trade expansion. After the coming into force of NAFTA, for example, a noticeable boost in trade between the parties can be found, especially from Mexico’s point of view. In the case of the EU-Mexico FTA, a trade enhancing effect should also be detected, because the general assumption of trade liberalisation is that it has a positive effect on trade. Nevertheless, differences in the impact on trade flows can be found when looking at the development of bilateral commerce, and questions arise as to which factors might be influential. Therefore, the purpose of this thesis is to provide a comparison of the two accords, and to look for reasons why the impact on trade flow between the parties is different. Why would a country engage in preferential regional trade agreements in the first place? Which determinants explain bilateral trade between the countries? Are some variables more important in one agreement than in the other? Those are the question this paper will try to answer. The relevance of this research purpose stems from the fact that the EU-Mexico FTA has not been in force for very long, but just about long enough to draw the first conclusions as to how influential it has been on bilateral trade between the parties.
The structure of this thesis is as follows. In section 2, an overview will be given about previous research that has been conducted on the subject, and section 3 summarizes the main contents of NAFTA and the EU-Mexico FTA. The theoretical analysis of why countries, especially Mexico in this case, would enter an FTA will be carried out in section 4, followed by an illustration of the real world trade developments in section 5. Section 6 provides the econometric analysis, and section 7 concludes the thesis.
2 Previous Research
Out of the vast pool of research that was conducted on both FTAs, only a small portion can be shown here, due to the limited space of the paper. Especially in the case of NAFTA, efforts had to be made in order to depict only the most important studies.
Among the first researchers to estimate the possible trade effects of NAFTA was Gould (1998). He applied a standard gravity model over the period 1980-1996 using quarterly data and found that, due to the FTA, US exports to Mexico have been 16.3 percentage points higher per year, and 16.2 percentage points higher in the case of US imports from Mexico. For bilateral trade between Canada and Mexico NAFTA is not found to have a significant impact.
The question how much trade growth between Mexico and the US can be attributed to the coming into force of NAFTA is also assessed by McDaniel and Agama (2003), who conduct an import and export demand function using quarterly time series data from 1983-2001. The variables they include are nominal income, prices paid by importers, prices of domestically produced goods, relative prices, and the actual tariff preference. The main result they find is that US import demand for Mexican goods is responsive to the changes in tariff preferences that were granted from the US to Mexico, in the form that a one percent decrease in US tariffs for Mexico corresponds with a 5.6 percent increase in US import demand for Mexican goods.
Tornell, Westermann and Martìnez (2004) analyse the Mexican growth performance of GDP and exports and ask to what extent fluctuations in the US economy account for that. In order to do this, they estimate a bivariate vector autoregression (VAR) model on US imports and US manufacturing, using quarterly data from 1988-2003. They conduct their analysis using growth rates. Their findings indicate that unexpected changes in Mexico’s export growth mainly result from shocks to its own economy (60% of the forecast error variance), whereas shocks to US imports account for 40% of the forecast error variance. Furthermore, Tornell et al. deduct from their analysis that the US economy does not fully account for the significant Mexican export increase in 1995 (after the FTA), nor for the recent fall in exports. They also compare the growth rate of Mexican exports with the one of US imports, where they find that 30% of the variance of Mexico’s total exports is explained by US imports.
Further studies on NAFTA were conducted by Ibarra-Yunes (2003) and Kondonassis and Malliaris (1996), which both focus on the theoretical part of regional economic integration and how multiple FTAs should be administered. Ramirez (2003), and Sargent and Matthews (2001) place emphasis on the effect NAFTA had on Mexico’s economy. Krueger (2000) conducts a study on NAFTAs general trade effects. Both Panagariya (1999) and Rivera-Batiz and Oliva (2003) used NAFTA as a case study for analyzing rationales for and welfare effects of regional preferential trade agreements in their books. Coffey et al. (1999) assessed NAFTAs effect for each of the member countries.
Studies on the EU-Mexico FTA, on the other hand, are not as richly found as for NAFTA, and only very few actually apply some kind of empirical model to assess the effects of the agreement.
Busse, Huth, and Koopmann (2000) provide an ex-ante analysis of trade effects through the agreement for different commodities, based on a model developed by Clague. The results of their estimation show that positive trade effects can be expected especially for agricultural and semi-finished products (from Mexico’s point of view), and for finished products like “Machinery and Transport Equipment” in case of the EU. Furthermore, the EU is expected to gain more from the FTA. In Mexico, trade creation is supposed to dominate trade diversion, even though trade will be diverted to some extend from the USA to Europe.
Slootmaeker’s (2004) findings for import and export flows between Mexico and the EU are based on a gravity model framework. She uses panel data over the period 1980-2003, which indicate that the effect of the FTA on exports is not statistically significant. The most important factor in determining exports and imports between Mexico, the EU and their trading partners, is US GDP. According to her results, a 1% increase in US GDP is related to a 3.8% increase in trade between the above named partners.
Further research was undertaken by Weisser (2003), Gehring and Kleider (2004), who try to analyze the effects the FTA had on trade relations between the EU and Mexico, as well as the future perspectives by using descriptive statistics only. Reiter (2003) intensely compares the contents of NAFTA and the EU-Mexico FTA.
As we can see from the discussion above, previous research has been focusing either on NAFTA or the EU-Mexico FTA. The authors who actually compare the agreements do not apply any econometric analysis, but rather base their work on showing the effects graphically. In this report, the method to analyse trade relations between the blocs will also include elements of descriptive statistics, but an econometric approach in form of estimating export and import demand functions will be taken as well, using time series methodology. Time series modelling is widely used in the literature in order to test whether the chosen variables are relevant in explaining trade between the observed parties. In this paper, one equation for export demand will be estimated for each FTA, using time series data taking into account the effects of the particular agreement. Both equations will contain more or less the same variables, so that a comparison is possible. Furthermore, one regression equation for import demand for each FTA will be estimated and then also be applied on the relevant accord. The new aspect to be found in this thesis is the actual calculation of transport costs, taking the cif (cost, insurance, freight) trade values and subtracting the fob (free on board) values. In addition to this, the application of basically the same regression equation to both accords has, to the author’s knowledge, not been performed in previous studies.
3 Contents of both Agreements
Taking first a look at the North American Free Trade Agreement, the Preamble points out that Canada, Mexico and the USA wanted to strengthen their friendship and cooperation among their nations, together with promoting a sustainable development. More defined objectives can be found in the articles 101 and 102 of NAFTA, stating that
(a) trade barriers shall be eliminated in order to facilitate the movement of goods and services,
(b) conditions of fair competition in the free trade area shall be promoted,
(c) investment opportunities in the territories of the Parties shall be increased, and
(d) intellectual property rights shall be protected.
The first talks in order to achieve these goals started between the USA and Mexico in 1990, and Canada joined the negotiations in 1991. One year later, President Bush (USA), President Salinas (Mexico), and Prime Minister Mulroney (Canada) signed the agreement. NAFTA officially entered into force on January 1, 1994.
The agreement was designed to eliminate most of the barriers to trade, so did US tariffs on Mexican products fall immediately from 3.3% average to 1.1%, Canadian tariffs came down from 2.4% to 0.9%, and the Mexican average tariff fell from 11% to 5%. About two thirds of the industrial and agricultural exports became instantaneously tariff free. It is worth noting that mentionable tariff reductions on Mexico’s side already took place in advance to NAFTA, sliding down from an average 25% to 13%, according to McDaniel and Agama (2003). By 2001, US average tariffs for imports from Mexico were 0.52%, for Canada 0.45% (McDaniel and Agama, 2003). A period of 10 to 15 years before the total tariff phase out was agreed on for sensitive products coming from the US textile and apparel industry, as well as for Mexico’s subsistence agricultural products (Ramirez 2003). However, all non-tariff barriers to agricultural trade were immediately eliminated. In the case of the USA and Canada no substantial change in trade relations or tariffs is to be mentioned, since these two countries have had an FTA since 1989 (CUSTA).
The rules of origin introduced in the agreement pinpoint that goods have to be grown, harvested or wholly produced in the free trade area in order to receive preferential duties. When products are originated outside of NAFTA, they have to be significantly transformed before they receive preferential duties when sent from one member country to one of the others.
Authors such as Tornell, Westermann, and Martinez (2004) argue that the reduction of trade barriers would not be the main benefit of NAFTA, but that the codification of the underlying trade rules was the most important ingredient of the agreement. This made the “silent integration” that was already taking place before 1994 official, and with that contributed to the locking in of reforms, enhancing Mexico’s credibility as a country for possible investments. The NAFTA investment provision also is an important means to mention in that context (McDaniel and Agama, 2003).
Besides the pure economical measures that were taken, the NAFTA Secretariat was established and empowered with Dispute Settlement Provisions. Side agreements on labour were also negotiated in course of the establishment of the accord. Furthermore, NAFTA is the first agreement where environmental issues are addressed, and the first between two highly industrialised countries and a developing one (Ramirez 2003).
3.2 The EU-Mexico FTA
"…this is a historic moment. It marks the beginning of a new era in Europe's relations with Mexico. We are confident that economic operators in Europe and Mexico will be quick to breathe life into this agreement and to improve our bilateral trading relationship significantly" (Commissioner Lamy, July 1, 2000).
Quoting Commissioner Lamy while welcoming the entering into force of the free trade agreement on July 1st, 2000, might give a good impression of the general expectations the EU put into the FTA. The official text of the agreement states the following objectives:
(a) the progressive and reciprocal liberalisation of trade in goods, in conformity with Article XXIV of GATT 1994;
(b) opening the agreed government procurement markets of the parties;
(c) establishing a cooperation mechanism in the field of competition;
(d) setting up a consultation mechanism in respect of intellectual property matter; and
(e) establishing a dispute settlement mechanism.
Not mentioned in the official text above is the preferential market access concerning the trade in services under Article V of GATS, which entered into force March 1st, 2001. The reason for this is that the issue on services had to wait for ratification of the EU-Mexico Economic Partnership, Political Coordination and Cooperation Agreement signed in December 1997. Article V of GATS has similar provisions for preferential trading in services as Article XXIV has for goods.
The innovative part of this “unnatural agreement” can be found in the fact that it is the first FTA between a Latin American country and the EU. It does not only include the lowering or abolishing of tariff-barriers for goods, but also facilitates the above named trade in services and public procurement, and further addresses intellectual property rights and competition issues.
According to the European Commission, the FTA will liberalize over 96% of the trade between the EU and Mexico till 2007. Right after the coming into force of the agreement, 82% of Mexico’s industrial export could enter the EU without any duties, and the rest was liberalized by January 1st, 2003. From the EU’s point of view, there is to say that 50% of industrial exports could enter Mexico free of duties after July 1st, 2000, and the rest has a maximum tariff level of 5% which shall be eliminated by either 2005 or 2007, depending on the good. Substantial liberalizations for agricultural and fishery products are also part of the agreement, so will 42% of the EU exports and 80% of the Mexican exports be duty free till 2010, reflecting the sensitivity of these commodities just as in NAFTA. For agriculture exports from Mexico, customs duties for coffee, cocoa beans, chickpeas, tequila, beer, mango, papayas and guavas were completely abolished. Mexico grants preferential tariff quotas to the EU when it comes to sugar, meat and dairy products, corn, bananas and orange juice, which is unique in this FTA. In contrast to NAFTA, the EU-Mexico FTA acknowledges the different development levels the two partners have, resulting in asymmetrical tariff reductions in favour of Mexico (Weisser, 2003). Furthermore, the transition periods for tariff phase out in this accord are shorter than in NAFTA.
The rules of origin in this agreement are similar to the ones in NAFTA, generally calling for goods being either originated in the partner countries or having been sufficiently changed. The importance of these rules becomes apparent when considering the fact that they actually occupy 177 pages in the official text of the agreement.
Concerning the trade in services, liberalization will take place over a period of ten years; excluded are audiovisual services, cabotage and air transport. The EU gained almost the same preferential access to Mexico’s public procurement market as the NAFTA members, in exchange for giving Mexico similar treatment in the EU under the plurilateral WTO Procurement Agreement, which Mexico did not sign (EU Commission, 2000).
4 Theoretical Framework of FTAs
Before stating the actual effects of the two accords on trade, the general rationales why a country would engage in free trade agreements shall be commented on. It is important to note that in this section the incentives to form an FTA are subject to explanation, rather than the incentives or benefits of multilateral free trade. As Bhagwati (1998) states, FTAs are not the same as free trade, since they provide free trade (eventually when finished) only to their member countries, but keep protection against non-members. For this reason, FTAs are often considered to be the second best solution, but, nevertheless, rather popular today.
A free trade area, per definition, can be established when the countries that are going to be members of the area remove the barriers to trade among them (completely or to a large extent), but keep barriers for non-members. There is no such thing as a common external tariff, but each country has the right to choose its individual restrictions to non-members (Sloman 2000). If the member countries of the free trade area had common external tariffs and quotas, the free trade area would, per definition, be a customs union, such as the EU. Art. XXIV of GATT and Art. V of GATS provides the basis for these two forms of regionalism.
4.1 Static Analysis
The static analysis deals with the questions whether or not preferential tariff reductions are welfare improving. If it was, this would be a reason for a country to form an FTA, since it wants to increase its welfare. The method to assess the welfare gains was first established by Viner (1950), who divided the effect of customs union (this also works for FTAs) into trade creating and trade diverting effects. If a FTA is trade creating, then a high cost producer of a good will be replaced by a low costs producer within the union after the tariff cut. Trade diversion occurs when a low cost outside producer is replaced by a high cost inside producer, who is only cheaper because of the preferential market access (Sloman, 2000) . A graphical demonstration facilitates the two possibilities.
illustration not visible in this excerpt
Let Smex be the supply curve of Mexico for the good X, and Dmex the demand curve (Figure 3.1). P1 is the price for good X with tariffs, so the price Mexico would have to pay in the absence of an FTA. At P1, Mexico produces quantity Q2 of good X and imports Q1-Q2 from, for example, the US. Imagine now these two establish a FTA, and let all tariffs be eliminated. Then the price would fall to P2, namely the price without tariff, and Mexico would produce Q4 and import Q3-Q4. A reduction in the price leads to a consumer gain of the areas 1&2&3&4, but also a domestic producer loss of area 1 and a loss of tariff revenues for the Mexican government of area 3. Nevertheless, there is still a net gain to be found of areas 2&4, due to trade creation.
In the case of trade diversion (Figure 3.2), an overall welfare loss can be the case. Let us assume that Mexico imports good X from Japan in the absence of an FTA, since Japan is the most efficient producer of good X in the world. Mexico then pays P1 for good X, consumes at Q2 and produces Q1, so Mexico imports Q2 - Q1. After joining NAFTA, let us assume that product X becomes completely tariff free when Mexico imports it from NAFTA. Mexico then has to pay only P2, consumes at Q4 and produces Q3 domestically. Trade diversion has happened, since the tariff free Japan price of good X, namely P3, would be cheaper than the tariff free NAFTA price, so consumption has switched to a higher cost producer. The consumer surplus still rises by areas 1&2&3&4, but Mexican producer surplus falls by area 1, and the government loses tariff revenue 3+5. In the end, we are left with a net gain of 1&2&3&4 minus areas 1&3&5, so basically (2&4)-5. If area 5 now is bigger than the areas 2&4, there is a net loss to Mexico due to trade diversion, even though consumers might still gain. Trade creation, therefore, is favourable, trade diversion might bring negative effects.
4.2 Dynamic Analysis
When turning to the dynamic analyses, we focus on the effect that will come with time, in contrast to the static analysis. Starting out with the most obvious reason, there is to say that the increase in market size that will result from an FTA can be beneficial for the member countries. A promotion in trade is expected, and coming along with this is more competition, which is likely to be an incentive for firms to become more efficient, and thus benefit the consumer. A larger market size is also considered to raise income per capita and enhance economic growth in general. The countries might be able to make better use of their internal economies of scale, since they now face a bigger market, and therefore more consumers (Rivera-Batiz and Oliva, 2003). As the market size increases, therefore, the costs per unit of output should fall as the scale of production increases. The scale of production is likely to increase because of the expected trade promotion through FTAs. This argument is especially valid when talking about small countries, since they encounter a relatively bigger market in comparison to the large countries that take part in the agreement. Therefore, in the case of Mexico, this can be seen as one of the reasons why it entered NAFTA and the accord with the EU.
Furthermore, through economic integration the spread of technology could be faster among the member countries, since trade relations intensify. The attraction of more foreign direct investment (FDI) is also often an objective that can lead countries into engaging in a FTA. Especially the last argument was one of Mexico’s main objectives when entering both accords.
A possible motivation for an FTA is also to be found when looking at external economies of scale. So might increased trade between the members lead to the improvement of highways, railroad and the like (Sloman, 2000). This, in turn, could lead to a lowering of transportation costs, which makes a preferential trade agreement among countries an interesting option and is likely to promote trade (Hummel, 1999a). Economic geography, therefore, does play a role in preferential agreements.
Along with the argument concerning proximity and trade, it should be mentioned that Panagariya (1999) opposes the transport costs argument when it comes to the aspect of welfare. He says that there is not any theoretical support to the idea that countries located far from each other, therefore having higher transport costs in trade, must gain less from trade.
4.3 Political Economy
Authors as Rivera-Batiz (2003), Panagariya and Findlay (1996), Baldwin (1996) and Ibarra-Yunes (2003) have drawn the attention from the pure economic rationales to enter an FTA to a political point of view. They say that the trade policies of a country are not purely based on welfare-maximisation, but are the result of the government’s interactions with interest groups. These lobbies do not seek the overall welfare improvement of the country, but only of their special area. Since the self-interest of the politicians is to get re-elected, they might follow the lobbies’ demands.
Cooperation in managing shared conflicts or large projects also falls under political economy reasons for engaging into a FTA. Those objectives are not as clearly present in NAFTA as in the EU-Mexico FTA, where financial cooperation regarding lagging regions and sectors and the sharing of administrative competence in fiscal, judicial and financial support to SMEs were discussed (Ibarra-Yunes, 2003).
Countries may see the advantage of improving their bargaining power with respect to the rest of the world, and therefore they maybe obtain better terms of trade, defined as the average price of exports divided by the average price of imports (Sloman, 2000). Additionally, the World Bank (2001) mentions the issue of “being noticed” in the world economy as another reason to participate in FTAs.
Rivera-Batiz and Oliva (2003) state that the formation of regional trade agreements can also be due to slow moving multilateral negotiations. Bilateral negotiations are expected to be faster, less costly, and less complex. This was one of the motivations for the USA to enter NAFTA and leave the path of multilateralism, since at that time further GATT rounds were more or less boycotted by the EC and regionalism seemed to be the way to go (Bhagwati et al., 1998).
Moreover, countries enter in FTAs in order to be insured against a shift to protectionist measures from their trading partner (World Bank, 2001). Both in NAFTA and the EU-Mexico FTA dispute settlement mechanisms are included, such as to provide help in the case of one country applying procedures that are not allowed under the accords. Especially in the case of NAFTA, this insurance against future protectionist measures is seen as another of Mexico’s major motivations to enter, even more important than the expected tariff reductions by the USA and Canada. These were bigger on Mexico’s side anyway, since the USA and Canada had lower tariff barriers to begin with. Along with the argument, it should be mentioned that an FTA does not only mean insurance in case the partner countries shift their trade policies, but it can also be viewed at as a possibility to promote and to lock-in reforms in any of the participating countries. Ibarra-Yunez (2003) mentions that preconditions of an FTA can include commitments by partner governments to maintain free trade and open capitalism, as was the case in NAFTA. Once a preferential trade agreement is concluded and liberalizing measures have been taken, it is not easy for future governments to reverse those settings.
 The complete legal text can be found on the NAFTA Secretariat homepage under www.nafta-sec-alena.org
 This summarized version of the rules of origin can be obtained from the United States Department of Agriculture under www.fas.usda.gov
 The complete legal text can be found under the external trade section of the EU homepage www.europa.eu.int
 Meaning an FTA in which members are from different continents, according to Frankel, Stein and Wei (1994), who quote Krugman (1991).
 Note that different figures exist. The European Commission text mentions 50%, Busse et al. talk about 52%, and Slootmaekers says even 60% (it does not become clear if she talks only about industrial goods, though). The author will follow the official EU statement.
 The rules of origins are listed in Annex III of the official agreement text and can be obtained from the external trade section of the EU homepage www.europa.eu.int
 Both Art. XXIV and V can be obtained from the WTO’s legal text section under www.wto.org