Introduction

Trade Agreements (TAS) stands as crucial tools for enhancing trade flows among nations or regions. These agreements, often referred to as "agreements between two or more countries that establish, through tariff reduction schedules, the conditions under which such tariff reductions can be applied, and which commonly include additional agreements concerning trade facilitation between the parties" (Van de Heetkamp and Tusveld, 2011) play a pivotal role in shaping global commerce. Consequently, it's reasonable to expect that Tas, such as Free Trade Agreements (FTAS), would foster increased trade flows among signatory nations.

In recent years, the employment of TAS has surged as a trade integration strategy in Latin American economies. Chile, for instance, currently boasts 26 active Tas spanning 65 markets globally (Ministry of Foreign Affairs of Chile, 2018). Notably, among the plethora of TAS to which Chile is a party, there are three inked with East Asia countries: the FTAS with South Korea and China in 2004 and 2006, respectively, and the Chile-Japan Economic Cooperation Agreement (ECA) in 2007 (SICE OAS, 2018). Chile stands out as a prime example of integration within Latin America, particularly with East Asia. An in-depth study of the Chilean case is imperative as it could offer invaluable insights and empirical evidence on the advantages of forging ties with this region for other Latin American countries.

Moreover, given East Asia's robust economic performance and its dynamism nature in establishing trade relations (Kawai and Wignaraja, 2010), is plausible that these specific TAS have yielded more substantial benefits for Chilean exports compared to others to which Chile is a party. This suggests that these agreements may hold significant potential for bolstering trade and economic growth within the region. Indications supporting this notion are discerned from the significant role that China, Japan and South Korea have assumed as destinations for Chilean exports (refer to Figure 1).

Figure 1
East Asia as an Export Destination for Chilean and Latin American Exports worldwide.
Source: ITC Trade Map (2018). Calculations conducted by the authors.

Figure 1 also illustrates the pattern of East Asia's emergence as a primary destination for Latin American exports, considering the export destination of Colombia, Brazil, Argentina, Peru, Ecuador, Venezuela, Mexico and Uruguay under the umbrella of "Latin America". While the growth of East Asia as an export destination appears to be a regional trend, it's evident that East Asia's significance as an export destination is disproportionately higher for Chile compared to the rest of the region. It's worth noting that within Latin America, only five TAS with East Asian countries exist, which are the focus of this study: MexicoJapan (2005), Peru-China (2010), Peru-South Korea (2011), Peru-Japan (2012), and Colombia-South Korea (2016), all signed within the period covered by this study (2001-2016).

Given all that has been proposed above, the question arises: Has the entry into force of trade agreements with East Asia had any significant and differentiated impact on the trend of Chile's export flows between 2001 and 2016? This question is addressed by estimating a gravity model for the period 2001 - 2016 using a rich panel of country-wide variables, while controlling for macroeconomic key variables, as well as cultural and geographical factors. This paper is structured as follows: Section 2 presents a review of existing theoretical and empirical literature. Section 3 outlines the empirical methodology employed. Finally, section 5 presents the results obtained by the gravity model, and Section 6 provides the conclusion.

Literature Review

Economic theory has long been dedicated to understanding trade dynamics among nations and the factors shaping them. Since its inception as a science, David Ricardo, in Principles of Political Economy established the theory of comparative advantage, which correctly implies the propensity to trade between two nations results from their interactions and specializations. Ricardo posited that: i) Specialization is a fundamental for the growth, as it enhances productivity, leading to efficient production; and ii) Given the productive gains from specialization, nations will find it advantageous to exchange goods in which they specialize for others, making trade mutually beneficial.

In recent decades, global economies have shifted from protectionist systems to trade specialization and openness, fostering an environment conducive to trade and productivity. Trade Agreements (TAS) have played a crucial role in facilitating this transition, opening national markets, fostering price and quality competition, and promoting innovations.

Modern theorists argue that promoting free trade relations, signing TAS, between nations is beneficial, as it prevents welfare and efficiency losses associated with protectionist policies (Krugman and Obstfeld, 1997). Numerous empirical studies corroborate a positive relationship between reciprocal TAS and increased trade between participating nations (Cipollina, Salvatici, and Luca, 2007). However, the impact of TAS, on trade size varies widely, ranging from 15% to 100% growth (DeRosa, 2007). Some studies suggest that the effects of TAS are not uniform, with mixed results regarding their impact on trade size (Kepaptsoglou, Kar-laftis, and Tsamboulas, 2010).

In Latin America, agreements like the Latin America Free Trade Agreement (LAFTA or CAN), have shown minimal effects on trade generation and diversification (Kepaptsoglou, Karlaftis, and Tsamboulas, 2010). For Chile, the trade agreement with the United States since August 2003 has had positive impacts on specific sectors, but its overall effect on Chile's exports to the United State is not universal (Jean, 2012). Conversely, the Chile-Canada FTA has significantly boosted trade between the two nations (FAIT Canada, 2013).

Finally, in a study on the effects that entry into force of the Trans-Pacific Partnership Agreement (TPP) could have on various economies in termes of the growth of their trade volumes, was found through gravity model estimations that increases in both the Chilean gross domestic product and that of its trading partners have a positive and significant impact on bilateral trade flow growth. Similarly, the border and language dummy variables also indicate a positive effect. On the other hand, the increase in distance has a negative and significant effect. Finally, the dummy variables that measure the impact that the presence of the TPP and the Trans-Pacific Strategic Economic Partnership Agreement (P4), an economic partnership agreement initiated by Chile, Brunei, New Zealand and Singapore in 2005, which years later would be expanded and renamed TPP, show that the estimates were not significant in both cases (Raffo, et al, 2018).

Despite the existence of this empirical evidence, there exists a lack of studies that analyze in depth and individually the effect that the current FTAS between Chile and Japan, South Korea and China have had on Chile's exports and evaluate the existence of a differentiated effect. This work aims to fill that gap in the literature and provide evidence impact on bilateral trade flows growth used by other Latin American countries to promote the signing of trade agreements with East Asia.

Empirical Strategy

The assessment of trade agreements' effectiveness in promoting exports, as well as the impacts ofvarious determinants of international trade, is typically conducted using augmented gravity models. The foundational gravity model (Tinbergen, 1962; Tinbergen 1966) posits significant relationships between countries' distance and the volume of trade between them. Building upon this, an extensive empirical and theoretical body of work has emerged, aiming to address trade policy questions through different iterations of the Tinbergen gravity model. This framework, consistent with orthodox theory, suggests that distance serves as a proxy for trade costs, implying that greater distances correlate with lower trade volumes. Moreover, novel approaches to measuring distance between countries have been proposed, incorporating not only geographical factor, but also cultural, economic and historical factors considerations into the costs associated with distance (Mejía and Hassan, 2017).

To validate these hypotheses, a gravity model will be employed, utilizing the following model specication¹:

where:

• i represents Chile, j refers to its trading partners, and t represents time in years.

• Exports _ijt refers to Chilean exports to each of its partners in period t.

• GDPi_t-1 and GDPj_t-1 refer to the gross domestic products of Chile and its trading partners during period t.

• Distance _ii represents the geographical distance between Chile and its partners.

• FTADummy_it and OADummy_ij controls for the existence of FTAS and other types of comercial agreements.

• BRDDummy_ij and LANDummy_ij, are dummies controlling for language and borders, respectively.

• CHNDummyij, jPNDummyij, and KORDummyij, are interaction variables that isolate the effect of a commercial agreement between Chile and China, Japan and Korea.

The gravity model should not be estimated in its log-linearized form, which involves taking the natural logarithm of variables on both sides of the equation. This method fails when the independent variable equals 0 and can introduce biases when linearizing the model for Ordinary Least Squares (OLS) estimation. This is because it diminishes the scale of differences in the observations of the endogenous variable, potentially rendering the estimator inconsistent. Another common issue in OLS estimation of the gravity equation is the assumption of homoscedasticity in the error term, which is often due to non-random differences between countries affecting their trade relations (Burguer et al., 2009). Consequently, heteroscedasticity in the error term can lead to an inefficient estimator (Santos and Teneyro, 2006).

To address these challenges, alternative techniques for estimating the gravity model are necessary. One such technique, increasingly employed, is the Poisson Pseudo-Maximum Verosmilitude (PPML) estimator (Kepastoglu et al., 2010). In this method, all explanatory variables are exponentiated. While Poisson models traditionally focused on count variables, the Poisson estimator can also be applied to continuous variables with values greater than 0, which aligns a case with the nature of trade flows, particularly exports in the context of this paper.

The gravity model is estimated using a panel data comprising annual net exports from Chile to each of its trading partners from 2001 to 2017. Specifically, this includes countries to which Chile exported during the period. Additionally, data on relevant variables of interest and control are incorporated. Estimation is conducted at the aggregate level, covering total exports from Chile to each partner, with separate estimations for agricultural and manufacturing exports. This disaggregation is facilitated by the system's harmonized classification.

Data series for endogenous variables, such as export trade balances at the aggregate level, and for agricultural and manufacturing sectors, were sourced from ITC Trade Map (2018). For the agricultural sector, data on Chile's exports to each partner were aggregated for the first 24 chapters of the harmonized system. For manufacturing sector, Chile's exports were aggregated for 100 manufactured products explicitly classified under manufacturing. It's worth noting that the study focuses on a representative portion of manufactured products due to technical constraints preventing a detailed examination of all items. Gross domestic product data for each country were sourced from the World Bank's "World Development Indicators" database (2018). Information for creating dummy variables controlling for Free Trade Agreements and interaction variables with China, Japan, and South Korea was obtained from the SICE OAS portal. Data for the border dummy variable were sourced from the website of the Ministry of Foreign Affairs of Chile (2018), while information for the language dummy variable (Spanish) was obtained from the website of the Association of Spanish Language Academies. Dummies isolating the effects of FTAS between Chile and East Asian countries were derived from the TLC Dummy variable.

Results

The Hausman test suggests that the model estimations for total exports and agricultural exports should utilize fixed effects to control for unobserved time-invariant heterogeneity across countries. Conversely, estimations for manufacturing exports should employ random effects to account for both observed and unobserved time-invariant heterogeneity across countries. To capture a comprehensive understanding of the potential effects of FTAS on Chilean exports at aggregate, agricultural, and manufacturing levels, and recognizing that fixed effects estimations eliminate effects that remain constant over time (such as border variables, language, and geographical distance), each estimation was conducted using both fixed and random effects.

For each of the three variables under examination, two types of estimations were conducted: the first type included only dummy variables indicating the presence of FTAS and other agreements, along with controls; the second type incorporated variables isolating the impact of commercial agreements with China, Japan, and South Korea.

In examining the variables of interest, an assessment of the estimates excluding those isolating relations with Japan, China, and South Korea reveals that FTAS exhibit a positive impact on total exports and agricultural exports, consistent with prior findings (Chumacero, 2004; Fuenzalida, 2016). However, for manufacturing exports, the estimation suggests a negative relationship with the F TA variable (see Table 1), aligning with qualitative analyses of Chilean exports behavior. These analyses highlight the potential effect of trade integration policies, noting that "the value added of the export basket remains low and sophistication low" (Dingemans, 2016), characteristics prevalent in manufacturing exports.

Table 1

Results

Note: Standard errors in parentheses - *** p<0.01, ** p<0.05, * p<0.1

Conversely, other types of trade agreements, such as Economic Cooperation Agreements, Partial Scope Agreements, Economic Integration Agreements, and Trade Protocols, demonstrate a similar pattern: a significant positive impact on total and agricultural exports, but a negative one on manufacturing exports. The unexpected negative impact of TAS on manufacturing exports could theoretically be supported by the idea that short-range economies of scale, which are not well- developed, may be displaced by long-range economies of scale upon entering competition (Krugman, 1979). This theoretical framework is relevant to this specific case, considering that the manufacturing sector is not Chile's primary comparative advantage, given its focused of mineral exports.

Turning to the analysis of variables isolating the effect of trade agreements between Chile and China, Japan, and Korea, it becomes evident that, for total exports, the impact of all trade agreements signed by Chile. This finding corroborates previous research indicating that CAS do indeed stimulate trade among their members (Baier and Bergstrand, 2007; Cipollina and Salvatici, 2010).

As for agricultural exports, the results are mixed: the Chile-Korea FTA stands out with an above-average effect on exports, whereas the Chile-China FTA and the Chile-Japan EPA exhibit below-average effects. Lastly, the estimated effects of the CAS with East Asia on manufacturing exports indicate a greater impact than the average effect observed for the three export destination countries under scrutiny.

Conclusions

East Asia stands as the primary destination for Chilean exports globally, absorbing 44% of Chilean exports in 2016. Moreover, there has been a consistent growth in Chilean exports to the region since the early 20th century. Chile's trade policy is oriented towards promoting Trade Agreements (TAS) not only with East Asia countries but also with nations and economic blocs worldwide, exemplified by the signing of agreements with South Korea (2004), China (2006), and Japan (2007).

The objective of this study was to empirically assess whether the FTAS signed by Chile with China, Japan, and South Korea have had a distinctive impact on Chilean exports. for this purpose, such quantification was carried out on total, agricultural and manufacturing exports, based on the use of a gravity model. The findings suggest that, contrary to the average effect, the impact of these FTAS is greater, except for the Chile-Japan Economic Cooperation Agreement (ECA) and the Chile-China FTA in the case of agricultural exports. In essence, the results indicate that Chile's FTAS with East Asian countries have exerted an above-average influence on the growth of Chile's export balances with each respective party. Consistent with prior studies, it is noted that when trade balances are generally low, such as in the case of manufacturing exports considered here, the Poisson Pseudo Maximum Likelihood (PPML) estimator tends to overestimate the effect of the variables in question (Gomez, 2013; Dueñas and Fagiolo, 2011).

Given the structural disparities among economies in the region, the findings of this study may not readily generalizable to other countries. Thus, it would be inappropriate to extrapolate the effects in this study for Chile to other nations in Southeast Asia. However, establish that the effect of the entry into force of trade agreements with Southeast Asian countries would have an effect that can be extrapolated in the direction or magnitude of that estimated in this study for the case of Chile. However, further research could explore the significance of East Asia as a platform for trade expansion not only for Chile but also for Colombia, Mexico, and Peru - countries in the region that have also engaged in CA signings with South Korea, China and Japan. Additionally, subsequent studies could investigate the influence of CAS with East Asia on the diversification of the Chilean export products.

References

Anderson, James (1979). "A Theoretical Foundation for the Gravity Equation," American Economic Review 69, 106-116.

Baier, S. and Bergstrand, Jeffrey (2007). Do free trade agreements actually increase members' international trade? Journal of International Economics, 71. 72-95. https://doi.org/10.1016/j.jinteco.2006.02.005

Banco Mundial. (1993). The East Asian Miracle: Economic Growth and Public Policy. World Bank - Oxford University.

Burger, M., van Oort, F. y Linders, Gert-Jan (2009). On the Specification of the Gravity Model of Trade: Zeros, Excess Zeros and Zero-Inflated Estimation. ERIM Report Series - Erasmus Research Institute of Management (ERIM). https://doi.org/10.1080/17421770902834327

Christopher Stevens, M. I. (2015). The impact of free trade agreements between developed and developing countries on economic development in developing countries. London: Overseas Development Institute.

Cipollina, M. and Salvatici, L. (2007). Reciprocal trade agreements in gravity models: a meta-analysis. Tra-deAG Project. https://doi.org/10.2139/ssrn.950929

Corvalán, Angélica. (2014). Efectos en la economía chilena de la firma de Acuerdos Comerciales. Univerisdad de Chile - Facultad de Economía y negocios.

DeRosa, D. (2007). Gravity model anaysis. Peterson Institute for International Economics. Washington DC.

Dingemans, A. (2016). El fin de una etapa exitosa. Los TLC en la estrategia comercial de Chile. Revista de Economía Institucional, 18 (34), 151-172. https://doi.org/10.18601/01245996.v18n34.10

Dirección General de Relaciones Económicas Internacionales - DIRECON, (2018). https://www.direcon.gob.cl/

Dueñas, M and Fagiolo, Giorgio (2013). Modeling the international treade- network: A Gravity Approach. Journal of economic interaction and coordination, 8. Issue 1, 155-178. https://doi.org/10.1007/s11403-013-0108-y

Fally, T., Structural gravity and fixed effects, J. Int. Econ. (2015), http://dx.doi.org/10.1016/j.jinteco.2015.05.005.

Foreign Affairs and International Trade - FAIT. (2013). The economic impact of the Canada Chile Free Trade Agreement. Office of the Chief Economist: Foreign Affairs and International Trade, Quebec.

Fuenzalida, D., Valenzulea, B., and Corvalán, A. (2018). La facilitación comercial y sus efectos en el comercio bilateral chileno de 2006 a 2014. Revista de la CEPAL N° 124, 173-191. https://doi.org/10.18356/630ad7fa-es

Herrera, E. G. (2011). Comparing alternative methods to estimate gravity models of bilateral trade. European Trade Studing Group.

Hertel, Thomas y Maros Ivanic (2005), "Agricultural trade policy and poverty in developing countries", ESA Working Paper, Roma, FAO.

Jean, S. (2012). Assesing the impact fo the EU-Chile FTA on international trade. *Working paper, GTAP Resource #3925

Kabir M., Salim, R., and Al-Mawali, N. (2017). The gravity model and trade flows: Recent developments in econometric modeling and empirical evifence. Economic Analysis and Policy, 56, 60-71. https://doi.org/10.1016/j.eap.2017.08.005

Kawai, M., and Wignaraja, G. (2010). Free Trade Agreements in East Asia: A Way Toward Liberalization? ADB Briefs No. 1.https://doi.org/10.4337/9780857930415

Kepaptsoglou, K., Karlaftis, M. G., and Tsamboulas, D. (2010). The Gravity Model Specification for Modeling International Trade Flows and Free Trade Agreement Effects: A 10-Year Review of Empirical Studies. The Open Economics Journal, 1-13. https://doi.org/10.2174/1874919401003010001

Kohl, T. , Brakman, S. y Garretsen, H. (2016), Do Trade Agreements Stimulate International Trade Differently? Evidence from 296 Trade Agreements. World Econ, 39: 97-131. https://doi.org/10.1111/twec.12272

Krugman, P., and Obstfeld, M. (1997). International Economics: Theory and Policy. AddisonWesley.

Linnemann, H. (1966), An Econometric Study of International Trade flows, North-Holland, Amsterdam.

Martin, Will and Pham, Chong (2008). Estimating the Gravity Equation when Zero Trade Flows Are Frequent. MPRA Paper, 9453.

Mejía, J., and Hassan, A. (2017). Proposing a new measure of distance in the gravitty setting: Evidence from latin America. Economía Aplicada, 135-198.https://doi.org/10.11606/1413-8050/ea154363

Okabe, M. (2015). Impact of Free Trade Agreements on Trade in East Asia. Discussion Paper Series. DP-2015-01, Economic Research Institute for ASEAN and East Asia (ERIA).

Piermartini, R. y R. Teh (2005), "Demystifying modelling methods for trade policy", Discussion Paper, N° 10, World Trade Organization.

Prehn, Sören and Brümmer, Bernhard and Glauben, Thomas. (2015). Gravity Model Estimation: Fixed Effects vs. Random Intercept Poisson Pseudo Maximum Likelihood. https://doi.org/10.1080/13504851.2015.1105916

Raffo L., Hernández, E., and Díaz, V. (2018). La Alianza del Pacífico y los efectos potenciales del Acuerdo Transpacífico de Cooperación Económica sin Estados Unidos. Revista de Finanzaz y política económica, 65-87. https://doi.org/10.14718/revfinanzpolitecon.2018.10.1.3

Ricardo, D. (1817). Los Principios de la Economía Política y La Tributación. Londres.

Rómulo A. Chumacero, J. R. (2004). Chile>s Free Trade Agreements: How big is the deal? Santiago de Chile: Comisión Económica para América Latina y el Caribe.

Santos, S., and Teneyro, S. (2006). The Log of Gravity. The Review of Economics and Statistics , 641-685. https://doi.org/10.1162/rest.88.4.641

Santos, Silva and Teneyro, Silvana (2011). Further simulation evidence on the performance of the Poisson pseudo-maximum likelihood estimator. Economics Letters, 220-2022. https://doi.org/10.1016/j.econ-let.2011.05.008

Sattinger, M. (1978). Trade flows and differences between countries. Atlantic Economic Journal, 22-30. https://doi.org/10.1007/BF02308226

SICE - OAS, (2018). Foreign Trade Information System: http://www.sice.oas.org/.

Siliverstovs, B. and Schumacher, D. Empir Econ (2009) 36: 645. https://doi.org/10.1007/s00181-008-0217-y

Silva, C. and Florez, Cortés (2017). ¿El fin del super ciclo del cobre y de las reformas sociales en Chile? Textos and Contextos, 16-2. 479 - 494. https://doi.org/10.15448/1677-9509.2017.2.29516

Squartini, T., and Garlaschelli, D. (2013). Jan Tinbergen's legacy for economic networks: from the gravity model to quantum statistics. Instituut-Lorentz for Theoretical Physics, Leiden Institute of Physics, University of Leiden. Leiden https://doi.org/10.1007/978-3-319-00023-7_9

Tinbergen, Jan (1962). The World Economy. Suggestions for an International Economic Policy (New York: Twentieth Century Fund).

Toro, J., Alonso, G., Esguerra, P., Garrido, D., Iregui, A. M., Montes, E., and Ramírez, J. M. (2012). El impacto del Tratado de Libre Comercio con Estados Unidos en la balanza de pagos hasta 2010. Banco de La República de Colombia.

Trejos, A. (2009). Instrumentos para la evaluación de impacto de los acuerdos comerciales internacionales: aplicaciones para países pequeños en América Latina. CEPAL - Serie Estudios y perspectivas - México - No 110.

Van de Heetkamp, A., and Tusveld, R. (2011). Origin Managment: Rules of Origin in Free Trade Agreements. Springer. https://doi.org/10.1007/978-3-642-19808-3

Waugh, Michael E. 2010. "International Trade and Income Differences." American Economic Review, 100 (5): 2093-2124. https://doi.org/10.1257/aer.100.5.2093

Wilhelmy, M. (2010). La trayectoria de Chile frente a la región Asia-Pacifico. Estudios Internacionales, Instituto de Estudios Internacionales - Universidad de Chile, 125-141.

Zhang, Yun Ling and Shen, Minghui, The Status of East Asian Free Trade Agreements (May 17, 2011). ADBI Working Paper 282, Available at SSRN: https://ssrn.com/abstract=1844293 or http://dx.doi.org/10.2139/ssrn.1844293

Notas