Regional Trade Agreement and International Trade: An Empirical Study of ASEAN.

• Author: Wang, Di

1. Introduction

   Despite the recent attrition in global trade, the world has witnessed a wave of globalization efforts in the form of regional trade agreements. In particular, members of the Association of Southeast Asian Nations (ASEAN) have achieved remarkable economic and social development. Through establishing the ASEAN Free Trade Area (AFTA), the grouping has experienced a significant increase in trade among members as well as between members and non-member economies.

   There are extensive studies focusing on the economic benefits of the regional free trade agreement in ASEAN nations (Frankel and Wei 1996; Elliott and Ikemoto 2004; Bun, Klaassen and Tan 2009; Ismail and King 2013). Hapsari and Mangunsong (2006), for instance, study ten-year trade data of selected ASEAN countries after the signing of AFTA and find that the nations are better off from tariff reductions. Similarly, Thu and Van Trung (2015) confirm enhanced exports among ASEAN member countries and increased economic welfare due to the free trade agreement. The imports from non-member countries have increased as well. A study by Wong, Liew and Arip (2017) uses more recent data to investigate the effects of AFTA on bilateral manufacturing trade between ASEAN nations and their trading partners. They use a dummy variable to compare the trade flows between the AFTA and non-AFTA member nations. Their results support the argument that the trade creation effect is attributable to the AFTA agreement. However, a handful of studies also find contrary results. Sharma and Chua (2000) employ the gravity model using data from 1980 to 1995 of five ASEAN countries and reveal that intra-ASEAN trade has not increased by the integration agreement, but trade with members of the wider Asia-Pacific Economic Cooperation (APEC) grouping has risen. A more recent study by Akhmadi (2017) focuses on whether Indonesian membership in AFTA has increased exports of its agricultural goods. The author notes that, although the country's agricultural exports are positively correlated with some of the gravity model variables, Indonesia's membership has not brought about the expected export boost.

   In regional trade agreement studies, import tariffs are typically examined in great detail. There are, in fact, numerous studies on the impact of import tariffs on a country's imports. While some of these study how a particular industry is influenced by the tariff, others use aggregate trade data to examine how a change in import tariffs affects the import levels of the country that levies them. For example, Hapsari and Mangunsong (2006) record that a reduction in tariff by 1 per cent will contribute to imports from trade partners rising by 2.44 per cent. Using data from 2001 to 2003, Manchin and Pelkmans-Balaoing (2008) find that, for ASEAN member countries too, lower import tariffs promote imports. Calvo-Pardo, Freund and Ornelas (2011) document a negative relationship between imports and import tariffs in ASEAN over the period of 1993-2007. All these studies confirm the expectation that a country's imports are inversely related to its import tariffs. However, to the best of our knowledge, none of the existing studies analyses the effect of changes in import tariffs on a country's exports. Our study fills this gap.

   Using directional bilateral trade flow data during 2001-15, we analyse the actual impact of import tariffs and of ASEAN membership on trade flows, which is different from the previous studies that predict the probable impact. By using the directional trade flow data, we can also distinguish between the effects of import tariffs on imports versus exports. Compared to imports, exports are often perceived to be more directly beneficial to a country's employment and economic growth and therefore receive more attention from policymakers. Previous studies use the level or growth rate of bilateral trade to measure trade flows. In our study, however, we use the ratio of imports or exports with a trade partner to the country's GDP or the ratio of imports or exports with a trade partner to the country's total imports or exports with all trade partners to measure the trade flows. When a country's total trade or economy size increases, naturally, its imports or exports with trade partners will increase. By using the share of imports or exports, we can control the natural increase in imports or exports and measure the effect of tariffs and ASEAN membership on trade flows more accurately. It must be noted that this "share of imports" concept was also used in Calvo-Pardo, Freund and Ornelas (2011). In addition, to take into account the heteroscedasticity in trade data and to take advantage of the information contained in the zero trade flows, we follow Santos-Silva and Tenreyro (2006) and use the Poisson Pseudo-Maximum Likelihood (PPML) regression for our analysis.

   We find that ASEAN has been successful in increasing both imports and exports, not only for the member countries but also for the non-member states. This shows that the trade-creation effect dominates the trade-diversion effect for ASEAN. The grouping increases global trade and welfare, which is consistent with the findings of previous studies (e.g., Calvo-Pardo, Freund and Ornelas 2011; Lee and Shin 2006; Wong, Liew and Arip 2017). We also find that the increase in imports and exports with member countries is larger than with non-member countries. In other words, member states benefit the most from the trade agreement.

   In this paper, we also delve into the prospects of the Regional Comprehensive Economic Partnership (RCEP). As a relatively new economic and trade agreement in the Asian-Pacific region, RCEP lacks sufficient data for study, yet receives substantial attention from academia, business, and politics. Since RCEP includes all ten ASEAN signatories plus five other countries--Australia, China, Japan, New Zealand and South Korea--we attempt to use our findings to explore the future possibilities for RCEP.

   Our study contributes to the literature in four ways. First, in terms of empirical testing, we keep the literature up to date by using the most recent and available actual bilateral trade data on ASEAN members and their trade partners from 2001 to 2015. The majority of prior studies either use data from before 2008 (e.g., Calvo-Pardo, Freund and Ornelas 2011; Lee and Shin 2006) or conduct policy simulations to capture the impact of regional trade liberalization (e.g., Itakura 2014).

   Second, we reexamine the relationship between import tariffs and imports, and more importantly, obtain evidence on the linkage between import tariffs and exports. It is a common belief that import tariffs directly impact a country's imports, and we reconfirm the negative relationship between import tariffs and imports in the context of ASEAN. However, whether they have any influence on a country's exports remains unclear. By using the directional trade flow data, we are able to examine the possible effects of import tariffs on exports. Specifically, we find that, if a country lowers its import tariffs, its exports increase at a rate higher than economic growth. The relationship between import tariffs and exports can be attributed to: the "friendly-signal" effect; the reciprocal requirement for access to each other's markets; the increasing purchasing power of the source country; the efficiency improvements due to the import of physical capital; and the advantages of being a member of a trade agreement.

   Third, we broaden the scope of studies on regional trade agreements. We find that ASEAN not only benefits member countries but also non-members around the world, so there are clear spillover effects of tariff reduction benefits. Our analysis adds to the strand of literature that studies the impact of regional trade agreements on global trade integration (e.g., Lee and Shin 2006; Calvo-Pardo, Freund, and Ornelas 2011). Moreover, due to the similarity of membership between RCEP and ASEAN, we predict that RCEP will have similar but stronger positive effects on international trade flows.

   Fourth, this paper lends support to the gravity model--more specifically, the PPML regression model. The gravity model has been widely used in recent international trade studies (Anderson, Larch and Yotov 2016; Bergstrand, Larch and Yotov 2015; Yotov et al. 2016). It does not assume a constant return to scale production technology, perfectly competitive markets, or product differentiation by origin. The PPML model allows us to account for the presence of heteroscedasticity in trade data and to take advantage of the information contained in the zero trade flows. The model can use three groups of fixed effect variables to absorb the source-time, destination-time and time-invariant pairwise fixed effects. By including these, we can address the concerns related to omitted control variables in the traditional gravity model, such as GDP growth rates and bilateral distance between the two nations (Anderson, Larch and Yotov 2016; Santos-Silva and Tenreyro 2006). The PPML model also provides better estimation compared to the linear (OLS) regression models employed in the previous ASEAN studies, as the former generates consistent estimates in the presence of heteroscedasticity.

   The rest of this paper is organized as follows. We present the methodology and the data sources in the next section. The results of the regression analysis and the robustness checks are included in the third section. The final section concludes.

2. Methodology and Data

   2.1 Model Specification

   Following previous literature, (e.g., Anderson 2011; Anderson and Van Wincoop 2003; Anderson and Yotov 2010; Anderson, Larch, and Yotov 2016; Bergstrand, Larch, and Yotov 2015; Yotov et al. 2016), we use the gravity model for our analysis. The model utilises new econometric specifications that consider multilateral resistances and bilateral trade costs.

   Since our dependent and independent variables...