Networks of free trade agreements.

Author:Li, Liaoliao

    In recent years, there are more literatures using network structures to study economics and social issues (Calvo-Armengol and Jackson, 2004;Corominas-Bosch, 2004; Krantonand Minehart, 2001; Wang and Watts, 2006).Motivated by the fast development of free trade agreements (FTAs), we are interested inexploringsocial welfare in countries that have formed FTAs and the stability of these agreements under the framework of free trade networks. In this paper, we are interested instudying networks of FTAs, in whichtwo countries are linkedwhen they sign an FTA to eliminate tariffs on imports and exports between each other. FTAs have been studied as trade networks by Li (2010), Mauleon et al. (2009), Furusawa and Konishi (2007), Goyal and Joshi (2006), Zissimos (2002), Das and Ghosh (2004). We add to this discussion by using an asymmetric way of introducing differentiated goods and transportation costs to the framework.

    In the present paper, we construct a three-country model to explore the incentives of countries to form free trade networks, i.e. the evolution and formation of free trade networks, and the stability of networks. We are interested in finding out how countries separated by distance and transportation costs may coordinate to sign an FTA, and whether a complete free trade network can be achieved. Because of transportation costs, trade is costlier for countries between regions than within them. Results from spaceless models may be invalid if imports and exports are very sensitive to the extra costs. In this case, countries are more likely to form an FTA if they are geographically close to each other, and free trade networks tend to be regional. Examples include the North American Free Trade Agreement (NAFTA). In 1994, Canada, the United States and Mexico formed the world's largest free trade area, and the three countries share the common borders. Ethier (1998) reviews the international commercial system.

    Li (2010) found that without the existence of transportation costs a complete free trade network, in which all countries have FTAs with each other, can be pairwise stable. Furusawa and Konishi (2007) investigated trade networks in a setting with n countries that trade a numeraire good and a continuum of horizontally differentiated industrial commodities when two countries are symmetric in the market size and the industry size and levy the same tariffs. The central conclusion in their paper is that when two countries are symmetric in the market size and the industry size, and levy the same tariffs, the stable global free trade network will be attained. Goyal and Joshi (2006) studied network games in an international trading system and concluded that if countries are identical, bilateralism is consistent and should be seen as a building block for global free trade.

    Mauleon et al. (2009) considered an n-country model in which the labor market is either unionized or nonunionized. If all countries are non-unionized, the complete free trade network is the unique pairwise stable and efficient network, but the same conclusion can only be drawn for unionized countries if they are also symmetric. Zissimos (2002) explained that the optimal tariffs are higher between closer countries because with lower transportation costs more rents may be shifted through tariffs. Similarly, Zissimos (2006) examined the issue of why custom unions (CUs) are regional. Saggi and Yildiz (2008) explore the trade liberalization in an FTA game, in which each country can form an FTA with either one of its trade partners, or both of them, or none of them. If countries are symmetric, free trade is uniquely stable. Bhattacharya et. al. (2008) used weighted networks to describe the social interactions in the static networks of population migration, international trade, and etc. Rauch (1996; 2001) examines the trade networks in the international trade system. Saggi and Yildiz (2008) talk about the development of bilateral free trade agreements. Herrmann-Pillath (2006) studies the Pareto-improving changes of trade policies, while Mukunoki (2005) talks about trade agreements from a theoretical aspect.

    Different from previous literatures, the present paper considers a trade network model in which three countries A, B, and C compete as Cournot oligopolists and produce differentiated products. They are located in different areas and then they must bear transportation costs when trading with each other, which might be so profound that the development of free trade networks can be interrupted or restricted. We start from a simpler situation where goods 1 and 2 are perfectly substitutable and then release this assumption in the later discussion. Under both circumstances, we investigate different cases where the three countries are located in the same regions; the three countries have the same distance between each other; and country i and country j are in the same region but the third country is father away. Based on the detailed analyses, we can have a thorough understanding of the formation and stability of free trade networks in a general setting. Our study enriches the literature by introducing transportation costs and oligopolistic competition into the study of free trade networks.

    The paper is structured as follows. In Section 2, we present the basicmodel and derive the quantities and prices in equilibrium, firms' profits, and the social welfare function. Then we analyze countries' incentives to form a free trade network in two specific situations, and develop the main results concerning the stability of free trade networks in Section 3. Section 4 concludes the paper.


    We use a three-country (countries A, B and C) trade network model in which each country...

To continue reading