The identification of valid instruments for endogenous variables is a common problem in the verification of the endogene. A valid instrument must be strongly correlated with the explanatory endogenous variable, but not with the notion of error, and we used a one-step delay of suspicious endogenous variables as an instrument to test possible endogenousness with the Durbin-Wu-Hausman test. The zero hypothesis of exogenous variables was not rejected either for final GDP or for trade openness (see Table 13 of The Annex). Other variables that can lead to endogenous problems are the PTA and the ILO, but due to the difficulty of obtaining valid instruments for these variables and the inadequacy of the use of their delayed forms, we are not testing their exonesis here. In this paper, we examine the trade effects of preferential trade agreements (EPAs) for a large sample of countries over the 1962-2000 period. The paper builds on existing literature by examining whether the significant effects of ATPs are due to a change in export diversity (large margin) or a change in the volume of existing products (intense margin). To tackle this problem, we use the frequently used gravitational equation, as well as an appropriate approach to solving potential self-selection problems. Our results indicate that exports are responding positively to the formation of an ATP between countries and that much of this increase in exports is done along the large margin. We also show that the large margin is more in response to the formation of a PTA among large exporters and large pairs of countries. Osnago A, Rocha N, Ruta M (2016) Do deep trade agreements encourage vertical direct investment? The World Bank Economic Review, Flight 30, pp S119-S125 Tobin J, Busch M (2010) A BIT is better than lot: bilateral investment agreements and preferential trade agreements. Global Policy 62 (1): 1-42 Hayakawa K, Kimura F (2015) To what extent do free trade agreements reduce trade barriers? Open Econ Rev 26 (4):711-729 Jaumotte F (2004) Foreign direct investment and regional trade agreements: the effect of market size verified (IMF working paper 206).
IMF, Washington, D.C An WEF can be estimated in different ways, including within the transformation, between effects or the LSDV approach. The latter was chosen because it allows the confinement of temporal variables. Kohl T, Trojanowska S (2015) Heterogeneous trade agreements, WTO membership and international trade: an analysis with a coherent economy. Appl Econ 47 (33): 3499-3509 The F test on the importance of fixed country effects rejected the zero hypothesis that there are no significant differences between countries (F – 27.46, p – 0.00) at the 5% meaning level, indicating that pooled OLSs do not fit. Medvedev D (2011) Beyond Trade: The Impact of Preferential Trade Agreements on Inflows. World Dev 40 (1):49-61. doi.org/10.1016/j.worlddev.2011.04.036 World Trade Organization (2011) World Trade Report 2011. Author, Swiss Bergstrand J, Egger P (2007) A model of knowledge and physical capital of international trade flows, foreign direct investment and multinational companies. J Int Econ 73 (2):278-308 Buthe T, Milner H (2014) Foreign direct investment and institutional diversity in trade agreements: credibility, commitment and economic flows in developing countries, 1971-2007. Global Policy 66(01):88-122 Each methodology has its strengths and weaknesses. National case studies have the advantage of being communicated more consistently and modifying certain control variables (e.g.B.