Snowballing alongside Domino on Proliferation of Preferential Trade Agreements

With regard to the third-country effects on bilateral incentives to sign new preferential trade agreements (PTAs), we develop a theory-based empirical approach that enables the assessment of how a country¡¯s pre-existing PTAs affect her formation of a new PTA. Assessment of this effect is not trivial because a country¡¯s pre-existing PTAs are also the pre-existing PTAs of her potential PTA partner¡¯s partner, generating two counter-acting effects on the bilateral incentives to sign a new PTA. Our empirical analysis shows that a country¡¯s pre-existing PTAs generate a positive effect on her incentive to sign a new PTA (i.e., Own PTA effect) but they generate a negative effect on her new partner¡¯s incentive to sign a PTA (i.e., Partner¡¯s PTA effect), as our theory predicts. The analysis also reveals that members of a CU are less likely to sign a new PTA than non-CU member countries as the Partner¡¯s PTA effect is magnified by the CU¡¯s joint negotiation requirement for a new PTA. Based on our n-country model of PTA proliferation, we conduct some calibration exercises that yield a varying degree of predictive power.