Will the real"natural trading partner"please stand up?

Adherents of the"natural trading partner"hypothesis argue that preferential trade agreements (PTAs) are more likely to improve welfare if participating countries already trade disproportionately with each other. Opponents of the hypothesis claim that the opposite is true: welfare gains are likely to be greater if participating countries trade less with each other. The author shows that neither analysis is correct. The"natural trading partner"hypothesis can be rescued if it is redefined in terms of complementarity or substitutability in the trade relations of countries, rather than in terms of their volume of trade. The author asks not whether a country should form or join a trading bloc but which partner or partners it should select if it does join such a bloc. He shows that the pre-PTA volume of trade is not a useful criterion for selecting a partner. The pre-PTA volume is equal to zero if the partner is an importer of the good sold to the home country and it is indeterminate if the partner is an exporter of that good. Among the author's conclusions: 1) The home country is better off with a larger partner country. First, a large partner is more likely to satisfy the home country's import demand at the world price. Second, the home country is likely to gain more on its exports to a large partner country, because that partner is likely to continue importing from the world market after formation of the trading bloc. And since the partner charges a tariff on imports from the world market, the home country is more likely to improve its terms of trade by selling to the partner at the higher tariff-inclusive price if the partner is large. 2) The PTA as a whole is likely to be better off if each country imports what the other exports (rather than each country importing what the other imports). Losses are similar but less likely, while gains are both more likely and the same or larger.