The previous chapters discussed some reasons for why economists have been divided on the wisdom of PTAs. Following Viner's (1950) demonstration that the netwelfare effects of PTAs are unpredictable and possibly negative, many attempts were made to refine the theory and identify member-country characteristics that would ensure welfare improvement and, thus, the welfare ambiguities associated with preferential trade (e.g., see Meade [1955], Lipsey [1958, 1960], and Johnson [1962], and a later synthesis by McMillan and McCann [1980]). However, these efforts yielded results that did not have any greater direct operational significance than did Viner (1950). That is, they did not yield any direct insights on which country characteristics would make trade creation rather than trade diversion a likely outcome and, therefore, which particular countries would be more desirable as partners in a PTA.

More recently, increasing emphasis has been placed on geographic proximity as a criterion for membership in a PTA. Regionalism in preferential trade has been argued by some as being key to generating better economic outcomes. Thus, Wonnacott and Lutz (1987), Krugman (1991), and Summers (1991) each proposed geographical proximity as a key predictor of trade creation and welfare improvement in PTAs, calling proximate trading partners “natural” partners for a PTA. Thus, for instance, Wonnacott and Lutz (1987) state: