It is no strange thing in economics to find out that the development of the theory on a particular topic and the development of empirical research have pursued rather different courses. It is difficult to imagine a better example of this divergence than the study of the effects of international economic integration (Mayes, 1988, p. 42). Various studies have attempted to measure the effects of international economic integration. Some of them intended to show the advance in integration as the interlace of economic ties among countries. Others wanted to measure the increase in welfare. Yet another group of studies strove to measure the distribution of the costs and benefits of integration. However, the theory of international economic integration encountered in this book is so complex that it cannot be represented mathematically with a great precision.
Econometric models which attempt to measure the effects of international economic integration can be divided into those which are ex ante and ex post. The former models are often founded on a simple extrapolation of trends. These models can not be based on reliable data, but rather on estimations. The problem represents the fact that one has no reliable data on future developments. It is assumed that the future flow of foreign trade is a function of income, production, relative prices, change in the level of trade barriers, substitutability among import sources, as well as between imports and domestic production.
The ex post models attempt to measure a hypothetical situation (the counterfactual world or anti-monde) which may represent what could have happened with trade in a situation without integration. The difference between the actual and expected imports of each of the participating countries represents trade creation, while the same difference in imports from outside countries shows trade diversion. These differences may be attributed to autonomous changes in prices, changes in income and competition, reductions in barriers to trade and to errors.