We investigate preferential trade agreement (PTA) formation when risk averse countries face demand uncertainty and, hence, have an insurance motive for pursuing trade integration. In this environment, when deciding which type of PTA − if any − they wish to form, countries seek to maximise their net welfare; that is, their expected utility less a risk premium. The desire for insurance influences, not just whether a particular PTA forms, but also the preferred depth of integration. We analyze the insurance implications of free trade agreements (FTAs), customs unions (CUs), and countries choosing to stand alone. We further distinguish between shallow CUs and deep CUs; in the former, members maximise the sum of their individual net welfares, while in the latter they maximise the net value of the sum of their individual expected welfares. We show that differences in country risk attitudes and the levels of risk they face, as well as the degree to which these risks are correlated with each other, each, and together, influence the formation and design of TAs. When countries’ demands are uncorrelated, they form a deep CU if their levels of risk aversion are sufficiently different. If, however, their risk attitudes are similar, countries opt for shallower trade integration − either a shallow CU or a FTA − if they face low levels of uncertainty, and choose to stand alone if one country faces a sufficiently high level of uncertainty. When countries’ demands are correlated, they tend to form a deep CU if their demands are strongly negatively correlated, a FTA if their demands are strongly positively correlated and a shallow CU when their demands are weakly correlated. Intuitively, differences in country risk attitudes (i.e., their degree of risk aversion) act as an additional source of comparative advantage. Deeper integration − particularly via a CU − permits less risk averse members to essentially export their relative partiality for risk to more risk averse partners, thereby effectively providing the latter with insurance.
Read full abstract