The impact of macroeconomic closures on long run trade projections and trade policy experiments

In this paper the impact of the macroeconomic closure in recursive dynamic computable general equilibrium (CGE) models on both long run trade projections and on the effect of trade policy experiments is examined. Six different closures of the trade balance are compared employing the WTO Global Trade Model: a fixed trade balance ratio; a Feldstein-Horioka rule; a converging trade balance ratio; the GTAP rate of return and fixed investment share rules; and a combination of the rate of return rule and the Feldstein-Horioka rule. The paper describes into detail how the different closures are coded, in particular the combination of the rate of return and Feldstein-Horioka rule and the treatment of the omitted region. The simulations on the impact on long run trade projections focuses on outcome variables such as changes in trade balances and the share in global trade of different regions. The simulations will show that long-run trade balances and trade patterns are strongly affected by the choice of macroeconomic closure. The simulations on the effect of trade policy experiments will focus on the projected changes in key outcome variables typically reported such as trade volumes, real GDP, and real income. This will for example shed light on questions such as how the pattern of trade diversion is affected by the macroeconomic closure and how the projected macroeconomic effects change with movements in international investment flows under different closures. An actual trade policy experiment will be employed such as the increase in tariffs and non-tariff barriers between the US and China.