Files
Abstract
This paper searches for a remedy to the ‘small shares stay small’ problem that haunts the Armington import specification. This problem is especially acute if the (CGE) model is used to simulated large shocks in trade policies, such as a complete liberalisation of border policies. The paper uses a gravity model at the sub-sector level to disentangle the trade effects of tariff- and quota barriers next to non-tariff barriers and the fixed cost of exporting (captured by traditional gravity variables such as distance and cultural proximity). After all small shares should stay small after tariff reductions if tariffs are not the true reason for small initial shares, but import shares should react if trade barriers are the main cause for small observed trade flows in the base. The results of the gravity estimations are subsequently entered into illustrative simulations with GTAPEM – a tailored sibling of the standard GTAP model that is used at the OECD Trade and Agriculture Directorate.