This paper revisits the empirical trade literature on East West trade in the early 1990s. Using the estimation technology commonly in use in 1989, I re-estimate a simple gravity model for 48 countries using 1988 data covering 88% of world trade and 92% of world GDP and find significant barriers to trade both in East West trade and against Chinese exports. My estimates that only cover direct trade effects and do not include spill-overs and dynamic effects (for example on long term growth) indicate that these walls together significantly reduced bilateral trade flows. Breaking down those walls according to my calculations increased world trade by 20% or 3.6% of world GDP (this amounts to a good third of the increase in global trade openness since 1990).
Typically the trade impact of walls is especially strong at the regional level. Comparing the regional impact to the global impact (using trade potential in per cent of GDP), I find that the regional impact was 2 to 4 time larger than the global impact (the ratios are: Western Europe 1.8, Asia 2.3 and Eastern Europe 3.8, respectively). In 1989 the Berlin Wall and Iron Curtain exerted a much stronger impact than the ‘invisible Chinese wall’ (at the global level about 15 times as large). This is even true in the Asian trade with China where low levels of bilateral trade in 1988 are by and large explained by low levels of GDP. This does not mean that this invisible wall was less significant per se, but rather that the regional conditions in this particular case in 1988 were such that trade would be low both with and without walls. An indication of the potential impact of the ‘invisible Chinese wall’ is that its influence in terms of percentage trade reduction was stronger in developed but far away markets (in particular the Netherlands and the Nordic countries) than in less developed nearby markets (such as Malaysia, Thailand and India). This suggests that the ‘invisible Chinese wall’ would have started to bite once development in Asia took off.
In order to check the latter hypothesis I perform a thought experiment introducing the visible and invisible walls in a gravity simulation for the world trade system in 2008, based on parameter estimates for 1988. I use 2008 population and GDP data and take account of the break down of a number of former Communist countries, shifts in capital cities in Nigeria and Germany and the German unification. The relative global impact of the regional walls is smaller in the less concentrated context of 2008. Introducing the hypothetical ‘invisible Chinese wall’, the simulation finds an impact on trade potential of 1% of GDP for an invisible Chinese Wall, which is comparable to the 1.5% that I find for a hypothetical ‘re-erection’ of the Berlin Wall and Iron Curtain.
|Series||ISS working papers. General series|