This article is mainly interesting due to its original approach and main findings. In order to explain barriers to trade in Central Asia and assess the magnitude of the “border effect” ― a literature initiated by McCallum in North America (1995) ―, the authors collected prices of goods along both sides of the borders (between Kazakhstan, Kyrgyzstan and Uzbekistan). They find that the impact of borders across different locations in Central Asia is smaller than conventionally thought. They explain the lower impact of international borders by costly “internal borders” (borders between oblast’s) because of numerous checkpoints. The findings are a little bit counterintuitive. However, when referring to similar work in developing countries, such as in Africa, the same results have been recorded like for the case of the Benin-Nigeria border. In this case, the extent of smuggling between both countries seems to explain price equalisation taking into account that unrecorded flows reach several billion of dollars. Similarly, in Central Asia, the extent of unrecorded border and international trade may also contribute to explain the low impact of the border. Despite the cost of corruption at the border, Central Asia or at least Kazakhstan, Kyrgyzstan and Uzbekistan could be more integrated. One point worth noting is that the thickness of the Uzbek border is the largest for the three studied countries. The recent Uzbek trade policy has probably led to a strengthening of smuggling if we refer to the boom of bazaars like Karasuu in the Kyrgyz part of the Fergana valley. We can only wish that this work will contribute to the launching of similar research in Central Asia, which would be complemented by assessing the extent of internal borders and unrecorded trade since the latter seems to be on the rise in this region.

Gaël Raballand, Observatory of Post-Soviet States, Paris
CER: II-7.1-563