Policy panel: Trade integration in South Asia

The panel was chaired by Tony Venables (IGC and Oxford University) and issues on the welfare costs of India’s partition, the weak connectivity of the region, and distrust among countries dominated discussions. Robin Burgess (IGC and LSE) presented his co-authored on-going project on “The cost of international disintegration”. The main question is what the effects of partition of India into modern-day India, Pakistan and Bangladesh, are on intra-national and international trade. The authors use a Ricardian general equilibrium model. By observing changes in trade frictions (making trading costs prohibitive), the authors capture the effects of GDP and welfare. Results show that from partition, Pakistan’s welfare falls by around 4% of GDP, whereas India’s welfare falls by 2% of GDP. The main policy message is that the cost of disrupting trade could be very large, and potential gains from reinstating free trade agreements is also large.

Nagesh Kumar (United Nations Economic and Social Commission for Asia and the Pacific) stressed that distrust and weak intra-regional trade are symptoms of the whole South Asia (SA) region, save a couple of exceptions. Studies show that intra-regional trade is only ¼ or 1/3 of its potential. Whereas regional production networks that have been driving industrial restructuring in other regions, have not been taking off in SA due to lack of trust, a more important concern is connectivity. The costs of doing trade within SA are much higher than any other region. Although there are national forums that try to address the issue of connectivity and shared infrastructure, there is a need for a more regional approach, a need to integrate national strategies.

T.N. Srinivasan (Yale) noted that prior to the partition India, Bangladesh and Pakistan were joined in political union, a customs union and a currency union. This deep integration, which Europe is struggling to achieve, was abandoned and South Asia is now the least integrated region in the world. However, Srinivasan expressed doubt about the benefit of India’s current focus on negotiating bilateral trade agreements.

Venables raised concern about the model used in Burgess’ work, in its reliance on the assumption that the substitutability between domestic and foreign goods is the same across all goods. He emphasised that this can’t be true: if a country has no domestic oil reserves, there is no easy substitute for imported oil. He noted that when you allow for differences of trade elasticities across sectors, this increases the model’s estimates of the gains from trade by a factor of about eight. This would suggest that the results presented by Burgess are in fact lower bounds. Venables also expressed some concern about the use of an “off-the-shelf” model that is not fine-tuned to the specific context, which again may contribute to low estimates for the gains from trade.

In the discussion following multiple audience members emphasised that the time of partition was not a “typical period” due to political upheaval, and for example, the large amount of cross-border migration. Burgess noted that for this reason, and the concerns raised by Venables, the authors will also analyse disaggregated sector and regional data that dates back to 1870. On the topic of regional integration schemes, Kumar, echoing Venables, emphasised that negotiations are no longer simply about bargaining for market access. Rather, the purpose of regional integration schemes is much deeper integration, concerned with foreign direct investment and engagement in production networks. Kumar advised that countries who are not pursuing regional integration do so at their own peril.