Thanks to steep fall in tariffs over the years and complex rules of origin, much of world trade is happening outside of FTAs
Eminent economist Jagdish Bhagwati in his 2008 book, Termites in the Trading System: How Preferential Agreements Undermine Free Trade, lamented how an ever-increasing number of free trade agreements (FTAs) are a threat to the world trading system. Ten years on we must know if the threat has materialised.
But why does Bhagwati consider FTAs to be bad? In an FTA, two or more countries agree to lower import tariffs and other trade barriers on each other’s products. Good for them at one level but bad for the overall trade, because of the two effects that take place as a result.
Economists term these trade diversion and trade creation. Trade diversion favours less efficient producers while trade creation stresses local producers. To understand, let us take the example of a shirt.
Let’s presume that all shirts are identical and have the same quality and a consumer will buy from the cheapest source. Cost of one shirt sold by the US is ?1,000 and the EU, ?1,100. If the import duty in India on the shirt is 20 per cent, cost of one shirt imported from the US will be ?1,200 and that from the EU, ?1,320. Now, since the price of shirts from the US is lower, Indian consumers will prefer to buy them. The game changes if India signs an FTA with the EU and eliminates import duty on shirts from the EU. So shirts from the EU can now enter India at ?1100, while shirts from the US will continue to come at ?1,200. So India will stop buying from the US and start buying from the EU. Notice that shirts from the EU cost more, but duty elimination through the FTA makes them less expensive.
Share of Preferential Trade
% share of trade with FTA partner
Countries in Global trade ( 2016 data)
County / Region Exports Imports
EU 13.6 13.6
US 45.0 34.1
US (excluding NAFTA) 13.2 8.3
India 19.4 18.1
China 30.1 23.2
ASEAN 31.1 42.1
Japan 20.8 24.1
South Korea 76.9 76.1
Australia 71.1 68.1
Source: The World Integrated Trade Solution (WITS)
Since the India-EU FTA diverted trade from the more efficient US to less efficient EU producer, the effect is termed trade diversion. Bhagwati considers it bad as the FTA rewards a less efficient producer.
Let us now understand the impact of an FTA on the local industry. Consider the shirt example again. A shirt produced in India sells at ?1,150. Pre-FTA, no imports will take place as this price is lower than the duty paid price of shirts from the US (?1,200) and EU (?1,320).
But position changes after the FTA. Now the shirts from the EU ( at ?1,100 per piece) will cost less than the shirts produced in India ( at ?1,150 per piece). The phenomenon is called trade creation as the FTA created new trade in the form of imports from the EU. Earlier there was none.
Soon imports from the EU will replace locally produced shirts. After some time, Indian shirt makers would shut shop.
Bhagwati called FTAs bad mostly because of trade diversion and trade creation effects. The big question is to what extent these effects distort world trade . The answer lies in finding out how much of the world trade happens through the FTA route.
Since FTAs allow trade at zero import duty on most products and world over 280-plus FTAs are operational, it is widely believed that most world trade happens through the FTA route. We did a data check and the results surprised us.
Global and bilateral export-import data show that much of world trade takes place outside the FTAs (this excludes intra-EU trade as the EU is an integrated economic entity).
Only 13.6 per cent of EU’s exports and imports take place through FTA partner countries. Most of its 41 FTA partners, except South Korea, supply raw-materials and low-end products.
For Japan, only 20.8 per cent exports and 24 per cent imports happen through the 17 FTA partner countries.
ASEAN and China are the centres of the East Asian International Production Network. But only 30.1 per cent of China’s exports and 23.2 per cent of imports take place through the FTA partner countries. And for ASEAN, 31.1 per cent of exports and 42.1 per cent of imports takes place through six FTA partner countries.
In the case of India, 19.4 per cent of its exports and 18.1 per cent of its imports takes place through the FTA partner countries. We do not have a full FTA with China yet.
Australia and South Korea are two exceptions where share with FTA countries is high. Australia’s 71 per cent and South Korea’s 76.9 per cent of exports go to their FTA partner countries.
We also noticed that an FTA does not always lead to increase in share or growth of trade between the partners. North American FTA (NAFTA) signed in 1994 among the US-Canada-Mexico is an example. NAFTA’s share in US trade was 28.9 per cent just before the signing of NAFTA in 1994. Today it is 31.8 per cent. And if we take out the NAFTA, only 13.2 per cent of the US export and 8.3 per cent of its imports come from the remaining 18 FTA partners.
These examples highlight that the share of most countries’ trade with their FTA partners is 20-40 per cent of their total global trade. But even most of this trade takes place outside of the FTA. There are several reasons for this. FTAs lower the import duty, but if it is already zero, FTAs can offer no benefit. And the import duty on products covering about half the world trade is zero, courtesy negotiations at the WTO.
Other reasons for low use of the FTAs are complex rules of origin. Rules of origin impose conditions like minimum value addition that need to be met when an input from a non-FTA partner country is used for making a product. Taking the trade between FTA partner countries and moderating the figure to count only preferential trade between them, we conclude that about 15-17 per cent of world trade is preferential.
The FTAs, in the early 1990s, became the central part of trade strategy of many countries, which believed that more the FTAs they sign, higher will be their trade. The tariffs in most countries were high, and no business-relevant decisions were taken at the WTO. Quickly this led to most countries signing FTAs with others. Today we have 284 operational FTAs. And more than 200 FTAs under negotiation.
But we are in different time zone now. Average tariffs have come down from over 100 per cent in pre-1990s to 0-10 per cent now. For most developed countries these are just 2-4 per cent. Any country doing an FTA with them will not gain much market access in most products. Also, the countries with high tariffs doing FTAs with those with low tariffs end up giving more benefits and market access to the FTA partners. The losses can hardly be balanced by gains in other areas like services or investments.
So while the alarm raised by Bhagwati was real and the FTAs had the potential to damage the multilateral trading system and world trade, they could not. Reason: Most (about 83-85 per cent) world trade takes place outside the FTAs. Only 15-17 per cent of trade is on preferential terms. But as the US action of increasing the tariffs and China’s response shows tariffs are still the central means of regulating imports. And these should be reduced through the FTAs only when economic benefits can be clearly demonstrated.
The writer is an Indian Trade Service officer. Views are personal. Data inputs by Dr Kishor Jadhav