Low tariffs, high protection

In a recent column (Wanted: Better trade policies, August 9, 2018), I lamented that the revival of higher protective customs tariffs over the past year (especially in the February 2018 Union budget and several times since) has been a damaging reversal of the cross-party reform commitment to reducing our tariffs to “East Asian levels” that had prevailed since 1991. I pointed out that such tariffs raise domestic costs and prices, hurt exports, foster inefficiency, encourage further tariff hikes, invite retaliation and generally weaken the development of an efficient and competitive manufacturing sector. They hurt trade and do little to reduce large trade and current account deficits.

Some have noted that most tariff increases thus far have been modest, usually not raising duty levels to above 20 per cent and hence may not lead to significant economic damage. This is wrong, because even relatively moderate tariffs of around 20 per cent can be highly protective and distortionary. To understand this one needs to grasp the concept of “effective rate of protection” (ERP) which comes into play whenever there are (tradable) intermediate goods in production activities, which is almost always the case. The concept was developed and explained by a number of economists in the 1960s, including Max Corden, Bela Balassa, and G Basevi.

Illustration: Binay Sinha

ERP focuses on the protection afforded to value added in producing a commodity. This contrasts with “Nominal Protection” (NP), which simply refers to the tariff rate on the product/output. A high ERP means that factors of production (capital, labour , land) would be attracted to the production of that commodity or sub-sector, while a low or negative one would not attract resources. Specifically, ERP for a product or activity is defined as: ((value added in domestic prices)-(value added in international prices))/(valued added in international prices) x 100. Consider an example where a product (say, cotton garments) is priced at Rs 100 per unit at international prices, its sole (tradable) intermediate input (say, cotton fabric) costs Rs 50 for the amount used in producing a garment and all tariffs are zero. Thus value added at international prices is Rs 100-Rs 50= Rs 50 and so is the value added at domestic prices. Both NP and ERP are zero by definition. (The currency does not matter; so “Rs ” will be dropped henceforth). 

Now assume tariffs are imposed at the rate of 20  per cent on both the output and input. The domestic price of the product goes up to 120, the domestic cost of the necessary input goes up to 60 and value added at domestic (tariff-inclusive) prices rises to 120-60=60. Hence the ERP, which was zero pre-tariffs, rises to (60-50)/50 x 100=20 per cent, equal to the rate of NP. This is Scenario A (see Table). A key implication is that when tariffs are imposed at a uniform rate (20  per cent in this case), there is no difference between the rates of NP and ERP for the product or output.

Let us consider some alternative scenarios. In scenario B the only difference is that the tariff on the input is lower (10 per cent) than the tariff on the output (20 per cent), which is a very common pattern in our customs duty structure. This results in a ERP of 30 per cent, substantially higher than the unchanged NP of 20  per cent. Scenario C is especially instructive: the proportion of valued added in total unit price/cost (at international prices) is assumed to be 20 per cent, instead of the 50  per cent assumed in previous scenarios. Such a low value added proportion is quite common in industrial activities. The tariff rates applicable in scenario B are assumed unchanged. In this case the ERP shoots up to 60 per cent, three times higher than the NP. Scenario D simply illustrates that changes in the value added proportion do not affect the earlier “result”: when tariff rates are uniform across inputs and outputs, there is no difference between the ERP and NP. This result holds even if the tariff rates are higher, at 40 per cent, for both input and output (Scenario E). 

The next two scenarios are the most worrisome and, regrettably, quite common. In scenario F the output tariff rate (40 per cent) is assumed significantly higher than the input rate (10 per cent), leading to an ERP of 160 per cent, four times the NP of 40 per cent! Similarly, in scenario G, even with a modest output tariff rate of 20 per cent one gets an ERP of 100 per cent, five times higher than the NP, when there is zero tariff on the input. Both these scenarios and the earlier scenario C show how easy it is to have high ERPs even when NPs look quite “moderate”. 

Sometimes, the output tariff is lower than the input tariff. In scenario H when the output tariff is zero, the input tariff is 20  per cent and the proportion of value added at international prices is 20 per cent of the unit price, the ERP becomes minus 80  per cent! This illustrates how damaging tariffs are for exports, which typically do not enjoy protection or subsidies; that is the “output tariff” is zero and the input tariff is positive.

These scenarios illustrate some key propositions for the design of customs tariffs if they are to be deployed at all: 

1) Tariffs , when deployed, should be low and uniform across inputs and outputs.

2) Since variations in value added proportions are inevitable across industries, even small dispersion in tariff rates can lead to large variation in ERPs across industries.

3) Even low product tariffs can lead to very high rates of effective protection, depending on value addition proportions and input tariff rates. 

4) Tariffs on inputs typically lead to negative (sometimes highly negative) effective protection (that is, discouragement) of export items, unless such tariffs are rebated. 

5) The complexity and variation in effective rates of protection, even with low nominal tariffs, offers large scope for both deliberate and unintended benefits for different industries and firms, which need to be guarded against.

Clearly, the design of tariffs is a difficult technical task if the problems above are to be avoided or minimised. Unfortunately, my 15-year experience in the Finance Ministry suggests that crucial concepts, such as ERP, are not well understood in the tariff-making bodies of government.

The writer is Honorary Professor at ICRIER and former Chief Economic Adviser to the Government of India. Views are personal