Arguments for protection: infant industry, dumping and tariffs against sales taxes

The case for free trade is that it raises output, income and employment, and keeps firms honest, since imports stop a domestic monopoly from raising prices (CUET 2022 keys A and B: maximisation of output, income and employment, and the prevention of monopoly). The case for protection has to beat that. Nine arguments are heard. Filter them by how well they hold.

The test to apply to any argument is the one economists call the first best test. If a market failure is real, is a tariff the best tool to fix it? Usually it is not, because a tariff also distorts consumption. A subsidy aimed at the failure itself does better. This is why most arguments land in the middle column: right about the problem, wrong about the remedy.

The oldest argument. A new industry has high costs at first, because its workers and managers are learning. If it were protected until it had learned, it could stand on its own. Infant industry protection is the policy of protecting a new domestic industry from lower-cost imports (JNU 2010), and UOH 2018 words it as a tariff designed to protect young manufacturing firms from foreign competition.

The argument can be tested with two questions, named for two economists. The Mill test: will the industry reach the world cost, so that it can stand alone? The Bastable test: do the later gains outweigh the cost of the protection? An industry can pass the first and fail the second. Try the lab. Costs fall with the learning curve: each time cumulative output doubles, unit cost falls by a fixed share.

With the starting settings (₹40, 25% learning, 8% discount) the industry reaches the world price in year 6, protection costs 335 in present value, and the later saving is 424, a net gain of 89. Drop the learning rate to 15% and the picture reverses: year 20, a cost of 681 against a saving of 17, a net loss of 664. It passes the Mill test and fails the Bastable one. The year of arrival, with learning rate r and a starting cost c₀ over the world price p_W, is where $c_0{t}^{{\log }_2(1-r)}=p_W$.

Even a pass on both tests does not settle it. If the industry will really be profitable later, why do private lenders not finance the early losses? An answer is needed: capital markets that cannot lend against future profits, or learning that spills over to other firms who do not pay for it. Without such a failure, protection only transfers money. And a tariff is not the best tool: a subsidy to the industry costs the same to firms and does not tax buyers.

Applied to a whole economy, the infant industry argument becomes a development strategy. Import-substituting industrialisation protects home industry with tariffs and quotas so that home production replaces imports. Several Hyderabad questions test the link: the infant industry argument is an important theoretical basis for import substitution (UOH 2014 and 2017), and in an infant-industry model of development, quantitative and tariff protection are given to domestic firms (UOH 2012). It is an inward-looking strategy (UOH 2021), set against an outward-looking one led by exports.

India followed an import-substitution strategy from independence until 1991, when the reforms began to dismantle the import controls. The costs were what the theory warns of: protection that was meant to be temporary lasted, and firms shielded from competition had little reason to cut costs.

Dumping is selling in a foreign market at a lower price than at home. UOH 2012 and 2017 define it so, and the second adds a lower price than the cost of production. The WTO describes it as introducing a product into the commerce of another country at less than its normal value. The most common form of price discrimination in international trade is dumping (UOH 2014).

Why would a firm do it? Because it has market power at home, and a lower willingness to pay abroad. A firm that can keep the two markets apart does best by charging each its own price. The lab shows the result.

In the default case the firm charges ₹55 at home and ₹25 abroad, a dumping margin of 54.5%, and earns 2,250 against 2,025 from a single price. Three kinds are told apart:

• Persistent dumping is this price discrimination. It lasts, and importing consumers gain from the low price.
• Predatory dumping is a price below cost, kept up to drive out rivals, after which the price is raised. This is the one anti-dumping law has in mind.
• Sporadic dumping is occasional, to dispose of a surplus.

Under the WTO’s rules a member may impose an anti-dumping duty only after finding that dumping is occurring, that the domestic industry making the like product is suffering material injury, and that the dumping caused it. The duty is meant to offset the dumping.

A neighbour of dumping is transfer pricing: the over- or under-pricing of goods traded between parts of the same multinational firm in different countries (CUET 2024). By pricing between its subsidiaries, a firm can show its profit where tax is lowest.

One more argument, not about industry: a tariff raises money. Governments of poor countries have often relied on it, because it is easy to collect at a port. But there is a cheaper way to raise the same money. UGC NET 2023 states it as an assertion: if governments did not care about protectionists’ rents, they would replace all import duties with sales taxes, since a sales tax is a more efficient means of raising revenue. Both statements are true and the second explains the first. Here is why.

For a revenue of 100, the tariff is ₹5 and wastes 50; the sales tax is ₹1.77 and wastes 3.1. The tariff is the worse instrument because it taxes two decisions at once. If the rents of the protected industry matter to the government, it may still choose the tariff, and that is the point of the assertion’s condition.

The standard model has many small firms and a price taker. Put one home firm against one foreign firm in the home market, each choosing a quantity (a Cournot duopoly), and a new argument appears: a tariff raises the foreign firm’s cost and so shifts output and profit to the home firm. JNU 2015 sets it up with demand Q = 200 − 2P, marginal costs of 10 and a tariff of 10.

Without the tariff each firm makes 60 and the price is ₹40. With it, the home firm makes 66.7, the foreign firm 46.7, the price is ₹43.33, and the government collects 467. This is the logic of strategic trade policy, associated with Brander and Spencer: in an oligopoly, a government can shift profit to its firms. The same models show how fragile the result is: it depends on the form of competition and on the other government not responding, which is why it is a case to understand, not a policy to recommend.

Two further ideas appear in the papers. The first is fair trade. Developing countries have asked developed ones to reduce subsidies to agriculture, which lets rich-country farmers undercut poor-country ones, and that is an argument for fair trade (UOH 2012), not for free trade or for protection. The second is non-tariff barriers, such as quotas, standards and licensing, which can restrict imports as effectively as a tariff, with less visibility. They are the means by which protection can go on after tariffs are bound or cut. The tariffs and quotas page shows how a quota differs from the tariff that would match it.

Read on 6 October 2026. The definition of dumping and the conditions for anti-dumping measures: the WTO’s anti-dumping page. India’s import-substitution period and the reforms that began in 1991: the Peterson Institute’s India’s trade reforms 30 years later and the NBER paper Dismantling the License Raj. The labs use stylised numbers, not data; the learning curve and the discount rate are assumptions you can change. For example, with the 20% learning rate the unit cost in year 4 is ₹25.6.