Trade policy under imperfect competition

Trade policy with monopoly • We will first consider a Home monopolist. – Under free trade, many new firms can sell in the Home market, thus eliminating the monopolist’s power. – Thus, free trade creates a perfectly competitive Home market. – Because the firm has market power however, tariffs and quotas affect the trade equilibrium differently.

• We will then consider the case of a monopolist Foreign exporting firm. – We will show that applying a tariff under a Foreign monopoly leads to an outcome similar to the large country case. – The tariff will lower the net-of-tariff price received by the Foreign exporter. – A tariff may now benefit the Home country.

2

Trade policy with monopoly Free trade

•

Home monopolist produces the same quantity and charges the same price as a perfectly competitive industry would.

With tariff

•

Losses due to the tariff are the same as under perfect competition. 3

Trade policy with monopoly Import quota

•

•

•

The effective demand facing Home monopolist is the old demand curve D minus the quota M2. The price is higher: P3 > PW + t. - This reflects the ability of the monopolist to raise its price once the quota is reached. - This occurs even if the quota allows the same amount of imports under tariff. - Tariff and quota are no longer equivalent! Higher price creates higher deadweight loss under a quota than a tariff.

4

Tariff to a foreign monopoly •

Since the Home country is paying a lower net-of-tariff price for its imports, it has experienced a termsof-trade gain as a result of the tariff.

• • • •

Decrease in Home CS: (c + d) Tariff revenue: (c + e) Net effect: e – d As in the large country case, Home welfare rises for small tariffs.

•

Note: graph assumes no Home firms. If there were Home firms too, there would be some production distortion, but also profit gains!

•

5

Trade policy with oligopoly • Consider the auto industry: U.S. versus Japan. – Toyota is the only Japanese auto company. – GM is only American auto company.

• Assume that the two firms produce cars that are interchangeable in the eyes of consumers. • Assume transport costs are zero and both firms have the same MC, so two firms split the market equally. • For simplicity, also assume that demands are identical across the two countries. • We will first suppose that the industry is structured as a Cournot oligopoly. • We will then move to Bertrand price competition. 6

Cournot model • Consider the following demand and MC: Q = 20,000,000  2,000P , and MCGM = MCT = $5,000 • GM’s and Toyota’s reaction functions for U.S. can be obtained as: 1 qGM  5,000,000  qT 2 1 qT  5,000,000  qGM 2 • Solving them, we get: qGM = qT = 3.33 million and P = $6,667.

7

8

Cournot – import tariff • Now consider a tariff on imported cars in the U.S., in the amount of $1000 per car (specific tariff). • This raises the marginal cost of servicing the U.S. market to $6000 for Toyota, and changes its reaction function to: 1 qT  4,000,000  qGM 2 • GM’s reaction function stays the same. • Solving for the new equilibrium, we get: qGM = 4,000,000, qT = 2,000,000 and P = $7,000. 9

10

• The tariff can raise US welfare by inducing Toyota to cede market share to GM. • “Rent-shifting” motive: Transfers some oligopolistic profits from foreign firm to a domestic one – this wouldn’t exist in a world of perfect competition. • Since prices rise, this rent-shifting occurs at the expense of U.S. consumers. • However, under the case analyzed, it is easy to check that U.S. welfare (consumer surplus + profits + tariff revenue) increases.

11

Cournot – export subsidy • Now consider an export subsidy of $1000 to GM for exporting to Japan. • This will decrease GM’s marginal cost of serving Japanese market from $5000 to $4000. • New equilibrium: qGM = 4.67 million, qT = 2.67 million and P = $6,333. • It is possible to verify in this case that U.S. welfare improves with export subsidy. – Again, a “rent-shifting” motive.

• Radical change from perfect competition models where a country can never raise its welfare with an export subsidy. 12

13

Cournot – import quota • Suppose U.S. imposes an import quota of 2 million cars. • Effects are identical to tariff. Toyota’s new reaction function will go horizontal at 2 million cars until the point where it voluntarily want to produce less than that (point A in the next graph). • New equilibrium at point E’ where Toyota exports 2 million cars, GM sells 4 million cars, and the resulting situation exactly the same as an import tariff of $1,000.

14

3.33 2

3.33 4

15

Bertrand model • Now consider price competition between GM and Toyota. • Take same demand and MC. • Recall that in a Bertrand model, firms undercut each other’s price until it is not feasible anymore. • Thus, initially without any trade policy tools used, both firms charge a price equal to their marginal cost ($5000) and make zero profits. • Each firm sells 5 million cars in each market.

16

Bertrand – import tariff • Now, consider a tariff of $1000 imposed by the U.S. government on imports. The tariff raises Toyota’s marginal cost for servicing the U.S. market to $6000. • Although both firms continue to split the Japanese market, GM will capture the entire U.S. market and charge a price that is slightly below $6000 per car, thus producing 8,000,000 cars in total. • Here there are no rents to shift, but there are production and consumption distortions. • Therefore, the tariff here can only lower US welfare.

17

Welfare loss

18

Bertrand – export subsidy • Now consider the case in which the U.S. government pays a subsidy of $1000 for each car GM exports to Japan. • GM’s marginal cost of serving the Japanese market will drop to $4000. • This time, GM will capture the entire Japanese market. It will charge a price slightly below its rival’s marginal cost ($5000) and sells 10,000,000 cars in Japan. • GM makes profit of $1,000 per car. • But this is exactly the fiscal cost of the subsidy. • Net effect on US welfare is zero. • Once again, there are no rents to shift, so the trade policy cannot help in the way it can with Cournot competition. 19

Bertrand – import quota • Now consider a quota that restricts Toyota’s US sales to exactly 5,000,000 cars per year. • At first glance, it would seem that this will have no effect since Toyota is not selling any more than that even under free trade. • However, this is not the case. Now, if GM decides to charge more than $5,000 per unit, it will not lose all of its customers because Toyota cannot increase the number of US customers it serves. • Hence, GM will gladly charge more than $5,000 and earn a strictly positive profit. • This is good news for Toyota because it will allow Toyota to increase its own price somewhat above $5000 and still sell 5 million cars. 20

• In summary, if Toyota still charges $5,000, GM will raise its price. • If GM is expected to charge more than $5,000, Toyota will do so as well. • There’s no pure strategy Nash equilibrium (outcome will be in mixed strategies). • But we know that both firms will charge more than $5,000. • Thus trade restrictions ‘facilitate’ cartel-like behaviour: both firms make zero profits under free trade, but positive expected profits under an import quota. • Since consumers will pay a higher price, there will a net deadweight loss at the end, so US will have lower welfare.

21

Sequential Bertrand – import quota • What if GM sets its price first and Toyota afterwards? In other words, what if GM is the Stackelberg leader? • GM anticipates that Toyota will slightly undercut GM’s price. Therefore, relevant demand curve for GM is the residual demand after Toyota’s 5 million cars are deducted: Qresidual= 15,000,000– 2,000P • GM sets the associated MR equal to MC, which yields: MRGM = MCGM  QGM = 2,500,000 and PGM = $6,250

• Toyota, on the other hand, will charge slightly below $6,250, and sell 5 million cars. • Once again, both firms make positive profits under the quota, compared with zero profits under free trade. • However, US welfare is lower since the total quantity is 22 reduced compared to free trade.

23