Master EPP, Eco-572 International Economics PC 4 Trade Policies

Exercice 1: Import quotas and free trade agreements US imports of sugar are subject to a quota. The figures used in this exercise are rounded up but close to reality. Thanks to the quota, US production of sugar is 6 million ton/year, instead of 5 million without the quota, and US consumption of sugar is 8 million ton/year, instead of 9 million without the quota. The US consumer pays $480/ton, whereas the world price is $280/ton. 1) What is the volume of the quota? Plot US supply and demand curves and show graphically the impact of the quota for consumers and producers. Why is the US price higher with the quota? 2) Calculate the gains and losses to US consumers and producers. How much would the latter be willing to secure the quota? 3) Calculate the quota “rent”. Who receives this rent? 4) Comment the following graph. Quota rents: the case of tariff-rate quotas Rent by importing country, in $Mn in 2001

UE Can USA

Japon

Suisse

Corée

Mex

World trade = $4277Bn in 2001. A tariff-rate quota is a value or a volume of a specific good that can be imported at a reduced tariff rate during a certain period. Any additional import is taxed at the normal tariff. Source: CEPII, MacMaps database.

5) Despite intense lobbying by US sugar producers, a free-trade agreement is concluded between the US and the EU. European sugar can now be imported tariff-free with no limitation in the US. The European price is $380/ton. The price of sugar in other regions (Brazil, Australia, Thailand, Cuba) is $280/ton, but each imported ton is subject to a $200 tariff in the US. 1

Show graphically the impact of the FTA compared to a situation where the US imports its sugar from other regions at the world price with the $200 tariff. Is there trade creation or diversion? Is it realistic to assume that the world price remains constant at $280/ton? Exercice 2: Protection for Sale in the data: Consider a simplified version of Grossman and Helpman’s “Protection for Sale”’ model. There is a continuum of individuals of mass one that share identical preferences given by: U = c0 +

n X

ui (ci )

i=1

where ci is the consumption of good i and c0 the consumption of the numeraire good. ui is an increasing 0 −1 concave function. di (pi ) is the demand for good i implied by these preferences (d (p ) = u i i i ()). The P indirect utility of an individual with income y is given by V = y + i [ui (di (pi )) − pi di (pi )]. Welfare W is equal to the sum of all agents’ indirect utilities. There are n + 1 inputs, labor and one sector-specific input for each sector. Each individual owns one unit of labor and at most one type of specific factors. αi is the share of people who own the specific factor i. Good 0 is produced one-to-one from labor so that the wage is equal to one. Goods 1...n are produced with labor and the specific factor that corresponds to the sector. The returns to sector-specific inputs only depends on the price: πi (pi ) = maxLi pi yi − Li . The government chooses specific trade taxes that introduce a wedge between the domestic and the world price: pi = p∗i + τi where p∗i is the exogenous world price and τi the trade tax (import tariff or export subsidy). The government redistributes the revenue from trade policy in lump-sum fashion, equally to all citizens. In some subset of sectors J = {1, ..., k} (k ≤ n), the owners of P specific factors are organized in lobbies. The share of the population that belongs to a lobby is αJ = i∈J αi . Lobbies pay a contribution Ci to the government in order to influence trade policy. Lobby i’s objective is equal to welfare of capital owners in the industry minus the contribution (Wi − Ci ). The government’s objective is a combination of welfare and contributions, with β the weight of welfare: X U G = βW + (1 − β) Ci i∈J

Assume that the interaction between lobbies and the government takes the form of a Nash bargaining game.1 Trade policies are thus selected to maximize the joint surplus of all parties involved: X X max βW + (1 − β) Ci + (1 − β) (Wi − Ci ) τi

i∈J

i∈J

1. Denoting by Mi (pi ) = di (pi ) − yi (pi ) the volume of imports in sector i, write the aggregate welfare W and the aggregate welfare of lobby i Wi . 2. Noting that the derivative of uj (dj (pj ))−pj dj (pj ) with respect to pj is equal to −dj (pj ) in optimum, write the first-order condition of the previous Nash bargaining game. Use this condition to derive the 1

The solution of that bargaining game is such that each agent receives a constant fraction of the total surplus, β in the case of the government. This simplifies the problem by removing the need to determine equilibrium contributions. One consequence is that the simplified model yields no clear prediction regarding contributions.

2

Table 1: Results from the Basic Specification (Goldberg & Maggi, 1999, Table 1) Scaling Factor µ=1 µ=1 µ=2 µ=2 µ=3 µ=3

Variable γ δ γ δ γ δ

Coefficient -0.0109 0.0111 -0.0155 0.0161 -0.0182 0.0193

StdErr .0043 .0055 .0062 .0080 .0073 .0095

optimal tariff set in industry i as a function of the import penetration ratio, Mi /yi and the import i elasticity, ei = ddlnlnM pi . Comment. When does this model predict zero trade costs? Goldberg and Maggi (AER, 1999) propose to test this prediction of Grossman and Helpman’s model. Their estimated equation writes: τi yi yi ei = γ + δIi + i 1 + τi Mi Mi with τi the ad-valorem tariff on good i, ei is the import demand elasticity on good i, yi is the domestic production of good i, Mi is the import demand for good i, Ii is a dummy variable equal to one if industry i is politically organized. i is an error term capturing variables potentially affecting protection that may have been left out of the theoretical model. The equation is estimated by maximum likelihood using a cross-section of sectoral non-tariff barriers. To measure protection, Goldberg and Maggi use coverage ratios for non-tariff barriers instead of advalorem tariffs. Coverage are obtained using an aggregation procedure across products within P iratios i where ni is a dummy equal to one if product k is covered by some w an industry: cri = n k k k k non-tariff barrier and wki is the weight of product k in sectoral imports. The relationship between this τi coverage ratio and equivalent tariffs is then assumed to be constant, 1+τ = µcri where µ = 1, 2, 3. i There is a strong uncertainty about the value of import elasticities (ei ) at the sectoral level. Goldberg and Maggi (1999) use estimates obtained from the literature. As a proxy for political-organized dummies (Ii ), the authors use data on political action committee campaign contributions. These contributions are all positive in the data at the 3-digit SIC level. A literal interpretation of the model would thus imply that all industries are politically organized. However, it may be that some of those contributions are not used to influence trade policy but domestic policies instead. An ad-hoc threshold is thus defined, below which the industry is considered non-organized. Finally, import penetration ratios are measured from actual data on sectoral outputs and imports. 3. How do the estimated coefficients map with the parameters of the model? What are the signs expected on both coefficients? 4. Why do you think Goldberg and Maggi use non-tariff barriers instead of tariffs? Discuss potential measurement error and endogeneity issues in the estimating equation. 5. Table 1 presents the estimated coefficients obtained by Goldberg and Maggi under different assumptions about the scaling factor µ. Discuss the results.