Welfare and market-access effects of piecemeal tariff reforms on environmentally preferable products

Abstract

We examine how welfare and market access are affected by piecemeal tariff reforms on environmentally preferable products (EPP) in a small open economy. We define EPP as clean goods that, when consumed, have no impact on pollution. First, we show that a uniform reduction of all tariffs improves welfare if a country’s imports consist only of clean goods. If a clean good is a net substitute for all other goods in excess demand, then reducing the highest tariff on the clean good improves welfare. Second, we show that a proportional tariff reduction leading to a welfare improvement also increases the value of imports if all tariffs are set at the same ad valorem rates. If the clean good is a net substitute for all other goods in excess demand, then reducing the lowest tariff on the clean good increases the value of imports. Finally, we explore the link between the change in welfare and the change in the value of imports in response to the tariff reforms, and we show that unlike a proportional tariff reduction, a tariff reduction on the clean good does not necessarily lead to improvements in both welfare and market access.

Keywords:

• Piecemeal policy reform
• trade policy
• tariff policy
• consumption externality
• market access
• trade and environment

JEL Classifications:

• F18
• H23
• Q56

Acknowledgements

I would like to thank Hikaru Ogawa, Shunsuke Managi and the seminar participants at Nagoya University and Tohoku University for their helpful comments and suggestions. I greatly appreciate the useful suggestions from the editor and anonymous reviewers concerning improvements to the paper. Any remaining errors are my own.

Notes

1. For example, the OECD (Organization for Economic Co-operation and Development) defines ‘environmentally preferable products’ as those that cause significantly less environmental harm at some stage of their life cycle than alternative products that serve the same purpose, and it lists bicycles, improved solid-fuel cooking stoves, and reusable shopping bags made of canvas or jute rather than plastic or paper as examples of such goods. See Tothova (2005), Potier and Tébar Less (2008), and WTO (2009) for a comprehensive survey of this issue.

2. This assumption is made for analytical simplicity. In fact, the exported good can be thought of as a single Hicksian composite commodity that consolidates many exported goods.

3. We do not consider pollution generated from production. The existence of production-generated pollution might affect our results depending on the importance of production-generated pollution relative to that of consumption-generated pollution.

4. Automobiles are an example source of consumption-generated pollution. For example, traditional gasoline vehicles negatively affect environmental quality (α > 0) by generating emissions as they are used by consumers. Their substitutes, such as hybrid vehicles, cause relatively less environmental damage. Electric vehicles produce zero emissions (α=0), making them even more environmentally preferable than gas or hybrid vehicles.

5. The case of no environmental policy is the same as the case in which a too weak (zero) pollution tax is imposed on the consumption of dirty goods. The presence of a positive pollution tax does not change any of our results for the second-best tariff reform. Moreover, in the presence of consumption-generated pollution, the first-best policy to correct the pollution distortion is not a tariff but a pollution tax on the dirty good. However, for informational reasons, a policy-maker might have difficulty imposing the relevant environmental policy and also implementing the second-best optimal tariffs. In that sense, this paper starts from the premise that the policy-maker faces these difficulties that stem from informational reasons and considers how a proportional reduction of all tariffs affects the small country’s welfare and market access.

6. The specific form of the expenditure function is introduced only for analytical simplicity. If the utility function, for example, takes a quasi-linear form and the pollution z is separable from c in the utility function, the expenditure function satisfies these properties.

7. None of our results will not affected even if we relax this assumption and instead have c _z > 0, a property that holds under the Cobb–Douglas utility function, among others. Nevertheless, this assumption significantly simplifies our analyses.

8. See Dixit and Norman (1980), Woodland (1982), and Feenstra (2004) for detailed properties of the GDP function.

9. The input vector is assumed to be fixed throughout the analysis and hence is omitted in the GDP function.

10. See Dixit and Norman (1980) and Neary (1995) for detailed discussions and interpretations of this condition.

11. This result is standard and has been extensively studied in the literature. See Dixit (1985) and Vousden (1990) for a comprehensive survey on this issue.

12. Again, see Dixit (1985) and Vousden (1990) for an excellent survey on this issue.

13. Intuitively, c_u /e_u = (c ¹ _u /e_u , …, c ^{h + 1} _u /e_u ) represents the vector of income effect terms. Thus, if all goods are normal, then cⁱ _u > 0 for i = 1, … , h + 1 because e _u > 0. See Dixit and Norman (1980) and Neary (1995) for detailed discussions and interpretations of this condition.