Trade in intermediate inputs, customs unions, and global free trade

• Published date: 01 May 2019
• Author: David Tsirekidze
• Date: 01 May 2019
• DOI: http://doi.org/10.1111/roie.12392

666

|

wileyonlinelibrary.com/journal/roie Rev Int Econ. 2019;27:666–693.

© 2019 John Wiley & Sons Ltd

1

|

INTRODUCTION

Article XXIV of the General Agreement on Tariffs and Trade (GATT) allows groups of countries to

eliminate tariffs on each other that they do not have to extend to other members of the World Trade

Organization (WTO). These preferential trade agreements (PTAs) might take the form of free trade

agreements (FTAs) or customs unions (CUs). In the first case, member countries are free to choose

their own external tariffs for imports from nonmembers, while in the second case members set jointly

optimal external tariffs leading to coordination benefit.1 In this paper, my emphasis will be on CUs

only where the members of a CU treat nonmembers in a discriminatory manner, and this has an ad-

verse effect on the nonmember country’s welfare.

Received: 24 August 2018

|

Revised: 8 January 2019

|

Accepted: 10 January 2019

DOI: 10.1111/roie.12392

ORIGINAL ARTICLE

Trade in intermediate inputs, customs unions, and

global free trade

DavidTsirekidze

Edgeworth Economics, Washington, District

of Columbia

Correspondence

David Tsirekidze, Edgeworth Economics,

1111 19th Street NW, Suite 200, Washington,

DC.

Email: dtsirekidze@edgewortheconomics.

com

Abstract

In a three‐country customs union (CU) formation game, I

introduce international trade in intermediate inputs and

rules of origin (RoO) restrictions. In the case of symmetric

countries, I show that as countries become more involved

in global supply chains, global free trade is less likely to be

a stable equilibrium outcome. RoO can help solve this

problem. In the case of asymmetry, depending on the de-

gree of the globalization, free riding (for high degree) or

exclusion motive (for low degree) prevents global free

trade. Correspondingly, I show that RoO can have helpful

or detrimental effects on attaining global free trade.

JEL CLASSIFICATION

F13, F15

|

667

TSIREKIDZE

Since we see that tariffs are declining over time, the importance of nontariff barriers rises.2 The

type of barrier I focus on in this paper is called “rules of origin” (RoO) and it restricts the use of inter-

mediate inputs used in the final goods production traded within the CU. Although one may attribute

RoOs only to FTAs as a way to prevent nonmember countries exploiting differences in tariffs they

face when exporting to FTA member countries, CUs too may impose RoOs in order to ensure imports

maintain a minimum amount of CU member content. For example, in the case of the EU Customs

Union, in order for the office machines (Harmonized System [HS] headings 8456 to 8466) to be traded

duty free within the member countries, all the nonoriginating materials used in the manufacture of the

office machines must not account for more than 40% of the price of the finished product. Similarly,

the EU Customs Union also has RoOs for its imports from countries under the Generalized System of

Preferences (GSP) to ensure that GSP countries only receive preferential treatment when their exports

contain at least a certain amount of GSP content. For example, in the case of umbrellas (HS heading

6601) nonoriginating materials used should not exceed 70% of the price of the product.3

What effect does this have on the creation of CUs and the global trade regime? Does its existence

help achieve global free trade? To investigate these questions, in this paper I consider the interna-

tional trade model with intermediate inputs and RoO restrictions. So, we have three countries, each

producing two goods and consuming all three. Therefore, each country’s import market is served

by two “competing exporters” as in Bagwell and Staiger (1999). Any pair of countries can initiate a

CU, and this determines the global trade regime. A stable equilibrium outcome of this endogenous

CU creation game is global free trade if there are no self‐enforcing deviations (i.e., deviations that

are profitable for deviating countries and neither of these countries want to subsequently deviate

unilaterally).

To understand why there might not be global free trade as an equilibrium outcome of the game,

we need to examine the incentives of countries to deviate from it. There are two major impediments

to global free trade: (1) a country prefers to stay out of the trade agreement of the other two countries;

and (2) a pair of countries prefer to exclude the third country. For the first effect, the RoO can mitigate

this by limiting the benefits that the nonmember country gets. At the same time, the RoO make mem-

ber countries better off, as most of the benefits from trade agreements now go to them, as opposed to

the nonmember country.

However, in contrast to the first case, in the second case all the RoO can do is to make the members

even more willing to exclude the nonmember (and they will be better off by doing so). Therefore,

RoO can only shrink the parameter space under which global free trade is a stable Nash equilibrium

outcome. As I show below, this is true in the case of asymmetric countries. The smaller ones, taking

into account the effect of their tariffs on the other CU member’s welfare, can effectively be better off

by excluding the larger country.

Note that depending on the degree of trade in intermediate inputs the first or the second effect is

operational. When countries are highly interconnected and their inputs are extensively used in the

final goods production abroad, the free riding effect is more important, and we are in the first scenario.

However, when that interconnectedness is weak, we have a typical case of CU members benefiting by

excluding the nonmember from the trade agreement and exploiting the terms of trade gains.

Given the intuition above, the conclusion I make in this paper is that RoO can be helpful in the

CU case when there is a high degree of globalization. However, in the other scenario, where there is a

small degree of trade in intermediate inputs, RoO only makes global free trade harder to achieve as it

incentivizes the smaller countries to exclude the larger country even more.

The rest of this paper is organized as follows. In Section 2, I review the closely related literature.

Section 3 introduces the basic model based on Saggi, Woodland, and Yildiz (2013). There, the CU for-

mation game is discussed, and the importance of country asymmetry is highlighted. In Section 4, the