Skip to main content
Log in

Free Trade Networks on Non-tariff Barriers

  • Published:
Journal of Industry, Competition and Trade Aims and scope Submit manuscript

Abstract

In this paper, we examine the formation of free trade agreements on eliminating non-tariff barriers as a network formation game. A firm faces an import volume quota in a foreign country, which it cannot exceed but can be removed through bilateral free trade agreements. We focus on the stable free trade agreements networks, considering the case where each country determines the quotas of the foreign firms. We show that global free trade may not be achieved in any stable network. Moreover, all countries may be autarkic in a stable network. Finally, we demonstrate that our main results continue to hold regardless of the type of non-tariff barriers, such as import licences, import share quotas, and technical measures for increasing the production costs of foreign firms.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Notes

  1. See the home page of the European Commission: http://trade.ec.europa.eu/doclib/docs/2009/october/tradoc._145203.pdf.

  2. Goyal and Joshi (2006) also briefly discuss the formation of networks on eliminating NTBs. However, they conclude that free trade is not a unique outcome in their game. On the other hand, we provide a similar model but show that free trade may not be achieved in any equilibrium outcomes.

  3. Several works discuss the formation of trading blocs by using coalition formation models. See, for example, Bond and Syropoulos (1996), Yi (1996, 2000), Bond et al. (2004) and Kiyotaki and Miyakawa (2013). However, none of them discusses the agreements on reducing NTBs.

  4. Imposing an import quota is prohibited by Article XI of GATT. However, there are several exceptions to the rule. Thus, there are several import quotas in the real world. See, for example, OECD (2005).

  5. See the home page of the UNCTAD: http://r0.unctad.org/trains_new/tcm.shtm.

  6. 6 We also denote \(\bar {q}^{V}\) as a vector whose elements are \(\bar {q}^{V}\).

  7. Goyal and Joshi (2006; Proposition 7) only show that a stable network is either g c or g ci. On the other hand, we show that g c is always stable and g ci is stable if and only if Eq. 1 is satisfied. These stability conditions are not changed even if we consider the stability notion considered by Goyal and Joshi (2006). That is, Lemma 1 is satisfied even if “stable” is replaced by “pairwise stable” .

  8. 8 In this section, we consider the model where each country can freely choose its quotas assigned to the firms of the countries with which it does not form an FTA. For example, if g i j = g i k = 0, then country i can choose both \(q_{i}^{Vj}\) and \(q_{i}^{Vk}\). Note that the results of this section do not changed even if the countries are forced to choose the same import quota for the firms of the countries with whom they do not form an FTA. This is because, as will be shown later, in each equilibrium, each country chooses the same import quota for all countries with whom it does not form an FTA.

  9. This is because consumer surplus is (Q i )2/2 in our model. This fact is also true as long as the twicely differentiated consumer surplus is positive; that is, \(-\left (P_{i}^{\prime }(Q)+P_{i}^{\prime \prime }(Q)Q\right ) >0\) for all Q.

  10. See also Yi (2000), Bond et al. (2004) and Furusawa and Konishi (2006).

  11. Furusawa and Konishi (2006) also consider FTAs to eliminate tariffs but their model is different from that of Goyal and Joshi (2006). In the model of Furusawa and Konishi (2006), national welfare is also an inverted U-shaped function of the tariff-rate.

  12. Goyal and Joshi (2006, p.776) give a similar model and state that “It follows then that free trade is the unique outcome in a game with quotas under our assumption that N ≥ 3.” However, in our model, free trade is not the unique outcome even if k = 1 and n ≥ 3.

  13. This does not mean that a country has an incentive to commit to an exogenous import quota as in Section 3. In Section 3, we discuss the situation where all countries commit to an exogenous import quota. Therefore, our result implies that national welfare of a country is increased if all countries conclude an agreement to commit to an exogenous import quota. However, obviously each country may have an incentive to deviate from the agreement.

  14. Note that in this model, a country adopts the protective trade policy if i sets \({\gamma _{i}^{j}}\left (g\right ) \) prohibitively high for all j. On the other hand, a country adopts the free trade policy if i sets \(\gamma _{i}^{j}\left (g\right ) =0\) for all j.

References

  • Belleflamme P, Bloch F (2004) Market sharing agreements and collusive networks. Int Econ Rev 45(2):387–411

    Article  Google Scholar 

  • Bond EW, Riezman R, Syropoulos C (2004) A strategic and welfare theoretic analysis of free trade areas. J Int Econ 64(1):1–27

    Article  Google Scholar 

  • Bond EW, Syropoulos C (1996) The size of trading blocs, market power and world welfare effects. J Int Econ 40:411–437

    Article  Google Scholar 

  • Furusawa T, Konishi H (2005) Free trade networks with transfers. Jpn Econ Rev 56:144–164

    Article  Google Scholar 

  • Furusawa T, Konishi H (2006) Free trade networks. J Int Econ 72:310–335

    Article  Google Scholar 

  • Goyal S, Joshi S (2003) Network of collaboration in oligopoly. Games Econ Behav 43:57–85

    Article  Google Scholar 

  • Goyal S, Joshi S (2006) Bilateralism and free trade. Int Econ Rev 47(3):749–778

    Article  Google Scholar 

  • Jackson M (2008) Social and economic network. Princeton University Press, Princeton

    Google Scholar 

  • Jackson M, Wolinsky A (1996) A strategic model of economic and social network. J Econ Theory 71:44–74

    Article  Google Scholar 

  • Kiyotaki F, Miyakawa T (2013) Barriers to global free trade through bilateral agreements. Rev Int Econ 21(3):536–548

    Article  Google Scholar 

  • Kosteki MM (1991) Marketing strategies and voluntary export restraints. J World Trade 25:87–99

    Google Scholar 

  • Mukunoki H, Tachi K (2006) Multilateralism and hub-and-spoke bilateralism. Rev Int Econ 14:658–674

    Article  Google Scholar 

  • OECD (2005) Looking beyond tariffs: the role of non-tariff barriers in world trade. OECD Publishing, Paris

    Book  Google Scholar 

  • Okumura Y (2012) Volume and share quotas in oligopoly, Mimeo, Hannan University. Downloadable from http://www2.hannan-u.ac.jp/~okuyasu/Okumura2012.pdf

  • Ono Y (1990) Foreign penetration and national welfare under oligopoly. Jpn World Econ 2:141–154

    Article  Google Scholar 

  • Saggi K (2006) Preferential trade agreements and multilateral tariff cooperation. Int Econ Rev 47(1):29–57

    Article  Google Scholar 

  • Saggi K, Yildiz HM (2010) Bilateralism, multilateralism, and the quest for global free trade. J Int Econ 81:26–37

    Article  Google Scholar 

  • Saggi K, Yildiz HM (2011) Bilateral trade agreements and the feasibility of multilateral free trade. Rev Int Econ 19:356–373

    Article  Google Scholar 

  • Yi S-S (1996) Endogenous formation of CUs under imperfect competition: open regionalism is Good. J Int Econ 41:151–175

    Article  Google Scholar 

  • Yi S-S (2000) Free-trade areas and welfare: an equilibrium analysis. Rev Int Econ 8(2):336–347

    Article  Google Scholar 

  • Zhang J, Cui Z, Zu L (2014) The evolution of free trade networks. J Econ Dyn Control 38:72–86

    Article  Google Scholar 

Download references

Acknowledgments

I thank two anonymous referees for their helpful comments. An early version of this paper was presented at Kansai Game Theory workshop in Matsuyama University. I am grateful to Yasukazu Ichino and audiences of the workshop for their comments and suggestions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Yasunori Okumura.

Appendix

Appendix

Proof of Lemma 1

Goyal and Joshi (2006) show two facts. One is that if η i (g) ≥ 2, then S i (g + g i j ) > S i (g). Two is that if η i (g) = η j (g) = 1, then S i (g + g i j ) > S i (g). Since we use a stronger stability notion, we need to additionally show several facts. First, we consider g c. We show that S i (g) > S i (gg i j1−⋯−g j j L ) for any \(\{g_{ij_{1}},{\cdots } ,g_{ij_{L}}\}\) satisfying \( g_{ij_{l}}=1\) for all l = 1, ⋯ , L in g c. Since S i (g + g i j ) > S i (g) if η i (g) ≥ 2, S j (g c) > S j (g cg j j1) > S j (g cg j j1g j j2) > ⋯ > S j (g cg i j1−⋯−g i j n − 2). In addition, since

$$\begin{array}{@{}rcl@{}} && S_{i}(g^{c})-S_{i}(g^{c}-g_{ij_{1}}{\cdots} -g_{ij_{n-1}}) = \\ && \frac{1}{2}\left( \frac{nk\left( \alpha -\gamma \right) }{nk+1}\right)^{2}+nk\left( \frac{\alpha -\gamma }{nk+1}\right)^{2}-\frac{1}{2}\left( \frac{k\left( \alpha -\gamma \right) }{k+1}\right)^{2}-k\left( \frac{\alpha -\gamma }{k+1}\right)^{2}>0 \end{array} $$

for all i, the complete network is stable. Next consider g ci. Let i be the isolated country in g ci. Then, i has no incentive to add any FTA in g ci if and only if Eq. 1 holds. This is straightforward from equation (A.8) of Goyal and Joshi (2006). Moreover, since

$$\begin{array}{@{}rcl@{}} && S_{i}(g^{ci})-S_{i}(g^{ci}-\sum\nolimits_{l\neq i}g_{il}) = \frac{1}{2}\left( \frac{\left( n-1\right) k\left( \alpha -\gamma \right) }{ \left( n-1\right) k+1}\right)^{2}\\ && +\left( n-1\right) k\left( \frac{\alpha -\gamma }{\left( n-1\right) k+1}\right)^{2}-\frac{1}{2}\left( \frac{k\left( \alpha -\gamma \right) }{k+1}\right)^{2}-k\left( \frac{\alpha -\gamma }{k+1} \right)^{2}>0 \end{array} $$

for all ij, no country in the complete component wants to sever any subset of FTAs. □

Proof of Lemma 3

First, we show that for a given g, the maximizer of S i is either \({q_{i}^{V}}=0\) or \(\bar {q}^{V}\). The equilibrium national welfare S i =

$$\begin{array}{@{}rcl@{}} && \frac{1}{2}\left( Q_{i}\right)^{2}+\left( \alpha -Q_{i}-\gamma \right) \sum\nolimits_{l\in K_{i}}{Q_{i}^{l}}+\sum\nolimits_{j\neq i}\left[ \left( P_{j}-\gamma \right) \sum\nolimits_{l\in K_{i}}{Q_{j}^{l}}\right] \\ &&= \frac{1}{2}\left( \sum\nolimits_{l\in K_{i}}{Q_{i}^{l}}+M_{i}\right)^{2}+\left( \alpha -\sum\nolimits_{l\in K_{i}}{Q_{i}^{l}}-M_{i}-\gamma \right) \sum\nolimits_{l\in K_{i}}{Q_{i}^{l}} \\ &&+\sum\nolimits_{j\neq i}\left[ \left( P_{j}-\gamma \right) \sum\nolimits_{l\in K_{i}}{Q_{j}^{l}}\right] . \end{array} $$

By the first order condition, for all lK i ,

$${Q_{i}^{l}}=\frac{\alpha -\gamma -M_{i}}{k+1}\text{ and thus }Q_{i}=\frac{ k\left( \alpha -\gamma \right) +M_{i}}{k+1}. $$

Therefore,

$$\begin{array}{@{}rcl@{}} \frac{\partial S_{i}}{\partial M_{i}} &=&\frac{k\left( \alpha -\gamma \right) +M_{i}}{\left( k+1\right)^{2}}-\frac{2k\left( \alpha -\gamma -M_{i}\right) }{\left( k+1\right)^{2}} \\ &=&\frac{-k\left( \alpha -\gamma \right) +\left( 2k+1\right) M_{i}}{\left( k+1\right)^{2}},\end{array} $$
(4)
$$\frac{\partial^{2}S_{i}}{\partial \left( M_{i}\right)^{2}} =\frac{2k+1}{ \left( k+1\right)^{2}}>0\text{.} $$
(5)

By Eq. 5, for a given g, the maximizer of S i is either \( {q_{i}^{V}}=0\) or \(\bar {q}^{V}\).

Second, suppose N i (g) = {i}. Then,

$$\begin{array}{@{}rcl@{}} S_{i} &=&\frac{1}{2}\left( \frac{k\left( \alpha -\gamma \right) }{k+1} \right)^{2}+k\left( \frac{\alpha -\gamma }{k+1}\right)^{2}+\sum\nolimits_{j\neq i}\left[ \left( P_{j}-\gamma \right) \sum\nolimits_{i\in K_{i}}{Q_{j}^{i}}\right] \text{ if }{q_{i}^{V}}=0, \\ &=&\frac{1}{2}\left( \frac{nk\left( \alpha -\gamma \right) }{nk+1}\right)^{2}+k\left( \frac{\alpha -\gamma }{nk+1}\right)^{2}+\sum\nolimits_{j\neq i} \left[ \left( P_{j}-\gamma \right) \sum\nolimits_{i\in K_{i}}{Q_{j}^{i}}\right] \text{ if }{q_{i}^{V}}=\bar{q}^{V}. \end{array} $$

Therefore, country i chooses \({q_{i}^{V}}=0\) if and only if n ≤ 2k + 3.

Third, suppose η i (g) ≥ 2. We show that country i chooses \( {q_{i}^{V}}=\bar {q}^{V}\) in this case. If \({q_{i}^{V}}=0,\) M i = (η i (g)−1)k(αγ)/(η i (g)k + 1). By Eq. 4, S i / M i ≥ 0 for all η i (g) ≥ 2. Thus, if η i (g) ≥ 2, then country i chooses \({q_{i}^{V}}=\bar {q}^{V}\). (Q.E.D.)

(Proof of Remark 1) Since \(q_{1}^{V\ast }(\hat {g}) = q_{2}^{V\ast }(\hat {g}) = \bar {q}^{V}\) and \(q_{3}^{V\ast }(\hat {g}) = 0\),

$$\begin{array}{@{}rcl@{}} S_{1}(\hat{g}) &=&S_{2}(\hat{g}) = \frac{1}{2}\left( \frac{3\left( \alpha -\gamma \right) }{4}\right)^{2}+2\left( \frac{\alpha -\gamma }{4}\right)^{2}=\frac{13\left( \alpha -\gamma \right)^{2}}{32}, \\ S_{3}(\hat{g}) &=&\frac{1}{2}\left( \frac{\alpha -\gamma }{2}\right)^{2}+\left( \frac{\alpha -\gamma }{2}\right)^{2}+2\left( \frac{\alpha -\gamma }{4}\right)^{2}=\frac{\left( \alpha -\gamma \right)^{2}}{2}. \end{array} $$

Since \(q_{3}^{V\ast }(\hat {g}+g_{13}) = q_{3}^{V\ast }(\hat {g} +g_{23}) = q_{3}^{V\ast }(g^{c}) = \bar {q}^{V},\)

$$\begin{array}{@{}rcl@{}} S_{3}(\hat{g}+g_{13}) &=&S_{3}(\hat{g}+g_{23}) = S_{3}(g^{c}) \\ &=&\frac{1}{2}\left( \frac{3\left( \alpha -\gamma \right) }{4}\right)^{2}+3\left( \frac{\alpha -\gamma }{4}\right)^{2}=\frac{15\left( \alpha -\gamma \right)^{2}}{32}. \end{array} $$

We have \(S_{3}(\hat {g}+g_{13}) = S_{3}(\hat {g}+g_{23}) = S_{3}(g^{c})<S_{3}(\hat { g})\). Moreover, since \(q_{1}^{V\ast }(\hat {g}-g_{12}) = q_{2}^{V\ast }(\hat {g} -g_{12}) = 0,\)

$$S_{1}(\hat{g}-g_{12}) = S_{2}(\hat{g}-g_{12}) = \frac{1}{2}\left( \frac{\alpha -\gamma }{2}\right)^{2}+\left( \frac{\alpha -\gamma }{2}\right)^{2}=\frac{ 3\left( \alpha -\gamma \right)^{2}}{8}. $$

We have \(S_{1}(\hat {g}) = S_{2}(\hat {g})>S_{1}(\hat {g}-g_{12}) = S_{2}(\hat {g} -g_{12})\). Therefore, \(\hat {g}\) is uniquely stable. Furthermore, \(S_{3}(\hat { g})>S_{1}(\hat {g}) = S_{2}(\hat {g})\). □

Proof of Proposition 1

At first, we show two facts. First, if two countries i and j that satisfy η i (g) = η j (g) = 1, then the FTA between them is not formed if and only if Eq. 3 is satisfied. This is because neither i nor j wants to form an FTA between them if and only if

$$\frac{1}{2}\left( \frac{k\left( \alpha -\gamma \right) }{k+1}\right)^{2}+k\left( \frac{\alpha -\gamma }{k+1}\right)^{2}>\frac{1}{2}\left( \frac{ nk\left( \alpha -\gamma \right) }{nk+1}\right)^{2}+2k\left( \frac{\alpha -\gamma }{nk+1}\right)^{2}, $$

which is equivalent to Eq. 3. Note that if η i (g) = η j (g) = 2 and g i j = 1 and Eq. 3 is satisfied, then the FTA between them will be deleted and thus g is not stable. In addition, if Eq. 3 is satisfied, then Eq. 2 is also satisfied.

Second, we show that if there is a connected component including at least two FTAs in a network g, then g is not stable if and only if Eq. 2 is satisfied. We show the necessity of this fact. Suppose that there is a component including at least two FTAs. Let L be the members of the component. Note that at least three members in L. Let iL be a country such that η i (g) ≤ η l (g) for all lL. Now, suppose i deletes all FTAs that i forms in g and let g′ be the network. If η i (g) ≥ 3, then η l (g) ≥ 3 and thus η l (g′) ≥ 2 for all lL∖{i}. Moreover, if η i (g) = 2, then η j (g′) ≥ 3 for all j such that g i j = 1 because L has at least three countries. Therefore, η l (g′) ≥ 2 for all lL∖{i}. By Lemma 3, the firms of country i can freely sells in the countries L∖{i} in g′. Thus, i will sever the FTAs that i forms in g and g is not stable if and only if Eq. 2 is satisfied.

Suppose that Eq. 3 holds. Then, by the facts introduced above, the empty network is uniquely stable. Suppose that Eq. 2 holds and Eq. 3 does not. If there are i and j such that η i (g) = η j (g) = 1, then g is not stable. Moreover, if there is a component including at least two FTAs, then g is not stable. Therefore, if n is even, then η i (g) = 2 for all i in any stable network g. If n is odd, then η j (g) = 1 and η i (g) = 2 for all ij in any stable network g. Finally, if Eq. 2 does not hold, then all networks are stable and all countries adopt the free trade policy. □

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Okumura, Y. Free Trade Networks on Non-tariff Barriers. J Ind Compet Trade 15, 223–238 (2015). https://doi.org/10.1007/s10842-014-0184-2

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10842-014-0184-2

Keywords

JEL Classification

Navigation