Bilateral trade agreements and the feasibility of multilateral free trade Kamal Saggi (SMU) and Halis M. Yildiz (Ryerson University) 1
1. Introduction By permitting countries to form free trade agreements (FTAs), Article XXIV of the General Agreement on Tariffs and Trade (GATT) provides an important exception to MFN. Notion of MFN is at the heart of the WTO system. Article XXIV has been controversial widespread concern regarding the effect of FTAs on multilateral trade liberalization. Does the option to form FTAs reduce the likelihood of obtaining global free trade? Bhagwati: Are FTAs building or stumbling blocs? Would world welfare be higher if countries could pursue trade liberalization only multilaterally? 2
Why do we care? Over 176 PTAs are in force today with another 140 or so at different stages of negotiations. An all encompassing web: only four WTO members Hong Kong, Macao, Mongolia, and Chinese Taipei donotformallybelongtoapta. On average, each country belongs to six PTAs (over 80% of which are FTAs). Most PTAs were concluded in the last 10 years or so: Since 1995 over 130 such agreements have been notified to the WTO. Prominent examples: NAFTA, MERCOSUR, and the numerous agreements of the European Union with other countries. Today, MFN seems to have become more of an exception rather than a core role of the WTO. 3
2. Approach of the paper We analyze the coalition proof Nash equilibria of two games of trade liberalization between three countries where each country is free to form multiple FTAs. Under the FTA game, countries can pursue either bilateral, multilateral, or no trade liberalization whereas under the No FTA game, they have only the latter two options. Analyze the stable (i.e. coalition proof) Nash equilibria of the above games allow countries to deviate jointly and isolate Nash equilibria that are immune to self-enforcing coalitional deviations. The underlying framework is one of intraindustry trade under oligopoly allows a neat separation of domestic surplus and export profits and makes asymmetries easy to handle. 4
3. Preview of main results The option to form bilateral FTAs reduces the likelihood of obtaining free trade when countries are symmetric (Weak Stumbling Blocs). However, under asymmetry there exist circumstances where multilateral free trade is an equilibrium only if countries are free to form bilateral FTAs (Strong Building Blocs). Free trade can fail to be an equilibrium even when bilateral FTAs are not permissible. Welfare improving bilateral FTAs can be feasible when global free trade is not (Partial Building Blocs) A hub and spoke type arrangement i.e. a regime where one country has a bilateral FTA with both its trading partners who in turn do not have an FTA with one another fails to be a stable equilibrium when countries are symmetric. 5
4. (Very closely) related literature Krishna (1998): an FTA reduces incentives for multilateral trade liberalization. Does not derive equilibrium FTAs. Aghion et. al. (2004) examine a leading country s choice between sequential and multilateral bargaining. Here: all countries are free to negotiate FTAs; countries can form a pair of bilateral FTAs; tariffs transmit externalities created by FTAs; and no transfers are allowed. Riezman (1999): a cooperative game theory approach; solves for the core under some numerical examples. Models of repeated interaction between countries see Bagwell and Staiger (1997 and 1998a), Bond et. al. (2001), Bond and Syropoulos (1996), and Saggi (2006). Levy (1997) and Freund (2000). 6
5. Underlying trade model Countries: a, b, andc and goods: x and y. Preferences: U(x, y) =u(x)+y. Good x produced by a single firm in each country at constant marginal cost. Cournot competition under segmented markets. Export profits of country j where j 6= i: Π ji =[p i (x i ) j t i ]x ji FOCs p i + p 0 i x ji = j + t i Assume dx ji dt i < 0 < dx ii dt i ;and dx i dt i < 0 7
Welfare of country i: W i (t) S i (t)+ X j Π ij (t) where domestic surplus S i (t) is S i (t) u(x i ) p i x i + Π ii + t X j x ji Assumption 1: S(t, t) >S(0,t) >S(0, 0) i.e. domestic surplus of each country is highest under no agreement and lowest when it practises free trade. 8
6. Endogenous FTAs FTA game: I) Each country announces whether or not it wants to form an FTA with each of the other two countries. II) Given trade agreements, firms compete in a Cournot fashion. No FTA game (Restricts the strategy space): I) Countries can choose only between free trade or no agreement. II) Same as No FTA game. Trade-off: An FTA lowers a country s domestic surplus but increases export profits. 9
Strategy set of country a: Possible regimes: Ω F a = {{φ, φ}, {b, φ}, {φ, c}, {b, c}} (i) No agreement h{φ}i no announcements match or everyone announces φ. (ii) Bilateral FTA h{ij}i both i and j announce each other s name. (iii) Two independent FTAs in which i is the common member h{ij, ik}i obtains when (1) j α i and i α j and (2) k α i and i α k. This is a hub and spoke type arrangement. (iv) Freetradeh{F }i obtains when α a = {b, c}, α b = {a, c}, andα c = {a, b}. 10
7. Remarks on the FTA game Different announcements can give the same outcome. Considerα a = {b, φ}, α b = {a, c}, α c = {φ, b}. These announcements give h{ab, ac}i. Suppose: α a = {b, c}, α b = {a, c}, α c = {φ, b}. Also yield h{ab, ac}i. Fewer possible deviations from h{f }i in the No FTA game but a unilateral deviation from h{f }i has different effects in the two games. In the No FTA game, any unilateral deviation from h{f }i results in no agreement. Not so in the FTA game if country c deviates from h{f }i, faces the FTA h{ab}i as a nonmember. 11
8. Equilibrium FTAs under symmetry Symmetry: i = for all i Lemma 1: Under symmetry, free trade yields higher world welfare than any other trade policy regime: ww(f ) > ww(ij, ik) >ww(ij) >ww(φ). Intuition: tariffs create a global deadweight loss. Proposition 1: Under symmetry, no agreement h{φ}i and free trade h{f }i are both Nash equilibria of the No FTA game. No unilateral incentive to deviate from free trade since any other announcement leads to no agreement (where everyone is worse off). 12
To derive Nash eq. of the FTA game, need to specify two conditions Condition 1: w k (ij, ik) <w k (ij). If condition 1 holds, each country prefers to be a nonmember under a bilateral FTA to being a spoke under a pair of bilateral FTAs. Suppose h{ik}i exists. The formation of h{ij}i lowers country k s export profits in both markets and this reduces the value of the original FTA h{ik}i to country i. Condition 2: w i (F ) <w i (ij). If condition 2 holds, each country is better off as a member of a bilateral FTA relative to free trade beneficial to exclude the third country. Obvious that h{φ}i is a Nash equilibrium of the FTA game. 13
Is h{ij}i an equilibrium too? Lemma 2: The formation of a bilateral FTA between countries i and j makes country k worse off relative to no agreement: w k (ij) <w k (Φ). Export profits of country k are higher under h{φ}i relative to h{ij}i ( Π(t, t) > Π(t, 0)) whereas its domestic surplus under both regimes equals S(t, t). Since ww(ij) >ww(φ), we must have w i (ij)+w j (ij) >w i (Φ)+w j (Φ) From symmetry w i (ij) =w j (ij) >w i (Φ) =w j (Φ) Hence, h{ij}i is a Nash equilibrium of the FTA game. 14
What about a hub and spoke type arrangement h{ij, ik}i? This is also a Nash equilibrium if condition 1 fails w k (ij, ik) >w k (ij). Why? Neither spoke country will break its link Lemma 2. Hub will not deviate either enjoys higher welfare than even free trade: w i (ij, ik) >w i (F ). Domestic surplus same as free trade but export profits are higher. 15
Is free trade a Nash equilibrium? Before proceeding, note the following result: Lemma 3: Under h{ij, ik}i, each spoke country (i.e. j and k) isworseoff relative to free trade. Since ww(ij, ik) <ww(f )andw i (ij, ik) >w i (F ) it must be that w j (ij, ik)+w k (ij, ik) <w j (F )+w k (F ) Symmetry implies w j (ij, ik) =w k (ij, ik) <w j (F )=w k (F ) 16
For free trade to be a Nash eq, need to rule out two deviations: 1. UF1: k s deviation from h{f }i to h{ij}i and 2. UF2: k s deviation from h{f }i to h{ij, ik}i. UF1 cannot occur because w k (F ) >w k (Φ) >w k (ij). UF2 cannot occur because each spoke is worse off relative to free trade (Lemma 4). Proposition 2: No agreement h{φ}i, abilateralfta h{ij}i, and free trade h{f }i are all Nash equilibria of the FTA game. In addition, if condition 1 fails then a pair of bilateral FTAs h{ij, ik}i is also a Nash equilibrium. 17
9. Stable Nash equilibria Allow countries to deviate jointly and isolate Nash equilibria that are immune to self-enforcing coalitional deviations. The No FTA game: h{φ}i and h{f }i. Which, if any, is stable? All three countries want to deviate from h{φ}i to h{f }i (Lemma 1). If no unilateral or joint incentive to further deviate from h{f }i, then h{φ}i is not stable. But since h{f }i is a Nash eq. we have: Proposition 3: h{f }i is the unique stable equilibrium of the No FTA game. Leaves little room for FTAs to act as building blocs. 18
Consider now the FTA game and start with h{φ}i. Countries i and j have an incentive to jointly deviate from h{φ}i to h{ij}i. Since h{ij}i is a Nash equilibrium of the FTA game, the initial joint deviation of i and j from h{φ}i to h{ij}i is self-enforcing and h{φ}i is not stable. Now consider h{ij, ik}i as a candidate. Countries j and k benefit from a joint deviation from h{ij, ik}i to h{f }i. Since h{f }i is a Nash equilibrium of the FTA game, no further deviations from it. As a result, h{ij, ik}i is also not stable. 19
Two candidates remain: h{f }i and {ij}. For h{f }i to be stable, need to rule out: JF1: Deviation of i and j from h{f }i to h{φ}i. JF2: Deviation of j and k from h{f }i to h{ij, ik}i. JF3: Deviation of i and j from h{f }i to h{ij}i. Since w i (F ) > w i (Φ) deviation JF1 cannot occur. Similarly, since w j (ij, ik) <w j (F )deviationjf2can be ruled out. If condition 2 fails (i.e. w i (F ) >w i (ij)) then deviationjf3beruledoutandfreetradeis(uniquely) stable. But what if condition 2 holds? Then, country i wants to deviate from h{ij}i to h{ij, ik}i. This implies that JF3 is not self-enforcing and h{f }i is stable. 20
Now consider h{ij}i. For it to be stable, need both conditions 1 and 2 to hold. Proposition 4: Free Trade h{f }i is a stable Nash equilibrium of the FTA game and it is uniquely stable if either condition 1 or condition 2 fail. A bilateral FTA h{ij}i is a stable Nash iff both condition 1 and condition 2 hold. Finally, no agreement h{φ}i and a pair of bilateral FTAs h{ij, ik}i are not stable Nash equilibria. Under symmetry, FTAs end up acting as weak stumbling blocs: while h{f }i is the unique stable equilibrium of the No FTA game, there exist conditions under which a bilateral FTA is also a stable equilibrium of the FTA game. Does a different role for FTAs emerge when countries are not symmetric? 21
9. FTAs under asymmetry Drop the assumption that the production cost of good x is equal across countries. The higher a country s cost of producing good x, the smaller its volume of exports and the larger its volume of imports. Assumption 2: w i (r v) < 0, w i(r v) > 0and w i(r v) < 0 i m n where m is an FTA partner of country i under regime r (but not regime v) whereasn is either a partner or not under both regimes r and v (i.e. its status with respect to country i isthesameunderregimesr and v). 22
Example: Consider regimes h{φ}i and h{ij}i. A2 says that a country s willingness to enter into a bilateral FTA with another depends (i) negatively on its own cost ( w i(ij Φ) < 0); i and (ii) positively on the cost of its partner( w i (ij Φ) ). m Also, welfare gain of bilateral liberalization with j declines with the cost of the excluded country ( w i(ij Φ) k ). Intuition:k s exports are low if its cost is high and the strategic advantage enjoyed by i under h{ij}i is less consequential. 23
10. Feasibility of free trade Sufficient to focus on a = b =0and c = >0. From Lemma 1 we know lim 0 w i (F Φ) > 0for all i = a, b. Define P as: lim P π ca (F )=lim P π cb (F )= 0. Since S(t, t) >S(0, 0),wehave: We conclude lim w c(φ F ) > 0 P w c (Φ) w c (F )iff φ Proposition 5: Free trade is the unique stable equilibrium of the No FTA game iff < φ. Or else we obtain no agreement. We say free trade feasible in the No FTA game only if φ. 24
11. FTAs as strong building blocs First note that the viability of free trade depends critically upon the high cost country: Proposition 6: The two low cost countries have no incentive (either unilateral or joint) to deviate from free trade. Let s be defined by w c (F ) >w c (s) iff < s. We know w c (Φ) >w c (ab)or w c (Φ F ) > w c (ab F ): Lemma 5: φ < ab. Condition 1A: w c (ab, ac) <w c (ab). 25
Proposition 7: Suppose condition 1A holds. Then, free trade is a stable equilibrium of the FTA game iff < ab. Proposition 5 and 7 together imply the following: Corollary 1: Suppose condition 1A holds. Then, the option to form bilateral FTAs acts as a strong building bloc whenever φ << ab. Insight: if global free trade is infeasible in the No FTA game, the fact that the low cost countries can form a bilateral FTA can make the high cost country a willing participant in global free trade since it is worse off as a non-member country under h{ab}i than it is under free trade. 26
When > ab free trade is not feasible under either game. What about bilateral FTAs? Proposition 8: If (i) condition 1A holds and (ii) > ab, then the bilateral FTA h{ab}i is a stable equilibrium of the FTA game. When > ab,ftasarepartial building blocs: h{φ}i under the No FTA game while h{ab}i under the FTA game. What other equilibria are possible under asymmetry? Proposition 9: Suppose condition 1A fails. Then, in theftagame(i)freetradeh{f }i is stable when ab,ac and (ii) the pair of bilateral FTAs h{ab, ac}i is stable when ab,ac. Corollary 2: If condition 1 A fails, FTAs act as strong building blocs when φ << ab,ac. 27
If condition 1A holds 0 φ ab ab,ac {F} - both games Strong Building Blocs {F} instead of {Φ} Partial Building Blocs {ab} instead of {Φ} 28
If condition 1A fails Scenario I: W ( Φ ) W ( ab, ac) c > c 0 φ ab,ac ab {F} - both games Strong Building Blocs {F} instead of {Φ} Partial Building Blocs { ab, ac} instead of {Φ} Scenario II: W ( Φ ) W ( ab, ac) c < c 0 ab,ac φ ab {F} - both games Stumbling Blocs { ab, ac} instead of {F} Partial Building Blocs { ab, ac} instead of {Φ} 29
12. A linear demand illustration {Φ} φ {F} t Figure 1: Stable agreements under the No FTA game 30
{ab} ab ab ab, ac { F } ; { ab } {F} ac {F} Figure 2: Stable agreements of the FTA game t Figure 1 and Figure 2 together imply that free trade is stable when production technologies are relatively similar across countries. Figures also illustrate when FTAs are partial building blocs and when stumbling blocs. 31
Scenario I: over the darker region in Figure 2, free trade is stable and Scenario II: h{ab}i is stable. Figure 3 compares the two games under scenario I: ab {F } under FTA game only φ {F} under both games t Figure 3: FTAs as Strong Building Blocs 32
Now consider scenario II where h{ab}i obtains. C C B φ ab ab, ac A ac t Figure 4: Building and Stumbling Blocs Region A: stumbling blocs; region B: strong building blocs; regionc:partial building blocs. 33
What can we say about the welfare effects of FTAs? Proposition 10: Under linear demand, the following obtains: (i) when FTAs act as building blocs (strong or partial), the option to form bilateral FTAs benefits low cost countries (as well as the world as a whole) whereas it hurts the high cost country and (ii) when FTAs act as stumbling blocs, all countries lose from being able to form bilateral FTAs. 34
13. Concluding remarks Paper analyzes the debate regarding FTAs in an environment that explicitly models the process of FTA formation and allows each country to form more than asinglefta. Our analysis sharpens the stumbling versus building bloc debate by highlighting conditions under which two effects are likely to obtain and why. There exist circumstances where global free trade obtains as an equilibrium only if countries are free to form bilateral FTAs. Literature has mostly overlooked this. FTAs can deliver welfare improving trade liberalization when multilateral free trade is infeasible. For hub and spoke to emerge, need asymmetry. Have utilized a very simple model of intraindustry trade. Do results extend to other trade settings? 35