Some aspects on regional integration, comparative advantage and consumer's diversity

Abstract

This article describes one simple Cournot oligopoly model with linear inverse demand and international linkage, and tries to analyse how the degree of competitiveness, the diversity in comparative advantage, consumer preference and market volume are closely interrelated with each other in the course of free trade areas (FTA) liberalization. The influence of income/substitution effects to firm's profit maximization is also examined. My analysis shows some basic results and implications regarding the tariff-setting strategies and the incentive for endogenous internal liberalization, as well as the role of FTA formation on a stream to world-wide liberalization.

Acknowledgement

The author is very grateful to the editor and an anonymous referee. Thanks also to Professor Winston Chang for having introduced related literatures in his lecture.

Notes

¹ FTA: free trade area, CU: customs union.

² Here I implicitly assume producers’ behaviour, which tries to maximize their profits, and which would eventually play an important role on each country's realized demand for each good through both the income and substitution effect. For example, Freund (2000) analyses this case, although she uses a linear demand, which is fixed and independent of firms’ profit (or sales) and prices of other goods.

³ Here I am relaxing the assumption of ‘an identical world price’, which implicitly assumes fully competitive markets across countries, reversible and instantaneous flows of goods over a world wide, and therefore the possibility of no-arbitrage for each good.

⁴ This may be somewhat a peculiar definition of ‘comparative advantage’, because ordinarily each country may produce both two goods, but lower marginal cost for the good the country has comparative advantage in. In effect, this is considered to be one extreme case, where the marginal cost of the good it does not comparative advantage in is infinity (∞). Therefore, all the results presented in this article do not lose any generality of economic implications by adopting this special definition.

⁵ and do not exist (or are set to be zero).

⁶ t _xa ¹ = t _yb ² = t _za ³ = t _zb ³ = 0.

⁷ Including this, in principle, I am using the similar notation as defined in Bond et al . (2003).

⁸ d ₂ > 0 and d ₂ < 0 imply that goods a and b are substitutes and complements, respectively.

⁹ CS: consumer surplus, PS: producer surplus, GR: government revenue.

¹⁰ On the other hand, CU's external tariff setting is derived by maximizing the joint welfare of the country with regard to τ, τ′, already set in (2.8) and t = 0.

¹¹ These best reaction functions do not change, in the long-run, even under the existence of some income/substitution effects (d ₁ > 0, ), as far as each firm, as assumed, takes the current inverse market demand, as given.

¹² In CU, the tariff setting for τ, τ′ and τ* can be calculated as follows:

¹³ n = m for ϕ(τ*,t) and n = m + 1 for ϕ*(τ, τ′,t).

¹⁴ In this case (0 <  < ∞), FTA country does not have an incentive to set the internal tariff t below this level, given that τ* is fixed.

¹⁵ Assuming some income/substitution effects (d ₁ > 0, ), the optimal internal tariff depends also on {1 + α,β,γ _F (·,·),γ _R (·,·)}, as well as on {n, m, c}.

¹⁶ is a relative price matrix.

¹⁷ Although, as explained in footnote 11, the income/substitution effects do not alter the form of best reaction functions, they possibly influence the optimal internal tariff .

¹⁸ Note that the smaller (slope) represents the larger market volume.

¹⁹ This is clear also from ∂Σ¹/∂t _za ¹| _{t za ¹( = τ)} > 0 and ∂Σ¹/∂t _zb ¹| _{t zb ¹( = τ')} = 0.

²⁰ See (3.4′) (footnote 12), the best reaction functions in CU. Assuming α = 0, it proves that τ( = ϕ _cu (τ*)) = τ'( = ϕ' _cu (τ*)). Therefore, τ≅τ' can be realized only through quasi-CU-like, cooperative regime, not through a quasi-competitive one like in FTA equilibrium.

²¹ The simultaneous change of external tariffs τ and τ′ in the same direction alleviates the distortion caused by substitution, that is and . Then, from symmetry, .

²² Consider, instead, the case where c = 0, n = m, but and . Then from (3.4) (footnote 12), one of appropriate rules for post-FTA negotiation is τ≅(1 + α)τ' and τ*≅βτ' (t = 0).

²³ If d ₁ is large enough, then .

²⁴ In MA, the tariff setting rule is agreed with by all countries, for example, as .

²⁵ β > 0 and γ = 1 are assumed here.

²⁶ This ‘post-FTA negotiation’ might be proceeded between old FTA and old ROW under the agreement that τ = τ* and t = 0.

²⁷ This is derived by remembering that, in NA, common tariffs should be set at t = τ = τ* = 3/10, and by comparing it with the results in FTA and CU.

²⁸ Then in this symmetry case, proves to be always optimal for each country.

²⁹ γ = 1/2 implies that old FTA and old ROW are the same in market volume.

³⁰ The larger β (intercept) represents the larger consumer preference, and the smaller γ (slope) represents the larger market volume.

³¹ ϕ(t,τ*) is a function only of t, and ϕ*(t,τ) is constant over t and τ.

³² So, the internal tariff reduction among FTA members always makes ROW better off.