Will a regional bloc enlarge?

Transcription

1 Will a regional bloc enlarge? Giorgia Albertin International Monetary Fund May 22, 2006 Abstract The recent and unprecedented spread of regionalism stimulated a buoyant debate on whether regionalism would lead to successive multilateral liberalization or, instead, to the fragmentation of the worldwide economy. This paper investigates whether a regional bloc would expand or remain stagnant in terms of its size by building a theoretical model where the enlargement of a regional bloc is the endogenous outcome of the interaction between the supply of, and demand for membership. We show that a maximum size of the regional bloc exists beyond which the regional policy-maker will prevent any further enlargement to avoid a loss in its political support. Furthermore this paper investigates how the deepening of integration within a regional bloc a ects its width. We show that deeper integration contracts the supply-side implied maximum size of the bloc while it boosts the demand of membership from third countries. As a result, we point out that deeper integration may lead to wider integration when the demand-side of membership is binding in determining the equilibrium size of the regional bloc, while the equilibrium size of the bloc will be una ected when the supply-side is binding. I thank Tony Venables, Steve Redding and all the participants in the International Economics seminar at the London School of Economics for their comments and suggestions. I am also really grateful to Frédéric Robert-Nicoud and Gianmarco Ottaviano for their useful comments.

2 Introduction The unprecedented wave of regionalism experienced since the early 990s has stimulated a buoyant debate on whether regionalism provides an impetus, or detract from the goal of multilateral non-discriminatory freeing of trade. ;2 This paper contributes to the debate by addressing the question of whether there are incentives for a regional trading bloc to enlarge through further extensions of its membership, or not to enlarge thus rejecting new membership requests and remaining stagnant in terms of its size. Notably, a regional trading bloc which enlarges through further extensions of its membership can be thought as a building bloc toward the multilateral freeing of trade while a bloc which does not enlarge can be seen as stumbling bloc to the goal of multilateralism. The contributions that have so far investigated the enlargement of regional trading blocs have taken two fundamentally di erent approaches. One strand of the literature has depicted the expansion of a regional trading bloc as exogenously determined, and focused on the implications of the enlargement on aggregate social welfare. Thus, in his seminal contribution Krugman (99) showed that in a world divided into symmetric trading blocs, an exogenous increase in their absolute size may reduce the world welfare. Bond and Syropolus (996) generalized Krugman (99) taking into account the existence of comparative advantages, and showed that an exogenous increase in the size of trading blocs may increase world welfare. A second strand of the literature has instead modelled the enlargement of a regional trading bloc as endogenously determined. In this context, the main contribution is Baldwin (995) who focused on the demand-side of the enlargement, and formalised the incentives of non-members countries to join a regional trading bloc assuming the supply of membership of the bloc as perfectly elastic. 3 However, the assumption that any country asking for membership in the bloc would be admitted is strongly counter factual since regional blocs seem to be closed clubs in which members decide whether or not a new country should be admitted, as the process of enlargement of the European Union points out. The novelty of this paper is to model the supply-side of membership of a regional trading bloc by formalising the scheme of incentives that underpin the decision to accept or reject new membership requests, and to integrate this supply-side with a demand-side à la Baldwin (995) to formalise the endogenous enlargement of a bloc as the result of the interaction between the supply of, and the demand for membership. Assuming a political economy approach à la Grossman-Helpman (994), we show that pro and anti-enlargement forces operate in the regional trading bloc, and the regional policy-maker s decision regarding whether to enlarge or not Since 990 more than 250 new regional trade agreements are reported to have been noti ed to the General Agreement on Trade and Tari s and the World Trade Organization. 2 See Winters (996), and Baldwin and Venables (997) for useful surveys of the literature in the regionalism versus multilateralism debate. 3 Recently, Alesina et al. (2005) investigated the related but di erent issue of the enlargement of an international political union modelled as a group of countries deciding together the provision of certain public goods and policies because of spillovers. 2

3 is a political equilibrium which balances these opposing forces. In this regard, we point out that the regional policy-maker will prevent the enlargement of the regional trading bloc beyond a size, labelled as the supply-side implied maximum size, to avoid a loss in the political support it receives. Furthermore, we formalise the enlargement of the regional trading bloc as the endogenous outcome of the interaction between the supply of membership of the bloc and the demand for membership arising from third countries. At this regard, we show that if the size of the regional trading bloc is smaller than the supply-side implied maximum size, the bloc will enlarge in the event of a request for membership from third countries. However, once the regional trading bloc has reached the supply-side implied maximum size, the regional policy-maker will not be willing to further enlarge the bloc such that any eventual request of membership from third countries will not be accepted. As a result, we point out that the equilibrium size of the regional trading bloc can not exceed but could be smaller than the supply-side implied maximum size. In fact, the equilibrium size of the regional trading bloc can not exceed the supply-side maximum size since the supply-side of membership will be binding on further enlargements beyond this size in case of a request for membership from third countries. However, the equilibrium size of the bloc might be smaller than the maximum size the regional policy-maker would be willing to achieve if the requests for membership are not numerous enough. Thus, if the supply-side implies a maximum size beyond which the regional trading bloc will not further enlarge, the demand-side of membership might be binding on further enlargements of the bloc, preventing it from reaching the maximum size that the regional policy-maker would be willing to achieve. Finally, we formalise the implications of the implementation of deeper integration among the members of a regional trading bloc on equilibrium size of the bloc. Thus, we show that deeper integration contracts the supply-side implied maximum size of the bloc that the regional policy-maker is willing to achieve, while it boosts third countries demand of membership. We point out that, as a result, the implications of deeper integration on the equilibrium size of a regional trading bloc are ambiguous and crucially depend on whether the supply-side or the demand-side are binding in the determination of the equilibrium size of the bloc. Notably, we show that deeper integration might lead to wider integration only when the demand-side of membership is initially binding, while the equilibrium size of the regional trading bloc will be una ected if the supply-side is initially binding. 2 The basic model We consider that the world economy is constituted by g countries, h of which are members of a regional trading bloc. 4 4 The theoretical framework we develop can also be used to formalise the formation of the regional trading bloc in the rst place, once we assume that none of the countries is initially in the bloc. 3

4 To simplify our framework, we assume that the regional trading bloc is unique, and we rule out the possibility of formation of other regional trading blocs. Thus, countries which are not members of the bloc, labelled as third countries, may ask for membership in the bloc but can not organize in any alternative form of preferential trade agreement. We assume that countries are symmetric, and are characterized by the existence of a manufacturing sector and an agricultural sector. The manufacturing sector is characterized by di erentiated products, increasing returns to scale and imperfect competition, while in the agricultural sector a homogenous product, constant returns, and perfect competition exist. In order to capture a feature of real economies, we consider that there are two di erent classes of agents in each country, labourers and rm owners, whose preferences, U L and U F, respectively, are: U L ( ) = C A CM () where U F ( ) = C A C M (2) C M = " X i # ( ) c ( )= i Thus, consumers have a love for variety, c i is the consumption of the manufactured variety i, is the elasticity of substitution between any two varieties, with >,while 0 and 0. Notably, labourers income is assumed to derive from the labour they provide to rms while rm owners income derives from rms pro ts. Focusing on technology, we assume that in the manufacturing sector the labor input requirement for a variety is: l i = + x i with and > 0 (3) where x i is the output of variety i, and is a xed cost. We consider that the number of manufacturing rms in any country is given and equal to k, new entry is ruled out, and each rm is su ciently small in the market to treat market aggregates as exogenous. 5 ;6 In addition, there is no 5 Market behavior in the manufacturing sector is like monopolistic competition but without free entry. See Baldwin (995), and Desrouelle and Richardson (997) for a similar approach. 6 Ruling out new entry guarantees that manufacturing rms have positive pro ts that can be used in lobbying activities. Alternatively, we could have assumed free entry and the existence of speci c factor of production owned by rm owners. In this case even if rms pro ts would be zero, rm owners would have a positive income to be used in lobbying activities. 4

5 possibility of relocation for manufacturing rms and that each rm is wholly owned by the residents of the country in which it is located. We also assume that in the agricultural sector the production function is linearly homogenous, the market structure is perfectly competitive, and units of the agricultural good are chosen such that the unit labour coe cient is unity. Finally, we consider that international trade in manufactured varieties is costly while trade in the agricultural good is costless. Notably, iceberg trade costs are assumed such that shipping manufactured varieties between any two countries melts a fraction of the shipment while no trade costs apply on domestic sales of manufactured varieties. Trade costs are assumed to be lower for ows of manufactured varieties arising between members of the regional trading bloc than for all other trade in manufactured goods. Thus, all trade in varieties between members of the regional trading bloc requires > units to be shipped for every unit sold while trade in varieties that does not arise within the bloc requires > units to be shipped for every unit sold, with >. 3 The equilibrium Since trade in the agricultural good is costless, the wage will be equalized across countries as long as the agricultural good is produced in any country, which is assumed henceforth. 7 Taking labour as the numeraire, labourers wage in any country will be equal to unity. Normalizing national workforces to unity, the aggregate income of labourers located in country j is: E L j = (4) Since rms owners income derive from rm pro ts and k manufacturing rms operate in each country, the aggregate income of rm owners located in country j is: E F j = k j (5) where j are the pro ts of a manufacturing rm located in country j. Since all manufactured varieties enter consumers demand symmetrically, and assuming the cost of introducing a new variety to be zero, no two rms will produce the same variety. There will be then only one rm producing a given variety, and no duplication will arise across countries. In addition, since 7 The non-full-specialisation (NFS) condition requires that no country has enough labour to satisfy the world demand for the agricultural good, i.e. that the world spending on this good is larger than the maximum value of its production that is possible in any of the countries. Given that the g countries are assumed to be economically symmetric, the NFS condition requires that g ( ) E L + ( ) E F > p A L, which is assumed to hold henceforth. 5

6 rms share the same manufacturing technology, all varieties produced will be symmetric. Denoting the mill or f.o.b price of a typical manufactured variety produced in country j as p j, we label as T ji the amount dispatched per unit received of a variety shipped from country j to country i, with T ji >. The iceberg transport costs imply that if the variety produced in country j is sold at price p j, the delivered or c.i.f. price at the consumption location i is p j T ji. Given that k symmetric varieties are produced in any country, the manufactured composite price index of consumers in country i is: 2 gx P i = 4 k (p j T ji ) j= ( ) Focusing on consumers utility maximization, in order to simplify the computations of the equilibrium demand patterns, we assume to be zero and to be strictly between unity and zero. 8 Therefore, given equations (), (4), and (6), the consumption demand in country i for any manufactured variety produced in country j is: c i = (p j T ji ) ( ) P ( ) i Thus, the aggregate demand of the manufactured composite good in country i is: 3 5 (6) (7) C M i = P i (8) Finally, given equations (), (2), (4), (5) and (6), the aggregate demand for the agricultural good in country i is: C A i = ( ) + k i p A (9) where p A is the price of the agricultural good A, and i indicates the pro ts earned by a typical manufacturing rm in country i. Turning to the supply, since the agricultural sector is perfectly competitive, the price of the agricultural good will re ect its marginal cost. Having chosen units of good A such that the labor input coe cient is unitary, the price of the agricultural good A in any country is: p A = Focusing on the manufacturing sector, the typical pro t-maximising rm located in country j faces an isoelastic demand curve expressed in equation (7) but due to trade costs to supply this amount of consumption, T ji times this 8 Our nal results would not change if we assumed to be strictly between zero and one. 6

7 amount has to be shipped. Summing up across countries, the total sales of the typical manufacturing rm located in country j is: x j = gx (p j T ji ) ( ) T ji i= ( ) P i (0) Since each pro t-maximising manufacturing rm sets its price taking the price index as given, the f.o.b. price of any variety produced in country j is a constant mark-up over the marginal cost: p j = () It follows that assuming the unit input coe cient just equals = ( ), all rms will charge the same unitary price. Given equations (0) and (), the operating pro ts (gross of xed cost) of the typical manufacturing rm in country j are: j = gx i= ( ) (T ji ) ( ) P i (2) Taking into account the asymmetric nature of trade costs, we can derive the equilibrium for the representative member of the regional trading bloc and the representative third country. The delivered prices or c.i.f. prices paid by consumers for manufacturing varieties will vary depending on consumers location. Thus, consumers in the representative member country will pay a price for any unit of manufactured variety produced in the regional trading bloc, and a price for any unit of manufactured variety produced in any third country. Consumers located in the representative third country will instead pay a price for any unit of variety produced in any other country, either member or third country. Given equation (6), the manufactured composite index price of consumers located in the representative member country and in the representative third country, P R and P N, respectively, are: P R = hk + k (h ) ( ) ( + k (g h) )i P N = ( hk + k (g ) )i (3) (4) It follows that, given equations (8), (3) and (4), the aggregate consumption of the manufactured composite good in the representative member country and in the representative third country, CR M and CM N, respectively, are: C M R = P R (5) C M N = P N (6) 7

8 Given equations (9), (3) and (4), the aggregate demand for the agricultural good in the representative member country and in the representative third country, CR A and CA N, respectively, are: C A R = ( ) + k R (7) C A N = ( ) + k N (8) Finally, given equations (2), (3) and (4), the pro ts of any manufacturing rm located, respectively, in the representative member country and in the representative third country, R and N, are: R = h ( ) P R + (h ) ( ) ( ) P R + (g h) ( ) P ( ) N i (9) N = h ( ) P N + h ( ) ( ) P R + (g h ) ( ) P ( ) N i (20) 4 The supply-side of membership To date the main contribution on the endogenous enlargement of regional trading blocs is Baldwin (995) who considered a bloc s supply of membership to be perfectly elastic. However, assuming that any country asking for membership of a regional trading bloc will be admitted appears to be counter factual since regional trading blocs seem to be closed clubs in which member countries decide whether or not a new country should be admitted. We develop a more realistic framework where a regional trading bloc may choose whether to admit or not a third country requesting for membership. 9 Thus, we formalise a regional trading bloc s supply of membership by focusing on the scheme of incentives underpinning the choice between accepting or rejecting third countries membership requests, under the assumption that the demand for membership is perfectly elastic. This assumption will be relaxed in section 6 where we will show how the demand-side of membership might be binding on further enlargements of the bloc. We consider that decisions in the regional trading bloc are centralized such that a regional policy-maker exists, faced with two trade policy options: to enlarge the bloc by admitting a third country or not to thus refusing the new membership request. 0 9 We assume that the eventual choice of the regional policy-maker to admit a country in the bloc is irreversible such that once a third country is admitted into the regional trading bloc it can not be forced to leave it later on. 0 This modelling choice is suggested by the existence of centralised political organs in the European Union. 8

9 Thus, we focus on the regional policy-maker s choice regarding a marginal enlargement of the regional trading bloc, that is an enlargement that would increase the size of the regional bloc by one additional member country. Relying on Grossman and Helpman (994), we consider that the regional policy-maker shapes its trade policy taking into account not only aggregate well-being but also the political contributions received from an organised interest group which participates in the political process to in uence policy outcomes. 2 Notably, the regional policy-maker trades o the contributions that would come from heeding the lobby s interests against the reduction in regional aggregate social welfare that would follow the choice of a socially costly trade policy. 3 All manufacturing rms in the regional bloc are assumed to be organized in a unique interest group that o ers a schedule of contingent (implicit) donations to the regional policy-maker to a ect its choice of trade policy. Notably, we assume that the organized industrial lobby speci es a donation contract or contribution schedule that stipulate how large a donation will be made for each of the two possible stances of trade policy open to the regional policy-maker, tailoring its contribution schedule to maximise the total welfare of its members, net of contributions. 4 The game is in two stages: in the rst stage, the donation contracts are announced by the organized interest group to the policy-maker while in the second stage the policy-maker sets the trade policy and collects the donations. Political contributions paid by the organized interest group to the policy-maker are then ex-post, that is they are paid after the policy-maker has chosen whether to enlarge or not. The regional policy-maker will decide whether to enlarge the bloc or not with the aim of maximizing the political support received which is assumed to depend positively on the political contributions obtained from the organised interest group, and on the aggregate social welfare, net of contributions, achieved 5 ;6 in the regional trading bloc. Our theoretical framework could be extended to investigate the choice of the regional policy-maker to enlarge the regional trading bloc by admitting more than one new member. 2 Recently, a number of studies have provided empirical evidence in support of the Grossman and Helpman s political economy approach to the formation of trade policy. At this regard see Goldberg and Maggi (999), Gawande and Bandyopadhyay (2000), Mitra et al. (2002), Eicher and Osang (2002), McCalman (2004), and Gawande and Krishna (2005). 3 Grossman and Helpman (994) noted that an incumbent policy-maker may make trade policy choices while being aware that its decisions may a ect its chances for re-election. Thus, the policy-maker may value political contributions since they can be used to nance campaign spending, and aggregate social welfare since voters are more likely to reelect a government which has delivered a high standard of living. As in Grossman and Helpman (994), we do not explicitly formalise the existence of an electoral process. See Grossman and Helpman (996) for a model with electoral competition where interest groups may use campaign contributions to in uence the outcome of the election, and the competing parties platforms. 4 We assume that consumers are not organised in any form of interest group. 5 See Grossman and Helpman (2002) for a detailed survey of alternative political economy approaches to the formation of trade policy. 6 We rule out the possibility of side-payments paid to the regional policy-make by organised 9

10 Labelling as D the political contributions that the regional policy-maker receives from the lobby group, and as W the net aggregate social welfare reached in the regional trading bloc as a whole, the objective function of the regional policy-maker is: G = a D + ( a) W (2) where a measures the extent of political distortion in the policy-making process in the regional trading bloc, with 0 a. Thus, when a is equal to zero the regional policy-maker will behave like a social welfare maximiser, while the greater is a, the greater is the weight that the interests of manufacturing rms receive in the policy-making process. The regional policy-maker will achieve di erent levels of political support depending on whether it chooses to enlarge the bloc or not to since the political contributions received and the regional aggregate social welfare will depend on the chosen stance of trade policy. 4. Aggregate social welfare: a pro-enlargement force We de ne the aggregate social welfare in the regional trading bloc as the sum of the aggregate welfare in any member country, in turn assumed to be the sum of the indirect utilities of the agents in the economy. So, the aggregate social welfare in any member country is the sum of labourers and rm owners aggregate indirect utilities. The aggregate social welfare achieved in the regional bloc if the regional policy-maker decided to enlarge or not to, labelled as f W ENL and f W NON, is: fw ENL = h ( ) ( ) P ( ) ENL + kh ENL (22) fw NON = h ( ) ( ) P ( ) R + kh NON (23) with: P ENL = hk + k h ( ) ( + k (g h ) )i (24) Comparing the regional aggregate social welfare that would be reached under the two alternative stance of trade policy, we can state the following proposition: Proposition The aggregate social welfare in the regional trading bloc will be greater if the policy-maker chooses to marginally enlarge rather than not to. Proof. See Mathematical Appendix for proof. Thus if the regional policy-maker behaved like a social welfare maximiser, it would choose to enlarge since this choice would imply a higher level of regional aggregate social welfare than the alternative of not enlarging. The choice of enlarging constitutes then the socially optimal choice of trade policy for any initial size of the regional bloc. interest groups located in third countries. 0

11 4.2 Lobbying: pro or anti-enlargement force? The organized industrial lobby speci es a contribution schedule which stipulates how large a donation will be made for each possible stance of trade policy open to the regional policy-maker. 7 Thus, the interest group s contribution schedule will comprise two items, D ENL and D NON, which are the political contributions associated with the regional policy-maker s choice to marginally enlarge the bloc or not to enlarge, respectively. 8 Following Grossman-Helpman (994) we restrict the lobby s contribution schedule to be truthful in the Bernheim-Whinston (986) speci cation such that the contribution schedule everywhere re ects the true preferences of the organised interest group. The assumption of truthful donation contracts implies that the contingent contribution schedule speci ed by the interest group will assume a speci c form. The interest group will pay to the regional policy-maker the excess, if any, of the lobby s gross welfare under the speci ed stance of trade policy, relative to an optimally chosen base level of welfare. Since the organized interest group is constituted by rm owners located within the regional trading bloc, the gross aggregate welfare of the group, labelled as V gf, is the sum of the gross indirect utility of rms owners located in the regional trading bloc. Thus, restricting political contributions to be non-negative, a truthful contribution function can be formally expressed as: h D = max 0; V g i F B where the organised interest group will choose B as such as to satisfy the voluntary participation constraint of the regional policy-maker. The truthful contribution function evidences that the organised interest group s net welfare will be equal to B whenever the group makes a positive contribution to the regional policy-maker. The interest group will then wish to make B as large as possible but without inducing the regional policy-maker to choose a policy outcome which is damaging to the group s interests. Since only one organised interest group operates in the bloc, the interaction between the regional policy-maker and the industrial lobby con gures as a principal-agent problem such that the regional policy-maker s voluntary participation constraint can be used to derive the lobby s choice of B in equilibrium. 9 7 We rule out the possibility for the interest group to o er contributions to foreign governments. See Grossman and Helpman (994), and Grossman and Helpman (995) for a similar approach. 8 It will never be optimal for the interest group to specify a positive contribution for both policy outcomes since then it could cut back equally on both of its o ers without a ecting the regional policy-maker s decision, and also since it will not wish to give the policy-maker an incentive to choose the trade policy outcome that it is contrary to the lobby s interests. 9 See Grossman and Helpman (994) for the derivation of the lobby s choice of B when di erent organised interest groups compete for protection in a common agency framework.

12 The interest group will choose B in order to make the regional policy-maker just indi erent between heeding the lobby s interests regarding enlargement, and the policy outcome that the regional policy-maker would implement in absence of any contributions, that is the socially optimal choice of trade policy. Since the interest group will favour the stance of trade policy that implies the greater group s gross welfare, we compare the gross welfare that interest group would achieve if the regional policy-maker decided to marginally enlarge the bloc or, alternatively, not to. The gross aggregate welfare that the organised interest group would achieve if the policy-maker decided to marginally enlarge the regional trading bloc or not to, labelled as ^V ENL F and ^V NON F, respectively, are: ^V F ENL = kh ENL (25) ^V F NON = kh NON (26) where ENL and NON indicate the pro ts that would be earned by any manufacturing rm in the bloc if the policy-maker decided to enlarge the bloc or not to enlarge. Having assumed the contribution schedule to be truthful, and given equations (22) and (23), the political contributions that the interest group would o er to the regional policy-maker for the two alternative stances of trade policy can be expressed as: D ENL = max [0; kh ENL B] (27) D NON = max [0; kh NON B] (28) The organised interest group will favour the stance of trade policy that implies the greater aggregate pro ts for the manufacturing rms located in the regional trading bloc. Thus, to evaluate whether the organized interest group constitutes a proor anti-enlargement force in the policy-making process, we compare the pro ts that a typical manufacturing rm located in the regional trading bloc would earn under the two alternative stances of trade policy. The pro ts a manufacturing rm located in the regional bloc would earn if the regional policy-maker choose to marginally enlarge or not to, respectively, are: ENL = h P ENL + h( ) P ENL + (g h ) PN ( )i (29) NON = h P R + (h ) ( ) P R + (g h) P N ( )i (30) 2

13 If the regional policy-maker chooses to marginally enlarge, a manufacturing rm located in the regional trading bloc would face di erent e ects in the three markets where it operates, that is the local market, the regional bloc market, and the third countries market. First, in an enlarged regional trading bloc, the rm would face greater competition in its local market since the price local consumers would pay for any manufactured variety produced in the new member would decrease from to. So due to a substitution e ect local consumers would reduce their consumption of local varieties, and the rm would experience an unambiguous pro ts reduction on its local market. Second, the rm would experience greater competition in the regional bloc market since the price consumers located in the bloc would face for varieties produced in the new member country would reduce from to. On the other hand, the rm would enjoy an enlarged regional bloc market since the size of the regional bloc market would increase from (h ) to h. As a result, the price paid for the rm s own variety by consumers located in the new member would decrease from to such that they will increase their consumption of the rm s own variety. These e ects would then have opposing implications since while the competition-e ect would reduce rm s pro t, the market-size e ect would increase it. Finally, the rm would face an unambiguous reduction in the pro ts earned in the third countries market since the size of this market would contract from (g h) to (g h ). Taking into account the pro t implications on the three di erent markets where the manufacturing rm operates, we can state the following proposition: Proposition 2 A size of the regional trading bloc, labelled as b h, exists that maximizes the pro ts earned by any manufacturing rm located in the regional trading bloc. Proof. See Mathematical Appendix for proof. A marginal enlargement of the bloc could then be either pro t-enhancing or pro t-reducing for manufacturing rms located within the bloc depending on its initial size. If the initial size of the bloc is smaller than the size that maximizes manufacturing rms pro ts, the policy-maker s choice of enlarging the bloc would imply greater aggregate pro ts for manufacturing rms located in the bloc than the alternative of not enlarging. On the other hand, if the initial size of the bloc is greater than or equal to the size that maximizes manufacturing rms pro ts, the policy-maker s choice of enlarging would imply a lower aggregate manufacturing pro ts than the choice of not to enlarge. We can then state the following corollary: Corollary 3 The organized interest group is a pro-enlargement force if the size of the regional trading bloc is such that h < b h while it is an anti-enlargement force if h b h. 3

14 The organized industrial interest group may act either as a pro- or antienlargement force in the policy-making process depending on the size of the bloc. If the size of the bloc is smaller than b h, the interest group represents a proenlargement force since this stance of trade policy would lead to a greater group s gross welfare. However, the interest group will actually contribute in equilibrium only if its own interests are in con ict with the trade policy outcome that the regional policy-maker s would choose in absence of any political contributions. Thus, for any size of the bloc smaller than b h, since the lobby s preferred policy outcome coincides with the socially optimal outcome, the interest group will contribute nothing in equilibrium. Instead if the size of the bloc is at least equal to b h, the interest group constitutes an anti-enlargement force, since this stance of trade policy would imply a greater group s gross welfare. Thus, the group s interests are in con ict with the socially optimal policy outcome, and it will o er a positive contribution to the regional policy-maker not to enlarge insofar as the group s net welfare is at least equal to the group s gross welfare under the socially optimal outcome, and a zero contribution otherwise. 4.3 The supply-side political equilibrium The choice of the regional policy-maker on whether marginally enlarge the bloc or not to is a political equilibrium that balances pro and anti-enlargement forces. If the size of the regional trading bloc is smaller than h, b the regional policymaker faces no trade-o in its choice of trade policy. In fact, the greater regional aggregate social welfare that could be achieved when enlarging operates as a proenlargement force in the policy-making process. Furthermore, the organised interest group represents a pro-enlargement force but it will contribute nothing in equilibrium since its preferred outcome coincides with the socially optimal one. Thus, the regional policy-maker will choose to enlarge and the interest group s contribution schedule will be DENL = D NON = 0. On the other hand, if the size of the regional bloc is greater than or equal to b h, the regional policy-maker faces a trade-o between heeding the industrial lobby s interests and the socially optimal outcome. In fact, the interest group is an anti-enlargement force, and it o ers a positive political contribution to induce the regional policy-maker not to enlarge, insofar as the resulting group s net welfare is at least equal to the group s gross welfare under the social optimum. Thus, given the political-support maximising behavior of the regional policymaker and the lobbying activity of the organised interest group, we can state the following proposition: Proposition 4 The regional policy-maker will choose to marginally enlarge the regional trading bloc if and only if the following condition is satis ed: a ( ENL NON ) + ( a) ( ) ( ) P ENL P R > 0 (3) 4

15 Proof. See Mathematical Appendix for proof. We label condition (3) as the supply-side condition since when it holds the regional policy-maker will choose to enlarge the bloc,while when it does not hold it will choose not to enlarge. Thus, if the size of the regional bloc is greater than or equal to b h and the supply-side condition holds, in equilibrium the regional policy-maker will choose to enlarge and the interest group s contribution schedule will be DENL = DNON = 0. However, if the supply-side condition does not hold, in equilibrium the regional policy-maker will choose not to enlargeand group s contribution schedule would be DNON = a 2a fwenl WNON f and DENL = 0. Then if the supply-side condition holds, a positive supply of membership arises since the regional policy-maker will be willing to further enlarge. However, if the supply-side condition is not veri ed, the supply of membership will be nil and any third countries request for membership will be refused. At this regard, we can state the following proposition: Proposition 5 If the political distortions in the policy-making process are strong enough, a size of the regional trading bloc, labelled as h, does exist beyond which the regional policy-maker will not be willing to further enlarge the bloc. Proof. See Mathematical Appendix for proof. While the supply-side condition is always veri ed for any size of the bloc smaller than h, b if the extent of political distortions is high enough, there will be a size of the bloc h larger than h, b for which the supply-side condition will no 20 ;2 longer be veri ed. Thus, for any size of the bloc smaller than h, the regional policy-maker will choose to marginally enlarge since this stance of trade policy implies a greater political support. However, when the size of the regional bloc has reached h, the regional policy-maker will choose not to marginally enlarge to avoid a loss in the political support received. We de ne h as the supply-side implied maximum size of the regional trading bloc, that is the equilibrium size that bloc would reach if the demand of membership arising from third country was perfectly elastic. In the following, we use graphical analysis to depict the political equilibrium of the supply of membership from the regional trading bloc. In Figure, the locus SS plots the right-hand side of the supply-side condition as a function of the size of the regional trading bloc. Notably, the locus SS is downward sloping since both the pro t and the price index di erential are decreasing in the size of the bloc. Corresponding with h, at the intersection between the locus SS and the horizontal axis, the supply-side condition will no longer be veri ed. Thus, h represents the maximum size to which the regional policy-maker would be 20 In the rest of the analysis we consider that political distortions in the policy-making process are strong enough to guarantee that h does exist. 2 If the regional policy-maker only cared the organised group s interest, i.e. a =, we would have that h = b h. 5

16 Figure : The supply-side political equilibrium SS SS h* Size of the bloc willing to enlarge the bloc since enlarging beyond h would imply a loss in political support received The demand-side of membership We follow Baldwin (995) in modelling the demand-side of membership thus assuming that the policy-maker in any third country considers the supply of membership as perfectly elastic. We will relax this assumption in section 6 where we will show how the supply-side of membership might be binding on further enlargements of the bloc. The policy-maker in any third country is assumed to be a political support maximiser a la Grossman-Helpman (994), and facing two alternative options of trade policy, that is joining the regional trading bloc or not to. The political support received by the policy-maker depends positively on the political contributions received from an organized interest group which represents local manufacturing rms, and o ers a schedule of contingent (implicit) donations to the policy-maker to a ect its choice of trade policy. 23 In addition, the political support obtained by the policy-maker depends positively on the net aggregate social welfare reached in the third country, and on the support of those groups that oppose or, alternatively, sustain joining the bloc on 22 More precisely, respecting the integer constraint, the equilibrium size of the regional trading bloc under the assumption of a perfectly elastic demand of membership is the highest integer that is lower than h. 23 We rule out the possibility for the interest group to o er contributions to the incumbent policy-maker in the regional trading bloc. In addition, we assume that consumers located in any of the third countries are not organised in any form of interest group. 6

17 non-economic grounds. 24 In any third country, the industrial lobby speci es a contribution schedule that stipulates how large a donation will be made for each of the two stances of trade policy open to the policy-maker. We restrict the contribution schedule speci ed by the organized interest group to be truthful in the Bernheim- Whinston (986) speci cation such that the organised group will o er the policymaker the excess, if any, of the group s gross welfare reached under the chosen stance of trade policy relative to some base level of welfare. Having assumed that in any third country the organized interest group is constituted by rm owners, the political contributions o ered to the incumbent policy-maker take the form: D IN = max [0; k IN C] (32) D OUT = max [0; k OUT C] (33) where C is a scalar chosen such as to satisfy the voluntary participation constraint of the incumbent policy-maker, and IN and OUT are the pro ts that any manufacturing rm would earn if the policy-maker decided to join the regional trading bloc, or not to. The organized interest group will favour the trade policy option that implies the higher level of group s gross welfare. Notably, the di erence in the pro ts that a manufacturing rm in a third country would earn if the policy-maker decided to join the bloc or not to, is: IN OUT = h P R P N + h ( ) ( ) i P ENL (34) First, the rm s pro ts on its local market would reduce if the policy-maker decided to join the bloc rather than not to since local consumers would pay a lower price for varieties produced in the bloc thus reducing their consumption of local varieties. Second, rm s pro ts in the regional bloc market would be higher if the policy-maker choose to join the bloc since the rm s variety would be available to consumers in regional bloc at the lower price such that they would increase their consumption of the this variety. 25 The loss in pro ts that the manufacturing rm would experience in its local market will be more than compensated by the gain in pro ts in the regional 24 Modelling a resistance to membership on non-economic grounds capture real world political concerns and enables to depict an equilibrium where some third countries will choose to enter the regional trading bloc while others will not. We could introduce the existence of resistance or willingness to enlarge the regional bloc on non-economic grounds but it would only rescale the relative strength of pro- and anti-enlargement forces. 25 The pro t earned by the manufacturing rm in the third countries market will be unchanged whether or not the country joins the regional bloc. 7

18 bloc market such that the pro ts di erential will be positive for any size of the regional trading bloc. Since it is pro t-enhancing for manufacturing rms located in any third country if the policy-maker chooses to join the regional trading bloc, the organized interest group constitutes a pro-membership force in the policy-making process. However, the organised interest group in any third country will actually contribute when its own interests are in con ict with the trade policy outcome that the incumbent policy-maker s would implement in absence of any political contributions. Thus, the organised interest group will o er a positive contribution to the policy-maker in order to join the regional trading bloc insofar as the noneconomic resistance to join would dominate economic considerations to join in absence of political contributions. Focusing on the aggregate social welfare reached in a third country, it is de ned as the sum of labourers and rm owners aggregate indirect utilities. Thus, the gross aggregate social welfare that would be reached if the policymaker choose to join the regional trading bloc or not to, respectively, are: fw IN = ( ) ( ) P ( ) R + k IN (35) fw OUT = ( ) ( ) P ( ) N + k OUT (36) Notably, the aggregate social welfare reached in the representative third country will be greater if the policy-maker decides to join the regional trading bloc rather than not to join so representing a pro-membership force in the policy-making process in any third country. As shown in Baldwin (995), the policy-maker in a third country will decide to join the regional trading bloc if the following condition is veri ed: b ( IN OUT ) + ( b) P R P N R (37) where b 2 [0; ] is a parameter which measures the extent of political distortions in the policy-making process, and R is the non-economic resistance to join the bloc. 26 We label this condition as the demand-side condition since when it holds the policy-maker will choose to join the regional trading bloc, while when it is not satis ed it will choose not to join the bloc The right-hand side of the demand-side condition will assume a positive value for those countries that have a resistance to join the regional trading bloc on non-economic grounds, and negative value for those countries that have a willingness to join the regional trading bloc. 27 Having assumed that some countries are characterized by a willingness to enter the regional trading bloc on non-economic grounds implies that for some countries the demand-side condition will be always veri ed, irrespectively of the size of the regional trading bloc. In turn, this guarantees that the regional trading bloc will be formed in the rst place. 8

19 Figure 2: The demand-side political equilibrium DD, RR RR DD h D Size of the bloc We use graphical analysis to derive the demand-side political equilibrium. In Figure 2 locus DD plots the right-hand side of the demand-side condition, and it is upward sloping since both the pro t di erential and the index price di erential are increasing in the size of the regional trading bloc. In addition, arranging countries in order of increasing resistance to membership on noneconomic grounds, locus RR plots the resistance to membership of each country. The size of the bloc at which the two loci DD and RR intersect, labelled as h D, is the equilibrium size that the regional bloc would reach if its supply of membership was perfectly elastic. 28 In fact, for all countries on the left of h D, the demand side condition would be veri ed such that they will join the regional bloc while for all countries on the right of h D, the demand side condition will not be veri ed so that they will not to join the bloc. 6 The equilibrium size of the regional bloc In this section we allow for the interaction between the supply of membership of the regional trading bloc, and the demand for membership arising from third countries, and formalise the equilibrium size of the regional trading bloc. De ning the equilibrium size of the regional trading bloc as the size beyond which no further enlargements of the bloc will take place, we can state the following proposition: 28 More precisely in order to respect the integer constraint, the equilibrium number of countries that will ask for membership in the regional trading bloc, given a perfectly elastic supply of membership, is the highest integer less than h D. 9

20 Proposition 6 The equilibrium size of the regional trading bloc will be reached when either the supply-side condition or the demand-side condition are not veri ed. When the supply-side and the demand-side of membership interact, an enlargement of the regional trading bloc will arise if and only if the supply-side condition is veri ed, and the demand-side condition holds for at least one of the third countries. Thus, the regional trading bloc will reach its equilibrium size when either the regional policy-maker is no longer willing to further enlarge the bloc so that any third countries request for membership is refused, or none of the third countries is willing to join despite the regional policy-maker might be willing to further enlarge. As a result of the interaction between the supply of and the demand for membership the equilibrium size of the regional bloc will not exceed but could be smaller than the supply-side implied maximum size of the bloc. Notably, while the supply-side of membership implies that a maximum size beyond which the regional trading bloc will no further enlarge exists, the demand-side of membership might imply that the supply-side implied maximum size is not reached in equilibrium. In the following, we use graphical analysis to derive the equilibrium size of the regional trading bloc, and show how the supply of, or the demand for membership might binding on further enlargements of the bloc. First, in Figure 3 we depict the supply-side political equilibrium, and the demand-side political equilibrium, and we show how the supply-side of membership might be binding in the determination of the equilibrium size of the regional trading bloc. In the case depicted in Figure 3, the number of countries that would rather join the regional trading bloc if the supply of membership was perfectly elastic exceeds the supply-side implied maximum size of the bloc, that is h D > h. However, the equilibrium size of the regional trading bloc can not exceed the supply-side implied maximum size since the regional policy-maker will not be willing to further enlarge the bloc beyond h. Thus, due to the interaction between the supply of and demand for membership, the equilibrium size of the regional trading bloc, labelled as h E, coincides with the supply-side implied maximum size of the bloc, that is h E = h. The number of countries that in equilibrium are in the regional trading bloc is then smaller than the number of countries that would have entered the bloc if the supply of membership had been perfectly elastic, that is h E < h D. Second, in Figure 4 we show how the demand-side of membership might be binding in the determination of the equilibrium size of the regional trading bloc. In the case depicted in Figure 4, the number of countries willing to join the regional trading bloc is lower than the maximum size that the regional policymaker would have been willing to achieve if the demand of membership had been perfectly elastic, that is h D < h. Since the requests for membership in the bloc are not numerous enough, 20