Seeking Rents in International Trade

Transcription

1 MSABR -6 Morrison School of Agribusiness and Resource Management Faculty Working Paper Series Seeking Rents in nternational Trade Andre Schmitz and Troy G. Schmitz April 9, This report is also available online at

2 SEEKNG RENTS N NTERNATONAL TRADE Andre Schmitz, University of Florida and Troy G. Schmitz, Arizona State University April 9, ABSTRACT We compare the current Canadian Supply Management regime in hich producers and importers benefit from rent-seeking activities that set production quota and import quota levels ith those under a tariff, in hich producers partakes in rent-seeking activities in order to induce the government to introduce a favorable tariff regime. We explore three different quota-setting games: () the import quota and production quota are set at a level that arises from a Cournot- Nash equilibrium beteen producers and importers; () the producer marketing board acts as a Stackelberg leader, taking into account the importers reaction to its production quota level; and (3) the importer behaves as a Stackelberg leader, taking into account producers reaction to its import quota level. We compare these quota-setting games ith to different tariff-setting games: () A non-cooperative game in hich the government sets the tariff at a level that maximizes tariff revenue; and () A cooperative game in hich producers, through rent-seeking activities, induce the government to set the tariff at a level that maximizes joint government and producer rents.

3 . NTRODUCTON There are many real-orld cases here producers are protected by import quotas. n addition, these producers are legally able to control production through supply management-type arrangements (i.e. quotas). They are essentially alloed to behave as profit maximizing monopolists subject to a voluntary export constraint placed upon importing firms [Harris, 985]. There are also cases in hich joint production and import controls exist. These controls come from both importers and producers and are enforced by their respective governments [Vercammen and Schmitz, 99]. Under these types of arrangements, producers earn rents above those attainable under competitive conditions. Correspondingly, private importers earn import rents hich ould not be available under a orld free-trade environment [Barichello, 98, Vercammen & Schmitz, 99, and Schmitz, Furtan, and Baylis, ]. The Canadian system of supply management for agricultural goods such as dairy and poultry represents one such case. There have been several proposals to eliminate both import and production quotas for supply-managed products in Canada. For Example, on August 3, 99, the Canadian nternational Trade Tribunal undertook an inquiry into the allocation of import quotas. One option as to replace import quotas under the Canadian supply management system ith tariffs, thereby diverting revenue from private importers to the government via tariffs. This so-called tariffication proposal has been described elsehere [Moschini and Meilke, 99; Schmitz and Schmitz, 994; Schmitz, de Gorter, and Schmitz, 996; Schmitz, Furtan, and Baylis, ]. n 995, as a part of GATT negotiations that led to the formulation of the WTO, import quotas for Canadian supply management ere replaced ith a system of tariff-rate quotas. Hoever, the over-quota tariffs for supply managed goods ere set at prohibitive levels, hile the ithinquota trade volumes (minimum access levels) ere set at essentially the same level as the import quotas that existed under the original supply management scheme. n addition, the original system of production quotas under supply management still remains in place.

4 Several studies have examined the issue of replacing quotas ith tariffs in hich the domestic firm has market poer. The ell-knon results on the "nonequivalence" of quotas and tariffs [eg. Bhagati, 965] state that tariffs and quotas are not equivalent in a orld hich is not perfectly competitive. Other studies have looked at the issue of tariffication in the presence of monopoly assuming different types of "non-economic objectives". For example, [Seeney et. al., 977] attempt to rank alternative tariff and quota policies by assuming that the government's objective is to attain a target ratio of imports to domestic production. McCulloch [973] examines the elfare cost associated ith tariff or quota protection alloing the cost to depend on the objective underlying protection. Another analysis [Vercammen and Schmitz, 99] concludes that if producers under supply-management are forced to choose beteen offering import concessions and abandoning supply-management, they ill, in specific circumstances, choose the former. While these papers have ansered many questions concerning the elimination of supply-management, none of them have explored the effects of replacing a quota ith a tariff under the assumption that producers and private importers noncooperatively determine domestic production and the level of imports. Vercammen and Schmitz [993] have modelled this endogeneity on the supply-managed side and analyzed the effects of a move to free trade, but the effects of government intervention have not yet been explored. The purpose of this paper is to compare a supply-managed quota system in hich producers and importers benefit from rent-seeking activities that set production quota and import quota levels ith those under a tariff here producers partakes in rent-seeking activities in order to induce the government to introduce a favorable tariff regime. Under the existing supply management regime, the level of production quotas and import quotas are jointly determined through the interaction of a marketing board that represents all domestic Canadian producers, and private importers hose import level and allocation of import quotas among producers is set by a single entity. We explore three different cases; () the import quota and production quota Bhagati defines tariffs and quotas as equivalent if the replacement of an explicit tariff by a quota set at the import level under the tariff ill produce an implicit tariff hich is identical to the explicit tariff. 3

5 are set at a level that arises from a Cournot-Nash equilibrium beteen producers and importers; () the producer marketing board acts as a Stackelberg leader, taking into account the importers reaction to its production quota level; and (3) the importer behaves as a Stackelberg leader, taking into account producers reaction to its import quota level. The above situation is compared to a scenario in hich the government eliminates import quotas and imposes a tariff in an attempt to collect tariff revenues. The possible outcomes under the imposition of such a tariff are many. The tariff revenue and producer rents that result from tariffication depend on the degree to hich the government and the producers cooperate. We present to situations that may arise under tariffication. The first situation assumes that the government sets the tariff at a level that maximizes tariff revenue. The second situation assumes that the government and producers cooperate in an attempt to maximize their joint revenue. We discuss the results by making comparisons across these alternative scenarios.. PRODUCERS AND MPORTERS RENT-SEEK (NTAL SUPPLY MANAGEMENT SCHEME) For the purposes of this analysis, e assume that producer and private import groups can successfully (and costlessly) lobby the government to set both the production quota and import quota at hatever level they jointly decide upon. n essence, the producer and importer determine the quota levels endogenously. We employ the small country assumption so that importers purchase the commodity from exporters at a constant orld price (incurring no transaction costs) and sell it to domestic consumers at a markup. The remainder of domestic demand is satisfied by domestic producers. Private importers collect revenue in the form of import rents hile producers receive economic rents that exceed those under competition, due to the poer of producer groups to lobby the government to set a statutory production quota, administered through the supply management marketing board. Figure shos a typical situation in hich producers and importers partake in rent-seeking activities. The consumers' demand curve is represented by D, the monopolist's marginal cost curve is given by MC, and P is the price at hich the importer can purchase the good from the 4

6 rest of the orld. Assuming (for the moment) that the quota level is fixed at and that domestic supply is fixed at s, the total quantity supplied to the domestic market ould be d s +. n equilibrium, the consumer price ould be set at P d hich is here the total quantity demanded s the total quantity supplied. Under this situation, importer rents are given by the markup (P d - P ) times the importer level and are represented by the cross-hatched region in Figure. Producer rents are determined by the area above the marginal cost curve and belo the domestic price, bound by the domestic quantity supplied s. These rents are represented by the shaded area in Figure. Notice that Figure is dran assuming aay the endogeneity of the quota level. This is just one arbitrary scenario in hich the level of domestic production happens to be set belo the level that hich ould exist under free trade. Hoever, if the quota is alloed to be determined endogenously then many different scenarios can exist. For example, consider the proposal to do aay ith supply management schemes altogether. Proponents of this ould argue that because supply managed industries provide unfair protection to domestic producers, eliminating quotas ould cause more imports to enter the country. Hoever, as [Vercammen and Schmitz, 993] point out, once the import quota is endogenized, situations can arise in hich the import quotas that exist under supply-management can actually be larger than the level of imports that ould exist under free trade.. MODEL SPECFCATON Let the country's inverse domestic demand schedule and producers marginal cost schedule be given by P d a b( + ) and MC α + β here,, Pd, and MC denote domestic production, imports, the price paid by consumers, and the marginal cost, respectfully, and a,b,α, and β are positive constants. These parameters can be expressed in terms of the orld price P, hat ould be the level of imports under orld competition (free trade), and the domestic quantity of production under free trade using the elasticity of supply ε (measured at P and 5

7 ) as ell as the absolute value of the elasticity of demand η (measured at P and + ) a ( + / η) P α ( / ε) P as: b (/ η)[ P /( + )] β (/ ε) P / Also, define to more parameters hich ill later be used for simplification purposes. Let r ε / η be the ratio of the elasticity of supply over the absolute value of the elasticity of demand under orld free trade competition. Define. For example, ν + ( + νr) and + ( + r). This implies that < iff x y since r is positive. Also, let the commodity in question. Finally, let θ + + x y < P d P M be the percentage mark up over the orld price here Pd is the consumer price of P be the producer's market share that ould be realized under orld free trade so that (-θ ) ould be the importers market share under free trade. Using the above notation the producer rents represented by the shaded region in Figure can be expressed mathematically as PR ( Pd α β / ) and the import rents (the crosshatched area in Figure ) are simply R ( Pd P ) MP. We can use the other parameters as ell to express these rents in terms of elasticities. For example, under orld free trade (here import rents are zero since Pd P) producer rents can be expressed as PR P / ε. These formulae, along ith general maximization principles are used extensively in the folloing sections to endogenously determine the quota and import levels as ell as rents under different types of interactions beteen producers and importers.. COURNOT-NASH DETERMNATON OF PRODUCTON AND MPORT UOTAS n the standard Cournot-Nash (CN) quantity setting game, e find the reaction function for the monopolist given imports () and the reaction function for the private importer given domestic quantity (). This process yields to equations in to unknons hich are solved simultaneously to determine the Nash equilibrium import quota and domestic production levels: The outcome ill be unique because the marginal cost and inverse demand curves are both continuous and monotonically increasing (decreasing) functions. 6

8 mporters: Producers: a P b { R } ( ) b a α b { PR } ( ) β + b Solving for and results in: cn ( + r), cn [ r ] With this information, the percentage markup over orld price can be expressed as: + r M cn. η + Producer and importer rents become: P PRcn ε cn, R cn P η cn Modelling the quota structure in a CN fashion has more intuitive appeal hen one player is not alloed to dominate another. Thus, this model makes more sense hen the producer's market share is near fifty percent (i.e. hen θ. 5 ). 3 n the folloing to sections e ill discuss the difference beteen this CN solution and the one in hich either player is alloed to dominate the other..3 THE PRODUCER GROUP S THE STACKELBERG LEADER n this section e assume that producers become the Stackelberg leader (ML) and importers follo. This model ould tend to apply to cases in hich producer market share is significantly larger than that of the importer (i.e. θ is closer to than ). 4 Under this scenario, 3 One could argue that the relative efficiency (in terms of marginal costs) ould be a more plausible indicator of hether the to firms compete in a Cournot-Nash fashion or in a Stackelberg manner. t may be the case that a firm that is relatively more efficient ould tend to dominate the less efficient firm. Hoever, in this model producers have linear marginal costs hile importers can be vieed as having constant marginal costs. These cost functions are not directly comparable because they are linear and cross only once. Thus, relative efficiency depends on the range of quantities under consideration. 4 This case in hich domestic producers have a relatively large market share ould apply, for example, in the Canadian supply-managed broiler industry in hich the domestic market comprises over 9 percent of total domestic consumption. 7

9 the producers maximize their economic rent given the reaction of the importer. That is, the producers knos () from above and maximizes accordingly: + + 4, ) ( )} ( { ml ml r r PR Using the endogenously determined quota level ml the percentage markup over orld price becomes: [ ] ml r M + 4 η Producer and importer rents become:, ml ml ml ml P R P PR η ε.4 THE MPORTER GROUP S THE STACKELBERG LEADER n this section e explore the case in hich importers are the Stackelberg leader (L) and the producers follo. This model ould tend to apply hen private importers have a larger market share than producers (i.e. θ is closer to than ). Under this scenario, the importer maximizes import rents ith respect to the level of imports, given the reaction function () from Section : [ ] [ ] l l r r R + + +, ) ( )} ( { Under this case, the percentage markup over the orld pricee received by importers becomes: [ ] l r M + η. Producer and importer rents become:, l ml l l P R P PR η ε.5 COMPARNG RENTS UNDER PRODUCER AND MPORTER ARRANGEMENTS 8

10 n this section e compare the three different supply management outcomes, in hich producers rent-seek in order to set production quotas and importers rent-seek in order to set import quotas. The three different outcomes under consideration are the Cournot-Nash (CN), Producer Stackelberg Leader (ML) and the mporter Stackelberg Leader (L). Comparisons among all the different possible combinations ill no be made in turn. A summary of these comparisons and comparisons made beteen these quantity-setting games and the tariff setting games of the next section are contained in Tables -5. First, consider the Cournot-Nash vs. the Producer Stackelberg Leader outcomes. The production quota under ML and CN can be compared by analyzing the ratio of the quantity under ML ith respect to the quantity under CN. Simplifying this ratio yields the folloing relationship: ML r + CN Since r (the absolute value of the ratio of the elasticity of supply over the elasticity of demand) is positive and and are positive, it must be the case that the production quota ill be larger hen producers lead hen compared to the Cournot-Nash outcome. A similar ratio can be used to determine the difference beteen import quotas set under these to arrangements. Division and extensive simplification yields: 5 ML K + + 3r K r CN Hence, import quotas under CN are alays higher than under ML. Domestic prices under each of these outcomes can be compared through a similar procedure, using the percentage markups over the orld price. After simplification, the ratio of the to domestic prices becomes: M M ML CN K + + 3/ r K + + r text. 5 K refers to a constant. This constant is different in each of the equations that it is presented throughout the 9

11 Hence, the domestic price under CN is higher than under ML. This also implies that the total quantity consumed under CN is alays loer than under ML, so that consumers prefer ML. What about producer and importer rents? As one ould expect, the ratio of production quota rents under the to outcomes becomes: PRML + + r PR + + 4r CN This expression is less than, hich implies that producer rents are alays higher under ML. Finally, the ratio of import rents reduces simply to: R R ML CN ML CN As to be expected, since imports are alays higher under CN, import quota rents are alays higher under Cournot-Nash than the situation in hich producers lead. To summarize: producers are alays off, importers are orse off, and the consumer is off under ML compared to CN. These results are summarized in Tables -5. Next, consider the Cournot-Nash (CN) vs. the mporter Stackelberg Leader (L) game. As expected, a similar analysis reveals just the opposite for this situation. Specifically, the production quota is alays higher and the import quota is alays loer under CN hen compared to L. Also, domestic prices are higher under CN hich implies that total consumption is loer under CN and consumers prefer L. Finally, producer rents are higher and import rents are loer under CN than under L. These results are also summarized in Tables -5. One can also make comparisons across the to Stackelberg games. Besides the comparisons that hold by transitivity through the CN case (i.e. that producer rents are larger and import rents are loer under ML than L), e can compare the domestic prices under ML and L. The ratio of the domestic price under ML ith respect to L simplifies to: M ML K M K L This relationship can be further simplified by multiplying the top and bottom of the price ratio by ( + ) and making use of the market share parameter θ that ould exist under free

12 trade. t turns out that if the producer has a larger free trade market share than the importer, the total quantity supplied to the domestic market ill be larger under ML than under L. The inverse is true if the importer has a larger market share, hile the total supply is the same hen θ.5. Thus, if θ >.5 the price is higher under L than under ML implying that consumer elfare is greater under ML. f θ >.5 then consumers are orse off under L. Again, these results are summarized in Tables -5. GOVERNMENT RENT-SEEKNG (A SYSTEM OF TARFFS) n Section e considered the case of importers and producers rent-seeking in order to non-cooperatively determine the domestic production quota and the import quota that results under the supply management regime. We derived the resulting price markups and producer and importer rents under various scenarios. n this section, e consider the case in hich the government eliminates the import quota scheme under supply management and replaces it ith a tariff. Under this scenario, importers do not receive rents from the import quota, because it no longer exists. nstead, the government captures import rents in the form of tariff revenue. Once the government enters the picture, the interaction among players (producers and the government) can no longer be modeled as a quantity-quantity setting game. Under this scenario the government sets the tariff, hich then determines the equilibrium level of domestic production and the equilibrium domestic price. We present the results for to plausible ays in hich the government might set the tariff: () The government chooses a tariff level that maximizes revenue; or () The government chooses a tariff level that maximizes joint revenue beteen the government and producers.. GOVERNMENT SETS THE TARFF THAT MAXMZES TS REVENUE The effect of eliminating supply management and replacing it ith a tariff that is set in order to maximize government revenue is depicted in Figure here MC is the marginal cost schedule for producers and D is total demand. P represents the orld price and is the

13 domestic quantity that ould be produced under free trade. P C represents the equilibrium price and quantity that ould exist under autarky in perfect domestic competition. Consider a tariff level t that results in a price P +t that is loer than P C in Figure. Within the range from P to Pc a reduction in the tariff causes a reduction in domestic production, yielding positive imports and thus positive tariff revenue for the government. Hence, if the government objective is to maximize tariff revenue, the government ill set the tariff at some value t so that the orld price plus the tariff is not higher than P C in Figure. This leads to tariff revenue of the cross-hatched area and producer rents to the shaded region. Retaining the notation from Section e can express this problem (GN) as { GR P + T α + β}. Where GR is government tariff revenue and α + β is producers T marginal cost. The solution to this optimization problem results in an ad-valorem tariff level of: T gn η nserting this solution into certain equations and simplifying yields: ( + r) gn +, gn These relationships imply that the domestic production quota gn ill alays be higher than or to the free-trade domestic quantity but less than c in Figure. n addition, the total quantity purchased by consumers (gn + gn) ill alays be loer under this noncooperative government rent-seeking game, hen compared to free trade. Producer rents and Government tariff revenue under this scenario are: P P PRgn gn, GRgn gn ε η The government ill import some quantity for hich it ill collect tariff revenue of the amount GRgn and producers ill receive rents of PRgn. The amount of this tariff revenue ill decrease in the demand elasticity parameter η.. GOVERNMENT SETS THE TARFF N ORDER TO MAXMZE JONT

14 PRODUCER AND GOVERNMENT REVENUE n this section e present an alternative in hich the government sets the tariff at a level that maximizes joint producer and government rents. There are actually to separate mathematical representations for this problem, both of hich must be explored. The choice of alternatives depends upon hether the tariff is set at or belo the prohibitive level that ould exist under autarky, or hether the tariff is set at a prohibitive level that is higher than hat ould exist under autarky. Each of these cases ill be discussed in turn. The first possibility, is that the tariff level gets set at or belo the prohibitive level. n this case, the joint maximization problem can be expressed as: T { JR PR + R} P + T PC The solution to the maximization problem results in an equilibrium ad-valorem tariff level: T. η This tariff level implies an equilibrium domestic production level to: ( + r) and an equilibrium level of imports of gc. Finally, producer rents and tariff revenues are given by: P PR, gc GR ε From the above analysis, a tariff level to T is prohibitive, and results in no imports ith a price and quantity solution that is exactly to the price (P C in figure ) and quantity that ould exist under autarky ith pure competition. Hence, regardless of the elasticities of supply or demand, T acts as a floor for the tariff that the government ill set if it is attempting to jointly maximize producer rents and tariff revenue. This implies that any cooperative joint maximization involving a tariff ill result in a prohibitive tariff ith no imports and no government revenue. n essence, the tariff that maximizes joint revenue is the same as the tariff 3

15 that solely maximizes producer rents. This leads to the outcome described belo, hich ill eventually turn out to be equivalent to the monopoly solution under autarky. n this case, the joint maximization problem (GC) can be expressed as: T { JR PR} P + T PC The solution to the maximization problem results in an equilibrium ad-valorem tariff level: + r T gc η Substituting and simplifying yields the folloing values for production and imports: ( + r) gc, gc Producer rents and tariff revenue under this case are to: P PR, gc gc gc ε This tariff is alays higher than the minimum prohibitive tariff, because: Tgc K + ( + r) r T K is alays greater than one. Hence, if producers can rent-seek in order to get the government to maximize joint revenue, it ould result in a tariff above the minimum prohibitive tariff (the tariff that ould induce autarky), hich ould be the same as the tariff level that maximizes producer rents and results in the monopoly level of production. This situation results in a price P M and quantity M in Figure hich occurs here marginal revenue (not shon) intersects marginal cost. Next, e compare the cooperative joint maximizing tariff game (GC) ith the noncooperative tariff game (GN). First, comparing quantities under GC ith GN yields: gc K + K + 3θ + rθ gn Hence, the quantity under the tariff game that maximizes producer rents (GC) ill be higher than the quantity under the tariff game that maximizes tariff revenue (GN) only if: 4

16 3θ r < θ mports and tariff revenue are alays loer under GC because imports are alays zero under the tariff game that maximizes producer rents, but are alays positive under the tariff revenue maximization game (GN). The domestic consumer price is alays higher under GC because the tariff is alays set above the prohibitive level, hile the tariff under GN is alays set belo the prohibitive level. Finally, producer rents are alays higher under GC hen compared to GN because it has already been shon that this solution results in the price and quantity combination that maximizes producer rents. V. COMPARNG UOTAS WTH TARFFS n this section e compare and contrast the results of the quota-setting games of section ith the tariff-setting games of section. V. Non-Cooperative Tariff (GN) vs. uantity-setting Games Here, e compare the non-cooperative tariff setting game (GN) of section ith the various quantity setting games of section. First, comparing quantities under the noncooperative tariff ith the Cournot-Nash outcome yields: gn ( + r) + + cn ( + r) + + This implies that the quantity produced under the non-cooperative tariff is larger than under the Cournot-Nash. Comparing the quantities under the producer-leader Stackelberg game to the non-cooperative tariff game yields: gn ( + r) + ml ( + r) + This implies that the quantity produced under the non-cooperative tariff is larger than under the producer-leader Stackelberg game. Through transitivity, the quantity produced under the noncooperative tariff must also be larger than under the importer-leader Stackelberg game. 5

17 The ratio of imports under GN hen compared to CN hen simplified becomes: gn K + 3 K + + cn Making use of the fact that and can be expressed in terms of the production and import shares that ould exist under free trade (θ and (-θ)), it turns out that imports under the government non-cooperative tariff setting game are greater (smaller) than imports under the Cournot-Nash quantity setting game if the import share that ould exist under free trade is greater than (less than) /3. The ratio of imports under GN hen compared to ML hen simplified is: gn K + K + + ml Further simplifying this relationship by substituting for θ reveals that if the import share that ould exist under free trade ere greater than ½, then imports under the government noncooperative tariff game are higher than imports under the producer-leader Stackelberg quantity setting game. Finally, comparing the ratio of imports under GN hen compared to L hen simplified yields: gn K + K + r il Hence, if r is less than (greater than) one, imports under the government noncooperative tariff ill be higher than (loer than) imports under the import-leader Stackelberg quantity setting game. n order to compare prices across the various cases, all e have to do is compare the advalorem tariffs under the tariff-setting game ith the markup rules under the quantity setting games. First, comparing prices under GN ith ML yields: Tgn M + r ml Since r is alays positive, the consumer price under the government non-cooperative tariff is alays smaller than the consumer price under the producer-leader Stackelberg game. This also means that the consumer price under the government non-cooperative tariff is alays 6

18 smaller than the consumer price under the Cournot-Nash quantity setting game through transitivity. Comparing prices under GN ith L yields: Tgn K + M K + r( + r) il Hence if r(+r) is less (greater) than one, the consumer price under the government non-cooperative tariff ould be greater (less) than the consumer price under the importer-leader Stackelberg game. n order to compare producer rents and import rents under the non-cooperative government tariff setting game ith the various quantity setting games, simulations ere performed over a ide range of values for θ and r. These simulations ere performed because the algebra involved ith attempting to simplify the relative amount of rents across each case as too complex. 6 t turns out that under a very ide range of values for θ and r, producer rents for the government tariff-revenue maximization game are alays loer than those for any of the three quantity setting games. n addition, import/tariff rents for the government tariff-revenue maximization game are also alays loer than those for any of the three quantity setting games. V. Joint imizing Tariff (GC) vs. uantity-setting Games Here, e compare the cooperative joint maximizing tariff-setting game (GC) of section ith the various quantity-setting games of section. We can make several generalized comparisons beteen the GC outcome and all three different quantity-setting games. First, imports and import rents are alays higher under all import and production quota-setting quantity games because they are alays zero under the joint maximizing tariff-setting game. Second, producer rents are alays higher under GC because GC is the solution that maximizes 6 For example, the comparison beteen producer rents under the Cournot-Nash game and producer rents under the game in hich the government sets the tariff in order to maximize revenue involves comparisons of polynomials that have combinations of fifth-order terms ith respect to r and fourth-order terms ith respect to θ. The simulation results are not provided in this paper due to space limitations. Hoever, these results are available from the author by request. 7

19 producer rents. Third, domestic prices are alays higher under GC than under any of the quantity-setting games, because the monopolist solution results in the largest reduction in the quantity produced, hich results in the highest price. Hence, consumers are alays orse of under the joint maximizing tariff-setting game. Again, these results are summarized in Tables - 5. All that remains is to compare production levels. Comparing the quantities under the joint maximizing tariff ith those under the Cournot-Nash quantity game yields: gc K + + r K cn So that the quantity produced under the joint maximizing tariff-setting game is alays larger than the quantity under the Cournot-Nash import and production quota setting game. Next, comparing quantities beteen GC and the producer-leading Stackelberg game yields: gc K + ( θ ) K + θ ML Hence, the joint maximizing tariff-setting level of production ill be higher than the production quota in the producer-leading Stackelberg game only if θ</. Finally, the joint maximizing level of production ill alays be higher than the production quota in the importer-leading Stackelberg game by transitivity, based on the fact that it is already higher than the Cournot- Nash quantity, hich in turn, is higher than the import-leading Stackelberg quantity. V. CONCLUSON We have derived and compared possible outcomes hen a government uses tariffs to replace endogenized quotas and thus engages in government rent-seeking activity in an attempt to collect tariff revenues. We begin ith one of three supply-management structures in hich a importers and domestic producers engage in different quantity setting scenarios and are alloed to endogenously determine the initial import quota and domestic production levels. Then e assume that the government enters the picture, eliminating quotas and interacting ith the producers in a non-cooperative and a cooperative tariff setting arrangement. We compare 8

20 consumer, producer, and importer elfare across these various scenarios. Several results emerge. f the government and the producers do not cooperate, the equilibrium tariff that the government sets ill not be prohibitive and consumers ill alays be off hen compared to any of the supply-management structure. On the other hand, if the government cooperates ith producers in setting the tariff, the tariff ill alays be prohibitive. Producers ill alays be off and importers and consumers ill alays be orse off under the cooperative tariff hen compared to any of the supply-management structures. Hence, if supply-management is replaced ith a tariff regime, there are major incentives for rent-seeking behavior on the part of producers in order to get the tariff level set at that hich maximizes joint government and producer rents. Producers ould be illing pay a considerable amount in rent-seeking activities in order to ensure the tariff is set at a prohibitive level. n essence, under the WTO, this is hat has happened. With a minimal amount of concessions in terms of alloing a small percentage of imports under TRs, producers ere able to lobby the government to set the over-quota tariff at a prohibitive level for all supply-managed products. 9

21 REFERENCES Barichello, R. R. (98). The Economics of Canadian Dairy ndustry Regulation. Economic Council of Canada Regulation Reference and the nstitute for Research on Public Policy. Technical Report E/, Ottoa. Bhagati, J.N. (965). On the Equivalence of Tariffs and uotas in R.E. Baldin et al., Trade Groth and the Balance of Payments. Rand McNally, Chicago The Canadian nternational Trade Tribunal, (99). Descriptive Staff Report: An nquiry into the Allocation of mport uotas. Research Branch, Reference No. GC-9-, Ottoa, Canada. Harris, R. (985). "Why Voluntary Export Restraints are 'Voluntary'", Canadian Journal of Economics 4, pp Kreps, D.M. (985). "Bertrand Equilibrium" in A Course in MicroEconomic Theory Princeton University Press, pp McCulloch, R. (973). "When are a Tariff and a uota Equivalent?". Canadian Journal of Economics 6, pp Moschini, G. and K. D. Meilke (99). Tariffication ith Supply Management: The Case of the U.S.-Canada Chicken Trade, Canadian Journal of Agricultural Economics 39, pp Schmitz, A. (983). Supply Management in Canadian Agriculture: An Assessment of the Economic Effects". Canadian Journal of Agricultural Economics 3, pp Schmitz, A. and T. Schmitz. (994). Supply Management: The Past and Future. Canadian Journal of Agricultural Economics 4(): Schmitz, A., H. Furtan, and K. Baylis. (). Agricultural Policy, Agribusiness, and Rent- Seeking Behaviour. Toronto: University of Toronto Press. Schmitz, A., H. de Gorter, T. Schmitz. (996). Consequences of Tariffication. in Regulation and Protectionism Under GATT: CASE Studies in North American Agriculture. A. Schmitz, G. Coffin, and K.A. Rosaasen, eds. Westvie Press: Boulder, CO. Seeney, R., Toer, E., and Willet, T. "The Ranking of Alternative Tariff and uota Policies in the Presence of Domestic Monopoly" Journal of nternational Economics 7 pp Vercammen, J. and A. Schmitz (99). Supply Management and mport Concessions, Canadian Journal of Economics 4, pp Vercammen, J. and A. Schmitz (993). "Deregulating Supply-Managed ndustries: The Unexpected Trade Effects". UBC Working Paper (Feb. 7, 993).

22 Table : Comparison of Domestic Production Levels under Various Trade Games Cournot-Nash Producers Lead mporters Lead Tariff Revenue Joint Revenue Production and mport uota-setting Games Cournot Producers mporters Nash Lead Lead smaller larger larger Tariff Revenue smaller smaller smaller Tariff-Setting Games Joint Revenue smaller larger if θ>/ smaller larger if rθ >-3θ Comparisons are made by taking the games don the left-hand side and then comparing them to the games along the top ro. θ is the production share of total consumption that ould exist under free trade. r is the absolute value of the ratio of the supply elasticity to the demand elasticity.

23 Table : Comparison of mport Levels under Various Trade Games Cournot-Nash Producers Lead mporters Lead Tariff Revenue Joint Revenue Production and mport uota-setting Games Cournot Producers mporters Nash Lead Lead larger smaller smaller Tariff Revenue larger if θ>/3 larger if θ>/ larger if r > Tariff-Setting Games Joint Revenue larger larger larger larger Comparisons are made by taking the games don the left-hand side and then comparing them to the games along the top ro. θ is the production share of total consumption that ould exist under free trade. r is the absolute value of the ratio of the supply elasticity to the demand elasticity. is the domestic quantity that ould be produced under free trade.

24 Table 3: Comparison of Consumer Surplus under the Various Trade Games Cournot-Nash Producers Lead mporters Lead Tariff Revenue Joint Revenue Production and mport uota-setting Games Cournot Producers mporters Nash Lead Lead orse orse if θ>/ Tariff Revenue orse orse if r(+r) < Tariff-Setting Games Joint Revenue Comparisons are made by taking the games don the left-hand side and then comparing them to the games along the top ro. θ is the production share of total consumption that ould exist under free trade. r is the absolute value of the ratio of the supply elasticity to the demand elasticity. is the domestic quantity that ould be produced under free trade. 3

25 Table 4: Comparison of Producer Rents under the Various Trade Games Cournot-Nash Producers Lead mporters Lead Tariff Revenue Joint Revenue Production and mport uota-setting Games Cournot Producers mporters Nash Lead Lead orse Tariff Revenue Tariff-Setting Games Joint Revenue orse orse orse orse Comparisons are made by taking the games don the left-hand side and then comparing them to the games along the top ro. 4

26 Table 5: Comparison of mport/tariff Revenue under the Various Trade Games Cournot-Nash Producers Lead mporters Lead Tariff Revenue Joint Revenue Production and mport uota-setting Games Cournot Producers mporters Nash Lead Lead orse orse Tariff Revenue Tariff-Setting Games Joint Revenue Comparisons are made by taking the games don the left-hand side and then comparing them to the games along the top ro. 5

27 Figure : Producers and mporters Participate in a Production/mport uota-setting Game P MC Pd P s < > d D 6

28 Figure : The Government Replaces mport uotas ith Tariffs P Pm MC P+T Pc P+t P <- t -> D m T t + 7