ENDOGENOUS TRADE POLICIES IN A DEVELOPING ECONOMY

Transcription

1 ENDOGENOUS TRADE POLICIES IN A DEVELOPING ECONOMY BY NGUYEN MANH HUNG AND NGUYEN VAN QUYEN Ths aer examnes the formaton of trade olcy for a small oen develong economy where lobbyng actvtes may be carred out to nfluence olcy-makng decson. The aer resents a three-sector economy n whch the manufacturng sector can lobby olcy makers for favorable olces. Under some lausble condtons t s demonstrated that lobbyng actvtes carred out by the owners of the secfc factor n the manufacturng sector would secure a rotectonst trade olcy through ether an mort tarff or an exort subsdy. The government also moses a consumton tax on agrcultural roducts whch further strengthens the rotectonst measure aled to the manufacturng sector. In general equlbrum there wll be a movement of labor to the manufacturng sector an outut exanson n the manufacturng sector and an outut contracton n the agrcultural sector exactly as suggested by factual evdence. Unversté Laval Québec Québec Canada G1K 7P4 Tel.: E-mal: nmhung@ecn.ulaval.ca Unversté d Ottawa rue Wlbrod Ontaro Canada K1N 6N5 Tel.: ext Emal: nvquyen@uottawa.ca JEL Classfcaton: F1 Address of Contact Author: Nguyen Van Quyen Tel.: ext Fax: Emal: nvquyen@uottawa.ca

2 1. Introducton Accordng to the new oltcal economy trade rotecton s endogenous n the sense that t results from motves lyng outsde the realm of ure economc effcency. To acheve oltcal goals governments often carry out olces that dstort trade flows and affect the dstrbuton of ncome. These dstortng trade olces could be exlaned through the oltcal-suort functon of Hllman 198 Long and Vousden 1991; the electoral cometton aroach of Brock and Magee 1978; the factor-ownersh dstrbuton cum voter artcaton costs under majorty votng of Mayer 1984; or the more recent nfluence-buyng aroach of Grossman and Helman 1994 GH hereafter who aled the theoretcal machnery of a frst-rce menu aucton develoed by Bernhem and Whnston 1986 BW hereafter. Accordng to Watkns and von Braun hgh tarffs and trade barrers are used by ndustralzed natons to kee out morts and each year the rch countres send US$ bllon to suort ther agrculture. Ths leaves bllons of eole n the develong world wth no market to exort ther agrcultural roducts. Ironcally the resonse of oor natons to Western rotectonsm s also to adot rotectonst measures but n the manufacturng sector deemed to consst of all nfant ndustres. In ths aer we attemt to analyze the consequences of nfluence buyng n a small develong country that s oened for nternatonal trade. There are three roducton sectors: a sector roducng a freely traded good a sector roducng a traded manufacturng good and a sector roducng a non-traded good the agrcultural sector. Followng Samuelson 1971 and Jones 1971 we assume that the roducton technology n these sectors requres the use of labor and a secfc factor. The roducton

3 structure n our model s the same as that of Corden and Neary 198 CN hereafter who drew on Jones o ct. n formulatng a model for analyzng the roblem of boomng and declnng sectors n a small oen economy. Most of comaratve statc generalequlbrum results used n our aer can be found n CN. However n our aer we go further than CN and offer an exlanaton for the endogenous determnaton of olces drven by lobbyng actvtes. In our model owners of the secfc factor n the roducton of the traded good s hghly concentrated whch makes t easer for them to get organzed as a secal-nterest grou to lobby olcy makers for the adoton of favorable trade olces. In contrast small farmers n the non-traded good sector are wdely dsersed and thus are not able to organze themselves to nfluence olcy makers. Workers could not exert such nfluence ether because there s effectvely no nsttuton such as labor unons that mght lay a role n settng wages. Thus n exchange for ther labor workers receve a comettve wage determned by market condtons. We would lke to ont out at ths ont that the wage rate s endogenously determned n our analyss not fxed as n the framework of GH. Trade olces therefore have an mact on the suly sde through ntersectoral allocaton n contrast wth the analyss of GH who consdered only the reercussons on the demand sde. Under some lausble condtons we are able to show that lobbyng actvtes carred out by the owners of the secfc factor n the manufacturng sector secure a rotectonst trade olcy through ether an mort tarff or an exort subsdy for ths sector. Also the owners of the secfc factor n agrculture n our model wll have to bear a tax. Thus the owners of the secfc factor n the manufacturng sector also beneft ndrectly from the consumton tax mosed on the consumton of the non-traded good.

4 Labor moves to the manufacturng sector and at the new equlbrum there wll be an outut exanson n ths sector and an outut contracton n the agrcultural sector. The rent made by the owners of the secfc factor n the manufacturng sector wll rse after the new equlbrum has been establshed whle the rent obtaned by the owners of the secfc factor n the agrcultural sector may rse or fall deendng on the arameters of the model. The aer s organzed as follows. In Secton we resent our model a three-stage game n whch we allow for labor moblty across the three sectors of roducton. Wthn a general-equlbrum settng we solve our roblem of endogenous olcy formaton n Secton. Fnally n Secton 4 we dscuss several ossble extensons of the model.. The Model Consder a small oen economy n whch three goods labeled good 1 good and good are roduced from labor and catal. Goods 1 and are traded on world markets whle good s non-traded. For a develong country good may ether be a nonexortable local servce or an agrcultural roduct whch are heavly rotected by ther ndustralzed tradng artners. Good 1 s freely traded and s chosen as the numérare. The world rces of goods 1 and are denoted resectvely by and. The 1 domestc rces of good faced by ts roducers and consumers are denoted by and q resectvely. For good the government can mose an mort tarff or an exort subsdy. In ths case q. For good the non-traded good we shall assume that a consumton tax subsdy s mosed on ts consumers. In ths case q. A combnaton wll be referred to as a olcy. 4

5 On the roducton sde labor s assumed to be homogenous and erfectly moble. Its fxed suly s denoted by L. Catal s sector-secfc and the catal stock n sector s denoted by K 1. The outut of good s gven by Y F L K where L s the labor nut and F s a standard neoclasscal roducton functon that exhbts constant returns to scale. The rent accrung to the sector-secfc catal s then gven by Y L where s the revalng wage rate. Sector solves the followng rent maxmzaton roblem: max F L K # L for 1. 1 L # We shall let L denote the labor nut used by sector and L K Y F denote ths sector s outut. The oulaton of the small oen economy s assumed to be a contnuum of measure 1. Each ndvdual n the oulaton owns only one tye of nut. We refer to grou as the grou of ndvduals who own the secfc factor K. For each 1 let denote the sze of grou relatve to the total oulaton. The workers as a grou wll be referred to as grou 4. If L < 4 then L reresents the number of ndvduals who 4 are ether unemloyed or workng n the nformal sector. Although ths feature s common n a develong economy t s not the subject of our analyss. Thus we assume that L 4. Ths assumton mles that grous 1 and obtan ther ncomes through the ownersh of sector-secfc catal whle grou 4 obtans ts ncome by sellng ts labor. We exect that 4 s large relatve to 1. A grou such as the owners of a secfc nut who see ther ncome ted to the rce of the outut that ths secfc nut hels to roduce have an ncentve to lobby olcy 5

6 makers for olces that enhance ther ncome. In many develong economes the owners of the secfc nut used n nfant manufacturng ndustres often become organzed to lobby the government for favorable olces. In ths aer we concentrate on ths roblem wth grou as the focus of our attenton. Indvdual references are dentcal and reresented by the followng quas-lnear utlty functon: x x x x u x 1 1 where x s the consumton of good and u s the sub-utlty functon assocated wth the consumton of good. The sub-utlty functon u s assumed to be contnuously dfferentable strctly ncreasng and strctly concave. An ndvdual wth an ncome m at hs dsosal solves the followng utlty maxmzaton roblem: max x x q x u x m subject to m # q. x Assumng that ths roblem has an nteror soluton the followng frst-order condton characterzes the demand for good : u ' x q. We shall let x q denote the value of x that solves. The ndrect utlty functon of an ndvdual s then gven by where v q q m m s q q q [ u x q q x q ] s q 5 s the consumer surlus enjoyed by ths ndvdual. 6

7 The game of endogenous olcy formaton has three stages and ts extensve form s as follows. In the frst stage grou lobbes the government by communcatng to the olcy makers a contngent ayment schedule C C where C : reresents the ayment that t s wllng to make to the government f the olcy s mlemented. The ayment C also called the oltcal contrbuton n the lterature s exressed n terms of the numérare and s valued by olcy makers because t can be used to fnance future electons or smly ut away for ersonal use. In the second stage of the game the olcy makers take as gven the contngent ayment schedule C and mlement a olcy accordng to some crteron whch deends on the ayment as well as on some measure of socal welfare. In the thrd stage of the game the roducers take as gven the olcy mlemented by the olcy makers n the second stage and carry out ther roducton lans to maxmze roft. As for consumers they make ther consumton decsons whch deend on ther ncomes as well as the rces they face. Grou wll then make the ayment t romsed the government n the frst stage. Suose that s the olcy mlemented by the olcy makers n the second stage. The general equlbrum of the small oen economy s characterzed by the followng market-clearng condtons: L 1 L # L x Y. 7 Together 6 and 7 consttute a system of two equatons n the two unknowns and. We shall denote by and the equlbrum rce of the non-traded good and the equlbrum wage rate that are nduced by the olcy. 7

8 8 The tax revenue collected under the olcy s [ ]. # x Y x T 8 We shall assume that the tax revenue collected by the government s redstrbuted equally to all the ndvduals n the economy. Under the olcy the ayoff of grou 1 s gven by [ ] # $ s T W 9 The gross ayoff of grou before the oltcal contrbutons are made s gven by [ ] # $ s T W 1 and ts net ayoff by. C W As for grou and grou 4 ther ayoffs are gven resectvely by [ ] # $ s T W 11 and [ ]. 4 4 # s T W 1 The gross socal welfare before the oltcal contrbutons are made s then gven by # $ $ $ s T W 1 and the net socal welfare as a functon of the contngent oltcal contrbuton schedule C and the olcy s then gven by. C W The ayoff of the government s assumed to be gven by. 1 W C C W C C # # $ 14

9 In 14 reresents the weght assgned to the oltcal comonent of ayoff. Note that n the ayoff functon net socal welfare receves a weght equal to 1. In order for the government to accet oltcal contrbuton t s necessary that > 1. If 1 a transfer from any grou to the government wll decrease the latter layer s ayoff. Now let # C arg max 15. C As defned C s the set of olces that are best aganst C. Defnton: Let C C be a feasble contngent ayment schedule and : a olcy. We say that the lst C s a Nash equlbrum for the game of endogenous olcy formaton f the followng condtons are satsfed. Frst # # C. Second for any feasble contngent ayment schedule C of grou we have W C # su C [ W ]. % $ C. The Endogenous Determnaton of Polces Let be the equlbrum wage rate that revals n the small oen economy when s the olcy mlemented by olcy makers. Welfare Maxmzaton If olcy makers objectve s to maxmze socal welfare then the government solves the followng maxmzaton roblem: max W 16. µ The followng frst-order condtons characterze the soluton of 16: D W [ D Zˆ ] [ D Yˆ ]

10 D W [ D Zˆ ] [ D Yˆ ] 18 where we have let ˆ Z x # Y and Y ˆ Y denote resectvely the equlbrum level of the mort of good and the equlbrum outut of the non-traded good under the olcy. Now defne # [ ˆ ][ ˆ ] [ ˆ ][ ˆ D 1Z DY DZ D1Y ]. Next note that as rses the demand for good declnes but ts suly goes u. Hence D Zˆ. Also we have D Yˆ. Also note that when rses the outut 1 < 1 < of good goes u but the outut of good goes down. Furthermore a rse n alone does not affect the demand for good. Hence D Zˆ and D Yˆ. < < Wthout further restrctons we cannot determne the sgn of unambguously. If we let and denote resectvely the elastcty of Z ˆ and the elastcty of Y ˆ wth resect to then can be rewrtten as follows: & ˆ ˆ # [ D 1Z ][ DY ] $ 1 '. % Intutvely we exect that the own-rce effect domnates the cross-rce effect.e. / 1 / < < Thus f both nequaltes n 19 hold then. In ths case the only values of > and that satsfy the system consttuted by 17 and 18 are ; that s the olcy that maxmzes socal welfare s that of non-nterventon. 1

11 Determnaton of Endogenous Polcy If s a olcy that grou wshes the government to mlements then the shortfall n the socal welfare comonent of the government s ayoff s µ W. To nduce the government nto mlementng ths olcy grou must romse a ayment of at least # # / 1. µ W The net ayoff of grou after the ayment has been made s then equal to # # µ W # # / 1. W Hence the olcy that the owners of the secfc nut used n the roducton of good would lke olcy makers to mlement s the soluton of the followng maxmzaton roblem : µ W % max µ. & W ' 1 # $ The followng frst-order condtons characterze the soluton of D Zˆ ] [ D Yˆ ] ˆD W 1 [ D Zˆ ] [ D Yˆ ] ˆD W. [ Note that n 1 and we have used 17 and 18 as well as let ˆ 1 <. The olcy say that solves the system consttuted by 1 and satsfes the followng condtons: ˆ & # ˆ $ 1 ' D 1W DY % and ˆ & ˆ # D W D1Z 1. $ ' 4 % 11

12 In and 4 j are the elastctes already defned whle j 1 j s the elastcty of W wth resect to j. In the same j 1 manner that we exect the nequaltes n 19 to hold we should also exect that for the gross ayoff of grou the own-rce elastcty wll domnate the cross-rce elastcty.e. <. 5 Hence when 5 and both nequaltes n 19 hold the exressons nsde the square brackets n and 4 wll both be ostve. Next recall that when both nequaltes n j 19 hold. Also recall that D Y. Thus > ˆ < wll be ostve or negatve deendng on whether D W s ostve or negatve. 1 Now W the gross ayoff of grou has three comonents: rent government transfer and consumer surlus. The rent comonent ncreases when rses. A rse n causes the consumer surlus assocated wth the consumton of ths good to declne. Also a rse n wll nduce a rse n the consumer rce of the non-traded good causng the consumer surlus assocated wth the consumton of ths good to declne. As for the government transfer comonent f good s exorted a rse n means an ncrease n the subsdy gven to the exort of ths commodty resultng n a heaver tax burden on each ndvdual of the small oen economy. If good s morted a rse n rases the tarff revenues collected on the morts of ths commodty. Furthermore f the tax on the consumton of the non-traded good s ostve the declne n ts roducton nduced by a rse n wll lead to a lower level of tax revenues collected from ths sector. On the other hand f the consumton of the non-traded good s subsdzed then a rse n wll 1

13 reduce the subsdy gven to the consumton of ths good. The overall mact of a rse n on the ayoff of grou deends on the net effects of these conflctng movements. However when the ownersh of the secfc nut n sector s hghly concentrated.e. when s very small we can gnore the government transfer and the consumer surlus comonents n the gross ayoff for good. In ths case the gross ayoff of grou can be aroxmated by the rents accrung to the secfc nut n ths sector and we have D W. We have just shown that f the ownersh of the secfc nut n sector 1 > s hghly concentrated and f the own-rce effects domnate the cross-rce effects then the owners of the secfc nut n sector wll be able to lobby olcy makers ether for a tarff or an exort subsdy for good. As for n general ts sgn deends on the sgn of D W. A rse n the tax on the consumton of the non-traded good deresses the wage rate and causes the roducer rce of the non-traded good to fall. Both sectors 1 and exand at the exense of sector. For the owners of the secfc nut n sector they see the rent they obtan rse wth. The consumer surlus that they obtan from the consumton of good also rses. As n the analyss on the sgn of f the ownersh of the secfc nut n sector s hghly concentrated then we have D W whch leads to. We have > roved the followng rooston whch s the central result of the aer: Prooston: Suose that the nequaltes n 19 and the nequalty n 5 hold.e. the own-rce effects domnate the cross-rce effects. If the ownersh of the secfc nut n sector s hghly concentrated then through ts lobbyng actvtes grou wll manage to obtan suort for ts ndustry ether through an mort tarff or an exort > 1

14 subsdy. Furthermore ts lobbyng actvtes wll nduce the government to mose a tax on the consumton of the non-traded good. Also the lobbyng actvtes of grou result n an exanson of sector and a contracton of sector. As for sector 1 t s not clear whether t exands or contracts. Proof: The frst two statements of the rooston have been establshed. To establsh the last two let us magne that startng from the ntal state of non-nterventon olcy makers mlement frst the olcy then the olcy. Frst note that when rses from to sector exands but sectors 1 and contract. The equlbrum wage rate that results from ths art of the olcy move s hgher than the equlbrum wage rate under nonnterventon. Next note that the move from to causes sector to exand further; sector 1 to exand; and sector to contract further. Furthermore the equlbrum wage rate dros n the second art of the olcy move. The overall macts of the lobbyng actvtes of grou are a net exanson of sector and a net contracton of sector. As for the net mact on the sector that roduces the numérare t s not clear whether ths sector exands or contracts. The Soluton of the Endogenous Polcy Problem The tax olcy gven by and 4 maxmzes the net ayoff of grou whle resectng the artcatng constrant of the government. Grou extracts the entre surlus generated by the artcaton of the government. A contngent ayment schedule that allows grou to maxmze ts net ayoff s C # C max[ W ]. 6 : µ 14

15 BW labeled such a contngent ayment schedule a truthful strategy. In adotng the strategy reresented by 6 grou only ams for a net ayoff of µ. More recsely f ts gross ayoff s less than µ then t wll not make any oltcal contrbuton. On the other hand any ayoff n excess of µ wll be offered to the government as oltcal contrbutons. For the government a best resonse to C s. Hence the lst C s a truthful Nash equlbrum for the game of endogenous trade olcy formaton. 4. What s next after Lobbyng? If we allow the owners of the secfc factor n agrculture to organze then obvously they wll lobby aganst the adoton of the consumton tax on agrcultural roducts. We can exect a lower tarff for the manufacturng good a subsdy for the agrcultural good and maybe a tax on labor ncome. Now one can exect the government s ayoff to rse above the level that t obtans under the olcy of non-nterventon. Fnally f workers are also allowed to unonze and lobby olcy makers then the government now the common agency wll see comettve bds from all grous n the economy. Accordng to GH these bds wll neutralze each other and the government wll extract the entre surlus generated n the game. Each grou must offer oltcal contrbutons to the government so that the latter layer wll not mlement any olcy that s unfavorable to ts own nterest: the end result accordng to GH s that of non-nterventon References Bernhem B. D. and M. D. Whnston Menu Auctons Resource Allocaton and Economcs Influence Quarterly Journal of Economcs

16 Brock W. A. and S. P. Magee The Economcs of Secal Interest Poltcs: The Case of the Tarff Amercan Economc Revew Corden W. Max and J. Peter Neary Boomng Sector and De-Industralsaton n a Small Oen Economy Economc Journal Grossman G. H. and E. Helman Protecton for Sales Amercan Economc Revew Hllman A. L. Declnng Industres and Poltcal-Suort Protectonst Motves Amercan Economc Revew Jones Ronald W. A Three-Factor Model n Theory Trade and Hstory n Trade Balance of Payments and Growth: Paers n Internatonal Economcs n Honor of Charles P. Kndleberger ed. J. D. Bhagwat et al. Amsterdam: North Holland -1. Long N. V. and N. Vousden Protectonst Resonses and Declnng Industres Journal of Internatonal Economcs Mayer Wolfgang Endogenous Tarff Formaton Amercan Economc Revew Mtra D. Endogenous Lobby Formaton and Endogenous Protecton: A Long-Run Model of Trade Polcy Determnaton Amercan Economc Revew Samuelson Paul A. An Exact Hume-Rcardo-Marshall Model of Internatonal Trade Journal of Internatonal Economcs Watkns Kevn and Joachm von Braun Tme to Sto Dumng on the World s Poor Internatonal Food Polcy Research Insttute Rernt from - Annual Reort. 16