Exploring trade flows between MERCOSUR and the EU: what does an export demand function tell us? *

Transcription

1 Exploring rade flows beween MERCOSUR and he EU: wha does an expor demand funcion ell us? * Ariel A. Barraud ** Alejandro D. Jacobo * This paper is parially drawn from he auhors projec on MERCOSUR-UE rade relaions. ** Universidad Nacional de Córdoba. Universidad Nacional de Córdoba and Universidad Caólica Argenina. 1

2 Resumen Ese rabajo examina la relación comercial enre el MERCOSUR y la Unión Europea a ravés de una función de exporación para el período Mienras las exporaciones a nivel agregado reaccionan moderadamene a los precios relaivos, su reacción con respeco al precio de los compeidores es imporane. Además, los ingresos de la UE afecan negaivamene al comercio y el efeco absorción parece esar ausene. El precio y las elasicidades ingreso sugieren que la UE no es un desino aracivo para las exporaciones del MERCOSUR. La esimación del modelo de corrección de error predice algunas dificulades para un ajuse hacia el equilibrio a largo plazo. Absrac This paper examines he commercial relaionship beween MERCOSUR and he EU hrough an expor demand funcion for he period While a an aggregae level expors reac moderaely o relaive prices, heir reacion wih respec o rade compeiors prices is imporan. Besides, rade parner s income adversely affecs rade and he absorpion effec seems o be absen. The average long-run price and income elasiciies sugges ha he EU is no an aracive desinaion for MERCOSUR expors. The esimae of he error correcion model appears o predic some difficulies for a quick adjusmen of variables o long run equilibrium. JEL Codes: F15 Key words: MERCOSUR, UE, inegración económica, exporaciones, EU, economic inegraion, expors 2

3 1.- Inroducion The recen proliferaion of regional inegraion agreemens and he emergence of regional rade blocs ha are reshaping he world economy have moivaed a large amoun of research in he area of inernaional rade. Mos of he exising conribuions sudy he effecs of economic inegraion on he volume of rade among member counries as well as he welfare effecs of rade agreemens on non-members. In spie of he increasing ineres abou he effecs of regionalism, however, empirical research on he deerminans of rade relaionship beween differen areas of economic inegraion is raher scarce. 1 The analysis of he deerminans of rade flows beween economic areas or blocs would reveal valuable informaion for relevan macroeconomic policy issues. In fac, when implemening a commercial or an exchange rae policy a an aggregae level, one major concern of policymakers is he response of expors o relaive price changes. This examinaion is paricularly relevan for blocs ha have signed preferenial agreemens since changes ha migh occur when one of he blocs modifies is policies will cerainly have a subsanial effec on commercial ransacions. Moreover, he poenial subsiuabiliy ha may exis beween one of he parners and a hird exporer (anoher bloc) may aler he policymakers abou he presence of a prospecive rival. The expors reacion o own income or rade parner s income is also significan for economic policy analysis. Given he lack of recen sudies and, consequenly, he scarciy of new informaion for policy prescripions, his paper explores he deerminans of recen rade flows beween he Souhern Common Marke (MERCOSUR) and he European Union (EU) hrough an expor funcion. The esimaion of an expor funcion beween MERCOSUR and he EU is based on he ineres in exploring he characerisics of he 1 To he auhors knowledge, only a few aemps have been made o examine bilaeral rade jus beween he EU and MERCOSUR. See for insance Cuadros e al. (1999), Canavella Jordá e al. (2003). 3

4 aggregae relaions beween wo areas of economic inegraion ha have shown a significan commercial relaion over he pas years. 2 The paper seeks o deermine he long-run elasiciies of MERCOSUR expors o he EU for he pas years and ries o find he evidence of a sable long-erm balance beween MERCOSUR expors o he EU and differen variables of relaive prices and income hrough an expor demand funcion. To his end, he paper proceeds as follows. Secion 2 makes some commens abou he empirical sudies and inroduces mehodological issues relaed o an expor demand funcion. Secion 3 presens he esimaion and resuls. Secion 4 concludes. 2.- Theoreical and mehodological background 3 The lieraure consisenly confirms he significance of expors for he economic prosperiy of a counry or region, and akes ino accoun he influence of prices and income on expors. However, depending on he counry, he period under sudy and he model used, he imporance and significance of hese variables on expors may be differen (Gonzalez e al., 2004: 231). In fac, subsanial empirical lieraure exiss on he esimaion of price and income elasiciies in inernaional rade, bu he values of price elasiciies vary considerably. For illusraive purposes only, Marquez and McNeilly (1988) find ha boh income and price elasiciies play a significan role in he deerminaion of rade flows for developing and developed counries, whereas Bahmani-Oskooee and Niroomand (1998) find ha his resul holds only for less developed counries. However, Rose (1990,1991), Osry and Rose (1992) and Bahmani-Oskooee and Alse (1994) find ha a real devaluaion has generally no significan impac on he rade balance, bu Reinhar (1995) provides, insead, srong suppor o he view ha depreciaions improve he rade balance. Likewise, Senhadji and Monenegro (1999) show ha expors reac o boh relaive prices and he rade parner s income in a large sample of boh developing and indusrial counries, while Hooper e al. (2000) find ha he price channel is weak 2 The invesmen relaion has also been an imporan one. 3 For furher deails see Barraud and Jacobo (2009). 4

5 wih respec o coninenal European counries and ha he income variable becomes more imporan when explaining he rade balance of his naions. 4 Despie some mehodological aspecs and oher minor differences ha may exis, he sudies confirm he influence of prices and income on UE expors o MERCOSUR (Canavella Jordá e al. 2003; Cuadros e al. (1999); Balaguer Franch and Marínez Zarzoso, 2000). 5 More specifically, wih respec o MERCOSUR-o-EU expors, here is no sudy on his direcion of rade flows, bu insead here exiss a preliminary esimae of an expor demand funcion by Barraud and Jacobo (2008). 6 In spie of he fac ha, as menioned above, he resuls obained for he differen variables are diverse, in he mehodological aspecs mos of he sudies ha esimae expor demand funcions are based on he elasiciies approach, adoping a parial equilibrium framework where he deerminans of expors are exogenous (González e al., op. ci.: 235). In fac, from a mehodological poin of view, he heoreical foundaion of he empirical analysis is he Imperfec Subsiues Model (Goldsein and Kahn, 1985). The basic assumpion of he model is ha neiher impors nor expors are perfec subsiues for domesic producs. Such a hypohesis is confirmed by empirical evidence: if domesic and foreign goods were perfec subsiues, a given counry would be eiher an exporer or an imporer. Since he world marke is characerized by he presence of bilaeral rade and he coexisence of impors and domesic producion, he hypohesis of perfec subsiuion can be easily rejeced. The main feaures of he imperfec subsiues model can be summed up as follows. Along wih he sandard demand heory, i is supposed ha he represenaive 4 See Barraud and Jacobo (2009). 5 According o Canavella e al. (2003), any variaions in relaive prices ha migh occur in he rade or exchange rae policies of eiher MERCOSUR or he EU will probably have a subsanial effec on commercial ransacions, bu, in conras, for he EU-NAFTA, price seems o have a lower influence and probably oher aspecs such as produc differeniaion may play a more relevan role on heir flows of rade. Besides, for hese auhors, EU expors appear o be sensiive o changes in income in boh MERCOSUR and NAFTA counries, alhough he esimae for NAFTA is larger as his is a greaer marke in boh qualiaive and quaniaive erms; hus his variable becomes and engine of expor growh for EU expors owards NAFTA (Canavella e al., 2003: 15). 6 The sudies do no use recen saisics. As consequence, valuable informaion for economic policy recommendaions is omied. See Barraud and Jacobo (2008). 5

6 agen maximizes his uiliy subjec o a budge consrain. The resuling demand funcions for expors herefore describe he demand (X) as a funcion of he prices of he exporer relaive o he impored produc s own price (PR) and of he level of moneary income in he imporing counry (Y*). Assuming homogeneiy of zero degree, a log linear relaionship beween variables, and he presence of omied facors ( ), he coefficiens imply elasiciies and he previously inroduced funcion is hen formulaed as follows: LNX α 0 α LNPR 1 α 2 LNY * μ However, as in many oher cases, MERCOSUR expors face compeiion no only from domesic producers in he imporing region bu also from hird area exporers. Therefore, oher prices should be inroduced in order o capure he subsiuabiliy beween MERCOSUR expors and he producion from compeing areas. This equaion can be expanded in order o capure boh he effec caused by he movemen of prices of counries or blocs ha compee wih exporable producs of he exporing region and he impac of domesic demand on naional expors. 7 More specifically, he srucure of he demand funcion for expors is: LNX α 0 α LNPRMEU 1 α 2 LNPRMUSA α LNPRMLA α 3 LNYEU α 4 LNYM μ, where, X represens he expors from MERCOSUR o he EU, PRMEU he relaive prices beween MERCOSUR and he EU, PRMUSA he relaive prices of MERCOSUR and he Unied Saes, PRMLA he relaive prices of MERCOSUR and Lain American counries, YEU he EU income, YM he MERCOSUR income, and he error erm. 8 The coefficien of (PRMEU) is expeced o be negaive, since an increase in he price of expors greaer han ha of he prices of producs made in he imporing 7 This funcion is based in a parial equilibrium model ha combines he elasiicies approach of Marshall, Lerner and Robinson, wih he absorpion effec of Alexander (1952), accordingly wih Cuadro Ramos e al. (1999) p. 49, and where expors can no be considered as perfec subsiues of naional goods. 8 From he above-menioned premises, expors from MERCOSUR are he endogenous variables, while he relaive prices and income are he exogenous variables. 6

7 counry markes resuls in a relaive increase in he firs and discourages is demands if domesic subsiues do no exiss in he imporer counry. The value of his parameer could be higher or lower han uniy depending on he degree of subsiuabiliy. 9 In order o capure he influence of prices of expored goods by MERCOSUR wih respec o he value of subsiue goods, we use he variable PRMUSA and PRMLA. The sign of he price elasiciy in boh cases is expeced o be negaive, because he exporable goods of he compeing counry (or counries in he case of Lain America) are very comparable and hey have a similar desinaion. To he exen ha goods are more subsiuable, a greaer elasiciy would be expeced. The variable income (YEU) capures he budge consrain of expors desinaion. An increase of he income level leads o a raise on expors if normal goods are raded, and he income elasiciy is posiive. However, due o he evenual consumpion of eiher inferior or luxury goods when income increases a he desinaion bloc, a differen resul may be observed hus indicaing ha MERCOSUR expors do no behave as he EU income. 10 The las variable used is MERCOSUR s income (YM). This variable is included in order o capure an evenual absorpion effec, which would adversely influence expors. In fac, economic developmen may resul in an increase in consumpion and hence in domesic demand. Besides, i migh be possible ha his does no happen, because a greaer income in he bloc means a greaer aciviy ha may in urn increase impors bu his does no necessarily operae in derimen of expors. The daa used are annual, covering he period, and are limied by heir availabiliy as well as by heir homogeneiy. In fac, in MERCOSUR counries, periods of high inflaion and macroeconomic insabiliy have hampered he availabiliy of series for exended periods of ime and have mined he consisency required for any empirical analysis. 9 A high posiive value would be indicaive of a shorage of subsiues. 10 This would imply a remarkably high absorpion when esimaing a demand funcion for expors o he EU case by making i dependen on he income of he EU. The resuls of some sudies seem o confirm his assumpion (see for example Cuadro Ramos e al.1999). 7

8 Thus, i is difficul o obain homogeneous daa ses. I is a common feaure o find differen base years hrough ime. Once he base year changes, he new series is no exended backwards a number of years, while he old one is officially disconinued. Due o his usual pracice, one may no discover if he changes in some values are due o real changes in he series or due o changes in he saisical procedures. Because of his feaure, daa collecion and sysemaizaion for Lain America s counries remembers he Sisyphean boulder: an inexorable ques (Mena, 1995). In oher words, his suggess ha someimes, when he series are chained, one can no disinguish wheher he observed differences in growh in gross domesic produc are due o changes in he srucure of he economy (inpu-oupu marix) or if hey merely reflec he peculiariies of saisical processes. In any case, he chaining of he ime series is adoped in his work, especially in he GDP (income) series from he Lain American economies. 11 The series consruced were checked wih he informaion provided by Economic Sudies of Lain America annual surveys, which include a brief yearly descripion of he performance of he economies of he region, which helps o avoid inroducing any inconsisencies in he daa and disorions in he series. The general procedure was o use he laes available daa and o consruc he series from he presen o he pas, wih he assumpion ha recen daa are properly elaboraed. Wih regards o he aggregaion of variables, fifeen counries were included in he case of he UE, while he four counries ha originally signed he Treay of Asunción were considered in he case of MERCOSUR. 12 Counries included in he empirical analysis were: Ausria, Argenina, Belgium-Luxembourg, Brazil, Denmark, Spain, Finland, France, Greece, Holland, Ireland, Ialy, Paraguay, Porugal, Unied Kingdom, Sweden and Uruguay. 11 The economeric esing in Lain American counries requires a deailed and specific knowledge of heir economic policy during he period of research. Such informaion needs o be incorporaed in he specificaion as well as in he esimaion procedures. These pressing resricions sugges he adopion of he second bes mehodology relaed o he chain of he series. 12 Since he inegraion of he Associae Saes in he case of MERCOSUR is variable in naure he paper has used he foundaional Saes o carry ou he analysis. 8

9 The variables used in his sudy for he esimaion of he equaion and he consrucion of deailed indexes are defined as follows: Xi: real expors of each of he member counries of MERCOSUR; X: real oal expors of member counries of MERCOSUR; IPI: index of domesic prices in UE counries; IVU: he uni value index of expors for each of he MERCOSUR counries; PIBi: gross domesic produc of each of he counries of he EU; PIB: oal GDP of he UE; PM: weighed price index of MERCOSUR; PEU: he weighed price index calculaed wih he UE for domesic prices; PRMEU: relaive MERCOSUR o EU prices; PRMUSA: relaive MERCOSUR o US prices; PRMLA: relaive MERCOSUR o Lain America prices 13 ; YM: income of he MERCOSUR counries (approximaed by he oal GDP of he four members of he common marke); YUE: income of EU counries (approximaed by he oal GDP of he fifeen counries considered); The MERCOSUR prices wih respec o he UE (PRMEU) are consruced from he following expression: PRMEU PM PEU 4 i 1 15 i 1 Xi IVUi X PIBi IPi PIB The relaive prices of MERCOSUR o hose of he Unied Saes (PRMUSA) and Lain America (PRMLA) have PM as numeraor and he IUV of Unied Saes and Lain-Americans expors in he denominaor Lain American prices are from Souh-American counries, bu hey exclude hose of MERCOSUR members. 9

10 The daa series are in consan U.S. dollars. Expor flows are from he World Bank daabase of unresriced access ( specifically from he secion Daa and Saisics, suiably adaped for his sudy. Nicia and Olarreaga have updaed he daabase, exending he original period and improving i in various ways (Nicia and Olarreaga, 2006). 15 The prices for MERCOSUR and EU15, and he income variable, are from World Economic Oulook. 3.- Esimaion and resuls In order o formally es for he presence of uni roos in he expor demand funcion, he Augmened Dickey-Fuller es (ADF) was implemened for each variable (Dickey and Fuller, 1979). The resuls are repored in Table 1. Table 1 Uni Roo Tes Variables Augmened Dickey-Fuller Tes levels differences LX LPRMEU LPRMUSA LPRMLA LYEU LYM Criical values: (1%), (5%) The null hypohesis of he exisence of uni roos can no be rejeced for any of he variables in levels, excep for he series of relaive prices beween MERCOSUR and he U.S. The same scenario bu for he variables in differences is rejeced in all 14 Barraud and Jacobo (2009) have ried oher price variables from differen counries whose producs may compee wih hose of MERCOSUR, as in he case of Ausralia, Canada or China. However, hey were no a all significan. 15 See Nicia and Olarreaga (2006) for addiional deails. 10

11 cases a 5% level of significance. 16 One concludes, herefore, ha variables in levels are no saionary and conain a uni roo. Coinegraion provides an appropriae mehod for esimaing long-run and shorrun elasiciies. The resuls of he coinegraion analysis are shown in Table 2. The saisic from he maximum likelihood es (Max) es he null hypohesis of he exisence of r coinegraing vecors agains he alernaive hypohesis of he exisence of r+1 coinegraing vecors. Table 2 Johansen's Maximum Likelihood Coinegraion Tes Max and Trace Saisics Max Criical Value(*) Number of Coinegraion Vecors r = r Noes: r is he number of coniegraing vecors (null hypohesi (*) Oserwald-Lenum criical values a 5 percen level. Following he procedure commonly used (see, for example, Cuadros Ramos e al. (1999)), he nonsaionary variables require he use of coinegraion for he regression analysis o avoid problems arising from heir non-saionariy when working wih variables in levels. The equaion o esimae in levels collecs he long-erm relaionship beween MERCOSUR expors o he EU, MERCOSUR relaive prices wih respec o he EU, MERCOSUR relaive prices wih respec o he Unied Saes, he income of MERCOSUR and he EU income. If hese variables are inegraed, any linear combinaion of hem will have a lower inegraion order. Johansen s mehodology provides he righ ools o work wih nonsaionary variables. 16 In some cases 1%. 11

12 The es saisic derived from maximum likelihood conrass he exisence of r coinegraion vecors agains he alernaive hypohesis ha here are exacly r+1 coinegraing relaions. From he resuling inegraion vecors, i is advisable o use he one ha had any economic sense according o he expeced sign and magniude (Cuadro Ramos e al. op. ci.; Muscaelli and Hurn, 1992), and is reproduced in Table 3. Table 3 repors he coefficiens for long-run relaionships ha have been esed in order o examine wheher hey are significanly differen from zero or no. For he MERCOSUR-EU equaion, Chi-squared saisics are greaer han 95 per cen criical value, hus rejecing he null hypohesis ha he coefficiens are zero. Table 3 Coinegraion Vecor Variable LX LPRMEU LPRMUSA LPRMLA LYEU LYM Coefficiens z P>ΙzΙ The resuls of he esimae sugges he following demand equaion for MERCOSUR expors o he EU: LX 0.16LPRMUE 4.61LPRMUSA 3.41LNMLA 4.85LYEU 4.67LBYM μ where he coefficiens of he variables show he relevan elasiciies, since he funcion has been esimaed wih variables in logarihms. The relaive price elasiciy of MERCOSUR o he EU does no have he expeced sign. An increase in he price of expors of goods relaive o he price of he producs produced in he imporing bloc does no discourage is demand. I should be noed, however, ha here is no much sensiiviy. 12

13 In order o capure how prices of goods expored by MERCOSUR reac o price movemens in oher markes, he sudies uses a second group of relaive prices variables whose sign is he expeced. They sugges, for example, ha any changes in prices relaive o he U.S. prices can resul in a significan negaive effec on MERCOSUR expors o he EU. 17 The same conclusion arises in he case of Lain America. Wih regards o he EU income variable, he share of luxury goods in he European consumer expendiures increases as income increases and MERCOSUR does no expor such goods; no a leas for Europeans consumers. In oher words, income growh in he EU is no ransformed ino an increase in MERCOSUR expors o ha desinaion, possibly because he EU spends is income on goods ha are no hose expored by MERCOSUR: goods produced in he EU or elsewhere. Finally, wih regards o he income of MERCOSUR, he coefficien does no seem o pick up an evenual absorpion effec; an oucome ha ofen goes ogeher wih economic developmen and a lower amoun of available expor producs. Possibly his is due o he naure of he region s expors producs (agriculural goods) and ha a greaer income in he bloc implies an increase in impors bu i does no operae in derimen of expors. Moreover, since he MERCOSUR s GDP variable is no ne of expors, a posiive sign in he resuls for his variable was expeced o he exen ha expors conribue o oal GDP by definiion. Once he long-run esimaes were obained, i is ineresing o have deailed informaion on wha happens in he shor run. The long-run relaionship is hen complemened wih a consrucion of a VAR in form of an error-correcion. 18 For his purpose, we incorporae no only he variables ha conain he dynamic informaion in he shor run, bu also he coinegraion relaion in he long-run previously esimaed. The srucure of his model is he following one: 17 This resul mus be inerpreed wih care, since i is obained a an aggregae level. 18 The differen rade variables ha have been analyzed so far are no always in equilibrium. They flucuae before aaining long-run equilibrium. The long-erm relaionship is herefore complemened by he consrucion of an error-correcion model. 13

14 LX LX -1 γ LPRMEU δ LPRMUSA η PRMLA π YEU θ YM RES -1 where he expresses he firs differences of he variables and RES is he error correcion erm derived from he long-run equaion. In order o obain a model ha explains beer he dynamics, quaniaive variables (dummies) were also inroduced wihou a significan modificaion of he resuls. 19 (I should be noiced ha all he variables incorporaed in he error-correcion model are saionary and herefore OLS can be applied). The closer o uniy he coefficien is, he greaer he adjusmen speed in he exising disequilibrium beween MERCOSUR expors and he res of he variables will be. This is he way hese variables achieve he long-run equilibrium. The general adequacy of his model is given by specific diagnosic es on residuals such as serial correlaion, funcional form, normaliy and heeroskedasiciy. No srucural changes were found during he sample period according o he differen sabiliy ess. Table 4 shows he error-correcion model we consider wih is resuls. Table 4 Error Correcion Model Variables Coefficiens (-saisics) LX (0.67) LPRMEU (-0.40) LPRMUSA 1.04 (2.17) LPRMLA (-0.77) LYEU 4.89 (2.67) LYM (-0.73) RES (-1) (-0.69) 19 See Barraud and Jacobo (2009). 14

15 Boh sandard saisics and diagnosic ess indicae ha he consruced model can provide a valid explanaion of he shor-run dynamics of MERCOSUR expors o he EU. The error correcion erm (RES(-1)) shows a reasonable adjusmen in his case and 35% of he long run adjusmen occurs in each period. The significance of his variable reflecs and adjusmen in he EU expors owards a long-run equilibrium of 35% per cen approximaely per year. This resul demonsraes, a he same ime, he exisence of he coinegraing vecor, which has been incorporaed in he error correcion model. All he variables finally included in he error-correcion model are joinly significan. However, in he shor run equaion only prices relaive o he U.S. and income in he recipien bloc are saisically significan. The srucure of he MERCOSUR-UE error correcion model incorporaes pas informaion of expors which helps o explain par of he dynamics on rade flows owards EU. 4.- Concluding remarks This paper analyses he aggregae commercial relaionships beween he MERCOSUR and he EU using an expor demand funcion for he period which considers he effecs of relaive prices and income variables on expors from MERCOSUR o he EU. The long-erm analysis suggess ha he price elasiciy of MERCOSUR-EU relaionship does no have he expeced sign, bu is of a moderae magniude, and ha he price elasiciy beween MERCOSUR and he U.S. and Lain-American prices have he expeced sign and a value ha suggess a significan subsiuion effec. Wih regards o he EU income variable, is increase does no raise MERCOSUR expors possibly because he EU spends his addiional income on goods 15

16 oher han hose expored by MERCOSUR, whereas he coefficien of he income of MERCOSUR does no reveal an absorpion effec. The resuls show ha, alhough he blocs mainain ineresing rade exchanges, he EU is no an appealing desinaion for MERCOSUR expors. This circumsance is unaccepable for blocs seeking o reinforce heir rade relaions, as i is also he significan effec on expors ha a change in he compeior s price has. This is paricularly discouraging in an environmen where imporan devaluaions ook place in he pas. To sum up, while expors reac moderaely o own relaive prices, heir reacion wih respec o rade compeior s prices is imporan. Besides, rade parner s income adversely affecs rade and he absorpion effec, if any, seems o be absen. The average long-run price and income elasiciies sugges ha he EU is no an aracive desinaion for MERCOSUR expors, a leas no ye. Accordingly, he economic policy recommendaions ha can be drawn from his paper probably should consider some of hese issues o ensure ha business relaionships hrive, and definiely srenghen rade ies. Finally, i mus be noed ha his is an approximaion of he many ha could be underaken and he resuls are limied in several aspecs. Among he consrains, i should be noed ha he sudy esimaes he expor demand funcion a an aggregae level, and if one breaks up he analysis by secors or expors iems, or even for periods of ime, some variaions may probably be found. Noneheless, he paper sill represens a conribuion, albei a small one, in he process of undersanding how cerain facors may influence he rade flows from MERCOSUR o he EU and why he EU is no ye an aracive desinaion for MERCOSUR expors. 16

17 References Alexander, S. (1952). Effecs of a Devaluaion on a Trade Balance, IMF Saff Papers, 2: Algieri, B. (2004). Price and Income Elasiciies of Russian Expors, The European Journal of Comparaive Economics, 1 (2): Bahmani-Oskooee, M. and J. Alse (1994). Shor-run versus Long-run Effecs of Devaluaions: Error-correcion Modeling and Coinegraion, Easern Economic Journal, 20: Bahmani-Oskooee, M. and F. Niroomand (1988). Long-run Price Elasiciies and he Marshall-Lerner Revisied, Economic Leers, 61: Balaguer Franch, M. and I. Marínez Zarzoso (2000). Análisis de los flujos comerciales Unión Europea-MERCOSUR, Información Comercial Española, 788: Barraud, A. y Jacobo, A. (2009), Exploring Recen Trade Flows Beween MERCOSUR and he EU: Wha Does an Expor Demand Funcion Tell Us?, IX Arnoldshain Seminar of he Inernaional Sociey for Comparaive Economic Sudies, Puna del Ese, Uruguay, manuscrip. Barraud, A. and A. Jacobo (2008). Una aproximación a los inercambios comerciales enre el MERCOSUR y la UE, Anales de Economía Aplicada, J. Tricás Preckler y C. Moslares García (Ed.), Dela Publishers, Madrid: Bouzas, R., R. Torren and Moa Veiga (2002). In Deph Analysis of MERCOSUR inegraion, is Prospecs and he Effecs on Marke Access for EU goods and Invesmens, in hp://mkaccdb.eu.in/sudies/sudy/32.doc. Calfa G. and R. Flores (2006). The EU- MERCOSUR Free Trade Agreemen: Quanifying Muual Gains, Journal of Common Marke Sudies, 44:

18 Canavella Jordá, M., A. Cuadros Ramos, I. Fernández and C. Suárez Burgue. (2003). A comparaive analysis of European expor demand elasiciies in MERCOSUR and NAFTA, VI Encuenro de Economía Aplicada, Granada, manuscrip. Cuadros Ramos, A., M. Canavella Jordá, J. Fernández Guerrero and C. Suárez Burgue (1999). Relaciones comerciales Unión Europea- MERCOSUR: Modelización de una función de exporación, Boleín Económico del ICE, 782: Devin, R. and R. Ffrench-Davis (1998). Towards an evaluaion of regional inegraion in Lain America in he 1990s, INTAL-BID, Working Paper 2. Dickey, D. and W. Fuller (1981). Likelihood Raion Saisics for Auorregressive Time Series wih a Uni Roo, Economerica, 49: Dickey, D. and W. Fuller (1979). Disribuion of he Esimaors for Auorregressive Time Series wih a Uni Roo, Journal of he American Saisical Associaion, 74: Engle, R. and C. Granger (1987). Coinegraion and Error Correcion: Represenaion, Esimaion and Tesing, Economerica, 2: Enders, W. (1995). Applied Economeric Time Series, John Wiley and Sons, Nueva York. Esevadeordal A., J. Goo and R. Sáez (2000). The New Regionalism in he Americas: The case of MERCOSUR, INTAL-BID, Working Paper 5, Buenos Aires. Febrel Melgarejo, I. (1996). La inegración económica de MERCOSUR, Boleín Económico del ICE, 2516: González, M., L. Peña and V. Vera (2004). La demanda de exporaciones: Qué dice la evidencia empírica? Análisis para el caso ecuaoriano ( ), Revisa Tecnológica, 17:

19 Golsein, N. and M. Khan (1985). Income and Price Effecs in Foreign Trade, en R. Jones y P. Kenen (Ediores): Handbook of Inernaional Economics, Vol. II: Gosh, S. and S. Yamarik (2003). Does rade creaion measure up? A reexaminaion of he effecs of regional rade agreemens, manuscrip. Grosman, G. and E. Helpman (1995). The Poliics of Free Trade Agreemens, American Economic Review, 85: Jacobo, A. and A. Barraud (2010). El comercio enre el MEROCOSUR y la UE: Siuación acual y perspecivas, Anales de Economía Aplicada, Sevilla Jiménez, M. y T. Torregosa Marí (Coord.), Dela Publishers, Madrid. Johansen, S. and K. Juselius (1990). Maximum Likelihood esimaion and Inference on Coinegraiuon wih Applicaions o he Demand for Money, Oxford Bulein of Economics and Saisics, 52: Johansen, S. (1988). Saisical Analysis of Coinegraing Vecors, Journal of Economic Dynamics and Conrol, 2: Hooper, P., K. Johnson and J. Markez. (2000). Trade Elasiciies for he G-7 Counries, Princeon Sudies in Inernaional Economics, 87: Keele, L., S. De Boef (2004). No Jus for Coinegraion: Error Correcion Models wih Saionary Daa, manuscrip. Khan, M. and M. Knigh (1988). Impor Compression and Expor Performance in Developing Counries, Review of Economics and Saisics, 70: Khan, M. and K. Ross (1975). Cyclical and Secular Income Elasiciies of he Demand for Impors, Review of Economic and Saisics, 57:

20 Iglesias, E. (2005). Cuaro décadas de inegración regional en América Laina y el Caribe, discourse pronounced in he seminar Los 40 años de la creación del INTAL, Buenos Aires, manuscrip. Marquez, J. and C. McNeilly (1988). Income and Price Elasiciies for Expors of Developing Counries, Review of Economics and Saisics, 45: Mena, H. (1995). Pushing he Sisyphean Boulder? Macroeconomeric Tesing in Lain American Counries, Review of Income and Wealh, 10: Muscaelli, V. and S. Hurn (1992). Coinegraion and Dynamic Time Series Models, Journal of Economic Surveys, 6: Nicia, A. and M. Olarreaga (2006). Trade, Producion and Proecion , manuscrio. Okoh, R. (2004). The Global Marke Place: How far can Nigeria go wih he presen Non-oil Produc Mix?, African Developmen and Povery Reducion: The Macro-Micro Linkage, Forum Paper. Olmedo, M. and A. Gómez-Crespo (1996). Relaciones UE-MERCOSUR, Boleín Económico del ICE, 2489: ONU- CEPAL. Esudios Económicos de América Laina, various issues. Osry, J. and A. Rose (1992). An Empirical Evaluaion of he Macroeconomic Effecs of Tariffs, Journal of Inernaional Money and Finance, 11: Paiva, P. and R. Gazel (2003). Mercosur: Pas, presen and fuure, Nova Economia, 13: Ramanahan, R. (1992). Inroducory Economerics wih Applicaions, Harcour Brace Jovanovich, Miami. 20

21 Reinhar, C. (1995). Devaluaion, Relaive Prices and Inernaional Trade, IMF Saff Papers, 42: Rose, A. (1991). The role of exchange raes in a popular model of inernaional rade. Does he Marshall- Lerner condiion hold?, Journal of Inernaional Economics, 30: Rose, A. (1990). Exchange Raes and he Trade Balance: Some Evidence from Developing Counries, Economic Leers, 34: Senhadji, A. and C. Monenegro (1999). Time Series Analysis of Expor Demand Equaions: A Cross-Counry Analysis, IMF Saff Papers, 46: