POLICY RESEARCH WORKING PAPER 2157 How Regional Blocs Affect Price data on exports to Brazil from countries excluded from Excluded Countries MERCOSUR show that preferential trading agreements hurt nonmember The Price Effects of MERCOSUR countries by compelling them to reduce their prices to meet Won Chang competition from suppliers L. Alan Winters within the regional trading bloc. FILE COPY The World Bank Development Research Group Trade August 1999 Poi (. RvsFAR(CII WORKING, PAPER 2157 Summary findings The welfare effects of preferential trading agreements are prices of nonmembers' exports to the bloc. T hese can he most directly linked to changes in trade prices - that is, explained largely by tariff preferences offerec to a the terms of trade. country's partners. Chang and Winters use a simple strategic pricing game Focusing on the Brazilian market (by far th largest in itn segmented markets to measture the effects of MERCOSUR), they show that noninembers' ?xport MERCOSLJR on the pricing of "nonmenmber" exports to prices to Brazil respond to both most-favorable-nation the regional trading bloc. Working with detailed data on and preferential tariffs. Preferential tariffs inc uce unit values and tariffs, they find that the creation of reductions in nonmember export prices. MERCOStJR is associated with significant declines in the This paper - a product of Trade, Development Research Group - is part of a larger effort in the group to un derstand the effects of regional integrationi. (Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Lili Tabada, room MC3-333, telephone 202-473-6896, fax 202-522-1159, Intcrniet address Itabada@ worldbank.org. Policy Research Working Papers are also posted on the Web at http://wwNw.worldbanik.org/html/ dec/Publications/Workpapers/lhome.html. The authors may be contacted at wclhang(iworldbank.org or l.a.winters (( sussex.ac.uk. AUgust 1999. (57 pages) | The Policy Research W'orkinig Paper Series disseminilates the findings of uwork in progress to enconrage the exchange of idea.: abot |development issues. Ant objective of the series is to get the findings out quickly, even ifthe presentations are less than fully polish d. The papers carry the namies of the authors and shozild be cited accordinigly. The findinigs, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the viewt of the W(orld Bank, its Executive Directors or tbe COU1ntries they represent. Produced by the Policy Research Dissemination Center How Regional Blocs Affect Excluded Countries: The Price Effects of MERCOSUR* Won Chang f and L. Alan Winters Keywords: Regional Integration; Terms of Trade; Imperfect Competition; MERCOSUR JEL classification: F13; F15; C33 t Won Chang is a research student at Columbia University, E-mail: wchang@worldbank.org. $ L. Alan Winters is Professor of Economics, School of Social Sciences, University of Sussex, Falmer, BRIGHTON, BN1 9QN, UK. Tel.: +44 (0) 1273 877273; Fax: +44 (0) 1273 673563/678466; E-mail: L.A.Winters@Sussex.ac.uk; Centre for Economic Policy Research, 90-98, Goswell Road, London, ECIV 7DB, UK; and Centre for Economic Performance, London School of Economics, Houghton Street, London WC2A 2AE, UK. * This work was partly conducted while the authors were Consultant and Research Manager in the Development Research Group of the World Bank. The views expressed in this paper are those of the authors and should therefore not be attributed to the World Bank or its member governments. The authors are grateful to Kyle Bagwell, Jagdish Bhagwati, Stephen Cameron, Richard Clarida, Antoni Estevadeordal, Junichi Goto, Ann Harrison, Ken Leonard, Will Martin, John McLaren, Andrew Newell, Robert Mundell, Marcelo Olarreaga, Maurice Schiff, Forhad Shilpi, Isidro Soloaga, Anthony Venables and Stan Wellisz for excellent comments and participation in the seminars at the Inter-American Development Bank, the World Bank, the US International Trade Commission, the University of Sussex and Columbia University. 1. INTRODUCTION 1.1 Introduction Preferential Trading Arrangements (PTAs) have now become an integral and enduring aspect of the multilateral trading regime. Between 1990 and 1997, 87 PTAs were notified to the WTO, and nearly all signatories of the WTO are currently members of at least one PTA. Despite such widespread existence, concerns continue about the welfare impacts of PTAs, especially on excluded countries. The effects of PTAs on the volume and quantities of trade are studied quite frequently but, as Winters (1997a, b) argues, these variables are not a reliable guide to welfare effects for non-member countries. The latter are more directly related to price effects, and of these there are few studies. Indeed, there is, to our knowledge, no published ex post study of the price effects of a PTA on its trading partners. This paper studies one of the most recently formed and controversial customs unions, MERCOSUR (between Argentina, Brazil, Paraguay, and Uruguay). It examines the effect that MERCOSUR has had on the prices of its imports from non-members, assuming that those countries export to two segmented markets, (1) Brazil and (2) rest of the world, in an imperfectly competitive setting with differentiated products. We concentrate on the Brazilian import market since it is a large market for imports, by far, the largest in MERCOSUR and it provides good data over the time period of interest.! We ' Yeats (1998) first raised the question of whether MERCOSUR may be a concern for non-members, since the most rapidly growing intra-MERCOSUR exports appear to be in products in which members do not have 1 postulate that changes in Brazilian m.f.n. tariff rates led directly to price changes by non- member firms exporting to Brazil, and that tariff preferences offered to members, e .g. Argentina, lead to additional 'strategic' price responses within the Brazilian market. We seek to identify both such responses in commodity-level import data from Brazil and in export data from its major overseas suppliers. MERCOSUR nations have made significant tariff adjustments over our sample period (1989-1996). In addition to unilateral reforrns over 1989-95, they largely abolished tariffs on imports from partners over 1991-95, as governed by the Treaty of Asunci6n, 1991. MERCOSUR's common external tariff (CET) is based on the Ouro Preto Protocol, agreed, after much contention, at the end of 1994 and implemented over the following two years. The different phasing of these adjustments, plus the exceptions to both the CET and internal free trade-see Olarreaga and Soloaga (1998)-mean that the margins of preference on internal trade show considerable variation both through time and across commodities. This helps us to identify their effects empirically. In the remainder of the paper, Section 1.2 summarizes the literature on the effects of PTAs on non-members and on identifying price effects empirically. Section 1.3 discusses some stylized facts and descriptive statistics on the major exporters to the Brazilian market. The formation of MERCOSUR seems likely to have had an immediate effect on the pricing of non-member exports to the Brazilian market. The Treaty of Asunci6n cut members' internal tariffs by more than 50% of the m.f.n. rate at the end of a comparative advantage. Nagarajan (1998) argues instead that intra-regional trade should be compared w ith extra-regional imports, not extra-regional exports, and that by focusing on the latter, Yeats may exaggerate the effects of MERCOSUR. Our work is quite different, referring to the prices not the values of trade flows. 2 1991, with the rest of the cut to zero following over the next four years. Intuitively, the response to such a large discriminatory tariff cut should be for members to increase their pre-tariff prices, while non-members reduce theirs. Section 2 briefly presents a model of this process. From this we derive reduced form estimation equations and a comparative statics exercise (Appendix I) to interpret their coefficients. The model has two firms, a 'non-member' and a 'member' firm, exporting a differentiated product to the Brazilian market. The two firms respond to each other's prices (as well as to their own tariffs, exchange rates, and wages), playing a Bertrand pricing game within the Brazilian market. We explore the game by examining relative member and non-member prices in Brazil, and, for certain exporters, the relative prices of exports to Brazil and to other markets. Section 3 presents the empirical implementation of the reduced form equations solved in section 2. It also provides details of MERCOSUR's tariff policy during the integration period and of the data and their limitations. Section 4 examines the final results which suggest strongly that m.fn. tariff changes and preferential tariffs both affect supplier prices significantly, and that MERCOSUR's preferential tariffs caused significant declines, ceteris paribus, in the prices of non-members' exports to Brazil. 1.2 Brief survey and motivation for the study One of the major influences on the welfare of any trading economy is its terms of trade, and thus questions surrounding trade policy should be concerned with this variable. 3 But given its importance in theory this issue is addressed surprisingly rarely in empirical studies. A seminal contribution was Kreinin (1961) who considered the effects of US m.f.n. tariff concessions during the post-war years. Kreinin notes that a reduction in US tariffs would most immediately affect import prices and that only through this medium would changes in the volume of imports occur. He also shows that US m.f.n. tariff concessions did indeed lead to considerable changes in foreign export prices.2 By the same token the empirical analysis of the effects of PTAs should be at least as concerned with price as with volume effects. An elegant but relatively unremarked theoretical examination of the terms of trade effect of regional integration is given by Mundell (1964). He elucidates the terms of trade effects in a 3-country model in which goods are gross substitutes, and in which price changes occur to restore balance of payments equilibrium after an initial preferential tariff shock occurs. He shows that for a single tariff change by one member, the preferred exporting partner's terms of trade unambiguously improve, while the excluded country's deteriorate. The net effect of the active country's tariff concessions on its own terms of trade is ambiguous, but when two countries swap preferential concessions, as in a PTA, they collectively improve their terns of trade vis-a-vis the rest of the world. More recent studies focusing on PTAs such as Bagwell and Staiger (1998, 1999) also show that the multilateral negotiations of the GATT and its principles of reciprocity and non-discrimination foster efficient outcomes which allow governments to escape from 2Kreinin states that "less than a third...of the tariff concessions granted by the US were passed on to the IJS consumer in the form of reduced import prices, while more than two-thirds.. .accrued to the foreign suppliers 4 a terms of trade driven Prisoners' Dilemma. The authors argue that PTA formation could enable member countries to exploit greater market power over their terms of trade and potentially undermine the efficient outcome of multilateral negotiations. The last result is potentially very significant, for the terms of trade is by far the most direct way in which PTAs affect the rest of the world (RoW). Precisely paralleling Kreinin's complaint, the usual empirical approach to assessing the effects of a PTA is to ask whether, as a result of integration, the RoW's exports to the integrating bloc increase (which is held to be good) or decrease (bad). Winters (1997a) shows that this is a very inadequate indicator: first, RoW welfare will be related to its imports not its exports, and second, in a competitive economy, marginal changes in quantities hardly matter, whereas changes in the prices of traded goods matter considerably.3 Given that the theoretical literature focuses so heavily on terms of trade effects, it is surprising that ex-post studies which examine these variables are so very sparse. Turning to quantitative studies of the effects of integration, Winters (1997b) observes that the RoW's terms of trade do figure in a number of ex ante studies (although frequently with little emphasis), but that no ex post study addresses the issue. Winters and Chang (forthcoming) started to do so in the case of Spanish accession to the EC, but were severely hampered by a number of intractable data difficulties. This paper continues our efforts in a much more satisfactory empirical environment and generates stronger and more and improved the terms of trade of the exporting nations." 3 Winters also argues that, contrary to the common belief, Kemp and Wan (1976) said nothing about whether RoW's welfare increases or decreases in the face of a PTA. They showed how it could be kept constant, completely obviating the need to discuss its determinants. 5 interesting results. Our focus is primarily on how regional schemes affect excluded countries: specifically, the effect that MERCOSUR has had on the prices of imports in Brazil since 1991. A useful empirical literature, on which we build, relies on the micro-foundations of imperfectly competitive and segmented markets. The 'pass-through' literature attempts to explain the lack of import price changes following changes in the exchange rate, and the consequent implication that foreign suppliers' markups change.4 Feenstra (1989) estimates a markup model for the US markets for motorcycles and trucks and obtains the usefi.l result that changes in the exchange rate and in tariffs have equal effects on the net price of imports--the so-called 'symmetry' hypothesis. Feenstra, however, considered only the rivalry between domestic and imported varieties and so examined only the pass-through cf the m.f.n. tariff. For the purpose of examining PTAs, however, we have to model the pricing game that occurs between rival foreign suppliers within a market under consideration. In imperfectly competitive settings, a firm's pricing depends not only on the tariff charged on its own product, but also on that charged on its rivals'. If a member- country firm receives a preferential tariff concession it becomes more competitive in PTA markets, and non-member firms are likely (although not bound) to reduce their prices in compensation. With this in mind we move on to present some stylized results and descriptive statistics. 6 1.3 Stylized results and descriptive statistics We present three simple calculations of the mean changes in prices (unit values) since the formation of MERCOSUR5: for various suppliers, the average price of exports to Brazil relative to those to non-integrating markets (RoW); the prices of exports to Brazil and RoW in absolute terms; and, using Brazilian data, the relative prices of imports from members (Argentina) and non-members. To render commodities comparable, the starting year price has been normalized to be I for each commodity so that we are essentially measuring price changes. To be precise we estimate and plot the following statistics: in Figure 1: - In n(s 2 D, i=(1,...,N) and t=(l,...,T), N j=, Pl90/P2i90 in Figure 2: IN n 5l$i) , i=(1,...,N) and t=(l,...,T), in Figure 3: IN ,i=(1,...,N) and t=(1,...,T). N =1 Pl1i'90 /p,i90g 4 Several recent studies analyze incomplete pass-through in the face of exchange rate fluctuations: for example, theoretical papers by Baldwin (1988), Dornbusch (1987) and Krugman (1987), and cross-sectional industry empirics by Knetter (1989), Froot and Klemperer (1989) and Schembri (1989). 5 Because no price data are available we have to use unit value data, but since these are available at the 6- digit level of the Harmonized System (HS-6) which distinguishes 5113 commodities, we can have reasonable confidence in their accuracy. The 6-digit Harmonized System became the standard classification for trade and tariff data across countries starting in 1989. Unfortunately, many countries started reporting well after that date, and there is no other way to obtain data of this level and precision for earlier years. 7 Where the first subscript, I or 2, represents prices paid in Brazil and RoW respectively, the second, i=l,...,N, the commodity, and the third, t=l,...,T, time, with the beginning year as base. The bars above the prices indicate that these are pre-tariff prices, and the superscript $ denotes prices in dollars. We have averaged prices only over the set of commodities for which we have observations for all years for both markets or suppliers. Figure 1 presents mean export prices for four major exporters to Brazil and RoW: the USA (for which 1356 commodities were exported to both markets in all years), Japan (580), Korea (99), and Argentina (686). The broken lines give the 95% confidence interval about the means. To infer from Figure 1 an effect of MERCOSUR on prices, we have implicitly to employ RoW as the 'anti-monde'. On this basis non-members' relative prices of exports to Brazil declined by approximately 15% between 1991 and 1996.6 Conversely, for the integrating partner, Argentina, relative pre-tariff prices to Brazil increased. This latter result is not significantly different from no change, however, possibly because data on the critical years 1991 and 1992, during which the major shocks occured, are missing. It is also interesting to see the pattern of the absolute export prices in Figure 2. For the USA and Korea absolute export prices declined by about 10% following the shock ol- MERCOSUR, and then began to rise somewhat afterwards. For Japan, absolute dollar prices to Brazil rose (presumably reflecting the yen's appreciation) but by less than exporn prices in general. 6 Similar results for USA exports have been obtained using the data provided in Feenstra (1997). 8 Finally, Figure 3 shows relative member/non-member import prices in the Brazilian market. Argentina's pre-tariff prices rise relative to USA, Korea, and the world as an aggregate. Japan is different presumably again explained by the appreciating Yen during the 1990-1995 period.7 These descriptive statistics match our a priori expectations surprisingly well. Moreover, they refer to significant volumes of international trade. In 1996, for example, Brazil imports of goods amounted to $56.5 billion: $12.5 billion from the USA (22.2% of the total), $7.1 billion from Argentina (12.6%), $5 billion from Germany (8.8%), $3.1 billion from Italy (5.4%), and $2.9 billion from Japan (5.1%). Other large suppliers examined are Korea and Chile, which account for $1.3 and $1.0 billion, (with 2.2 and 1.8% import share) respectively. At the commodity level the USA has a share of 10% or more of Brazilian imports in 60% of the HS-6 headings, Argentina in 17%, Germany in 30%, Italy in 16%, and Japan in 12%. Korea and Chile each have approximately 5% of HS-6 headings which have 10% or greater import share. 2. THE MODEL 2.1 Export Pricing under Imperfect Competition and Segmented Markets While the pricing figures above are very informative, they are also very crude, and so we now include a series of controls to model the effects of MERCOSUR more formally. 7 The Yen appreciated by 54% from 144.8 in 1990 to 94.1 Yen/$ in 1995. 9 We use a parsimonious model of export pricing to illustrate the effects we expect to find. For each good we distinguish two segmented markets, Brazil and the Rest of the Worlcl (RoW), and two exporting firms, a non-member firm from outside MERCOSUR and a member firm from inside (always Argentina in our case).8 The firms supply differentiatecl products9 and maximize profits in their own currency by manipulating duty-paid prices in their markets (p). They take their input costs, exchange rates and tariffs as given. Costs (c(x,w)) are homogeneous of degree one in the price of a composite factor, loosely referred to here as the wage (w). Thus c (x, w) = wc(x), where x is output and c(x) is unil costs. The demand for the non-member's differentiated product in Brazil (market 1) is given by, xI(p1,p1t,Q1,YI), a function of the its own price, p, its major rival's (Argentina) product price, p*, the aggregate price index, Q, and nominal national income, Y, in Brazil. The demand for its product in the RoW (market 2) is a function of its own price, the aggregate price level and national income in RoW, x2(p2,Q2,Y2). We are assuming here that Argentina is a sufficiently large supplier to the Brazilian market that the non-member firm's demand may be related to Argentina's prices, but that it is so insignificant in RoW markets that no separate Argentina price effect will be identifiable.' The non-member firm's objective function and first order conditions may thus be written: ' We concentrate on the two largest traders of MERCOSUR, Argentina and Brazil because data on Paraguay and Uruguay are so sparse. 9 We use Arnington's (1969) distinction between a 'good' and 'product'. 'Goods' are distinguished only by kind whereas 'products' are distinguished by kind and origin of supply. 10 Max [PI XI(PI 'PX l Ql XY ) + e2 P2X2 (P2,I Q2, Y2 ) - Cl (XlI)W - C2 (X2 )W(l) P,,P2 T2 with F.O.C.s plllw+ W Clx(XI(P(. PiIQQIY D)) =0 i=- P' (la) P21 + C W]- 2e C2x (X2 (P2 Q22)) = ° 772p = &2 P2 (lb) where, in addition to the variables already defined, x1, and t2 are the ad-valorem tariff factors (I+t) charged by Brazil and RoW, and e, and e2, the supplier countries' currency prices of a Brazilian REAL and RoW currency. Note that price elasticities, ,n, and rj2, are affected by the same variables as demand. The member (Argentinian) firm's objective function and first order conditions may be written similarly, except in that demand in RoW depends explicitly on both Argentina and non-member prices, with the latter being treated as exogenous. M4ax' elpx'X) (2) Max* p;x;(p,,P p;Q,,Y,)+ e2. P2X2-(P29P21Q21Y2)-C*(Xl`)W' -C2(X)' 2 F.O.C.s P;t + . ]- , c;(x;(P1,p;,Q,, Y)) = 0 *P. I Pi (2a) 10 Argentina's price is effectively rolled into the general price level in the rest of the world, captured by the world's price deflator Q2. The assumption is not unreasonable. Argentina's share of Brazil's imports exceeds 5% in 22.6% of all HS-6 headings, but in only 3.1% of headings in RoW even using our limited set of exporters to define world sales. 11 I *T *6i ; s[ iw 2* ___(2b) p2[+ C21(x2(p2,p,Q2,YD))=O ; = 17 The first order conditions imply that, for any market and supplier, an increase [n either the tariff or the supplying country's exogenous wage, or a decrease in the exchange rate will increase the marginal cost of delivering exports. The supplying firm must therefore increase its marginal revenue by altering its landed price (p). We have shown in Appendix I, that the nature of this change depends on how the price elasticity of demand changes as costs change. By assuming that the two markets are segmented and have independent cost functions we are making them strategically separable, so that we can develop two separate pairs of price equations." In Brazil: PI =f,(-,Pi,Q1,}1) (1a) P. = Y.(.l,p,zX) (2a) and in RoW: P2 = f2 (-, Q2, Y2) (lb) e2* P; = A ( W. ,P2, Q2, Y2) (2b) e; I There is strong evidence to support that markets are in fact segmented-see for example Knetter (1 989) and Marston (1990). 12 These equations are homogeneous of degree one in costs, competitor's price, the aggregate price and nominal income in local currency. Our assumptions imply that firms play an interactive pricing game in the Brazilian market, solving (la) and (2a) simultaneously, while in RoW the solution is recursive with (lb) affecting (2b) but not vice versa. For estimation purposes we log-linearize equations (1) and (2) and estimate reduced form equations for prices. Thus, . . ln P;= A, +,BIlnWl + 61 nW[1t+a,ln Q, +2i,n Yl (3a) '2 e, el . . lnpj =A ; + 61 In w +/,B; In , '+a lnQa + I lnh Y1 (3b) el e, lnp2 = A2 +±82 ln-+a2 lnQ2 +22 InY2 (4a) e2 w~~~~~~~~~~~~~~~~~~~~~~~~~~~~ In p; = A; + 52 In-w+,8 .I2n-, + a* In Q2 + X21 n Y2 (4b) Equations (4a) and (4b) are written without tariffs in the RoW, i.e., without r2 and T2, because these variables are considered fixed over our sample period, and thus are absorbed into the constant term.'3 Feenstra (1989) uses a variant of equation (3a) to show that for US imports of Japanese trucks and cycles, the long-run pass-through of tariffs and 12 In accordance with the symmetry hypothesis we have given the tariff and wage the same coefficients in these equations, but in our estimations we separate out the tariffs. 13 In fact these rates did actually change a little over time, but much less than in MERCOSUR. In any case, since we have no data on 'world' tariffs, these variables must either be taken as constant, or absorbed into the error term as white noise. 13 exchange rates are statistically identical. Essentially, it focused on the m.f.n. effects, P of the equation, whereas the coefficient of interest in the 'strategic' pricing relevant to PTAs is 81*. If marginal costs are fixed then the expected sign of 81* depends only on how its 'perceived' price elasticity of demand gets altered from the preferential tariff inducedl reduction of its rival's price. If the non-member's demand becomes more elastic, then the optimal response is to reduce price, hence 8,* > O."4 Detailed analysis and interpretations of the coefficients and comparative statics is relegated to Appendix I. While (3) and (4) are estimable directly it is intuitively easier and econometrically more efficient to combine them into a series of relative price equations. Subtracting (3a) from (3b) generates an equation for the relative prices of member and non-member country exports to Brazil. Using the homogeneity assumption, i.e., a1, =1-,6, -E, -Al, ancl a. = 1-f,l -86 - X, we get: ln PL = A + (51 - WV1 wn l +W(V8-t5; ) 1n +Y(X-A )l (5)15 pI e, , , , Q, 14 Using the framework of Bulow, Geanakoplos, and Klemperer (1985), we say that the strategic interaction between these rivals' pricing would be 'strategic complements'. This is what one would expect under price competition. The less likely outcome is also possible: a reduction in the Argentine price can cause the non- member's demand curve to become less elastic, at least locally, hence making it optimal to raise price. Thus 'strategic substitutability' is also a possibility, though probably rare. 15 If we were willing to assume symmetry between (3a) and (3b) such that B, =,6; = ,a = a , and = (5) would simplify to a form expressing relative member/non-member pre-tariff prices for a product as a function of relative costs and the tariff preference margin: In P' = A + (8 - w/e1 + 5 -,6) The bar over the price denotes pre-tariff prices. 14 Figure 4, summarizes the effect of a preferential tariff shock on the relative prices. Panel A describes the 'normal' effect of a preferential reduction of tariffs on a trade partner. The reduction shifts the member's reaction function rf,* to rf2*, less than proportionately if there is incomplete pass through. If this were all, and the new equilibrium were M, the partner price and the price relative (p*/p) would have shifted by no more than the proportionate change in the tariff factor T*. But, in fact, non-partner exporters react to the price change, ultimately shifting equilibrium to N. Here both prices have fallen but the price ratio has fallen by less than at M, and hence certainly less than proportionately to the tariff shock. In terms of equation (5) the elasticity (I3-o6*) lies between 0 and 1. It is also possible to have cases such as panel B, where a very responsive member reaction function causes the elasticity to be greater than 1, and panel C, in which a very responsive non-member implies a negative elasticity. We have shown that the cost elasticities can have a wide range, but it is also clear that in all three panels the non- member price falls. To measure this effect directly we need to isolate 8,*. Turning to the non-members' equations (3a) and (4a) we can compare relative export prices to Brazil and RoW. Applying homogeneity again, Pi / Q, w~~~~ 1i [vi ln_ -c_ 811 . __ T 2n'~~2iQ =c/ln - 1-0f21n[ +,51In * +A In1 L In (6) P2 / Q2 [eQlQ |e2QI e1Ql ] l] Q2 Similarly equations (3b) and (4b) for Argentina imply In__Q__ *+A w~z1 *I I *, WV i wi. Y . P;IQ2 4 LelQ, A eljQe2Q2J 1 eQ, 2 e2Q2J QQ2 (7) 15 In summary, while equation (5) shows how much the non-member's export price changes in Brazil relative the member's, export price, equation (6) shows how much the non-- member export price changes relative to non-member exports to RoW, and (7) how much the member export price changes relative to its export prices to RoW. Our interest is primarily on how the tariff preferences inherent in MERCOSUR have changed Argentinian and non-member export prices--i.e. on the coefficients on t, in these equations. Figures 1 and 2 suggest that there were significant effects through time and (5)-(7) help as to identify whether those are due to tariff changes (MERCOSUR) or to other factors such as exchange rates or costs. 3. EMPIRICAL IMPLEMENTATION 3.1 MERCOSUR Tariff Policy MERCOSUR (Mercado Comuxn del Sur) was established under the Treaty oi Asunci6n, signed by the Presidents of Argentina, Brazil, Paraguay and Uruguay in 26 March 1991 and ratified on 29 November 1991. This treaty extended the borders of the association between Argentina and Brazil dating from 1985 and culminating in The Treaty of Integration, Co-operation and Development of November 1988.16 16 Nogues and Quintanilla (1993) note that regional integration efforts between Argentina and Brazil did not go beyond 'declarative' statements until the Protocols initiated between 1985-1989 on capital goods which was mainly designed to substitute imports from cheaper sources. 16 Article 5 of the Treaty of Asunci6n defined a path of tariff liberalization to achieve zero internal tariffs and the elimination of non-tariff barriers by the end of 1994. The immediate reduction of the internal applied tariff rates was by 47% of the m.fn. rate after the ratification of the Treaty on 29 November 1991. Subsequent preferential reductions relative to prevailing m.f.n. rates were to occur semi-annually and automatically according to the following time table: 54% December 1991, 61% June 1992, 68% December 1992, 75% June 1993, 82% December 1993, 89% June 1994, and finally 100% December 1994.'7 Members were allowed to declare upto 300 exceptions to internal free trade, but by 1995 approximately 95% of intra-regional trade was duty-free--Laird (1997). In fact Brazil had only 27 exceptions and so effectively had open borders for its MERCOSUR partners. MERCOSUR member countries had originally planned to align their external tariffs on the MERCOSUR common external tariff by 1 January 1995. However, this proved politically impossible and little progress was made in defining the CET until the Protocol of Ouro Preto was signed in December 1994. Under the Ouro Preto Protocol the CET was to be introduced beginning 1995. Each member was again allowed an exceptions list, the tariffs on which were to be aligned by 2001 for Argentina and Brazil, and 2006 for Paraguay and Uruguay, see Olarreaga and Soloaga (1998). Brazil named approximately 200 tariff lines in the exceptions list, mainly sensitive industries such as computers, electronics, chemical, agroindustry, textiles, capital goods (machinery), and the automotive industry. Unilateral liberalization followed by this negotiated changes reduced tariffs 7 Article 3, Annex 1, Trade Liberalization Program, Treaty of Asunci6n, 1991. 17 substantially in MERCOSUR countries, from an average of 50% in 1988 to a CET average of 12% in 1995. However, it remained the case that trade policy in Brazil was subject to vigorous debate and to frequent changes to meet short-run political objectives. For example, tariffs on textiles, toys and motor vehicles in particular were increased to 70% for non-members in 1995.18 The different phasing of internal and external tariff reductions, the large number of tariff rates and the use of exceptions mean that over 1989-96--our sample period-tariffs and preference margins varied widely over time and commodities. This allows us a good chance of identifying their effects empirically. 3.2 Data Our trade data, used to obtain unit values from quantities and values, were taken from the UN's Comtrade database, at the Harmonized System (HS) 6-digit level. Although it was introduced in 1989 several countries did not start to use HS until somewhat later. Hence our sample periods vary by country. HS 6-digit data offer two major advantages over other sources. First, they are very disaggregated--over 5,000 commodities are distinguished. This helps to minimize heterogeneity within each heading, which in turn improves the quality of our unit value " Motor vehicles have been a special issue within Brazil. The Brazilian government applied special local content rules. Foreign multi-national fiirms which produced vehicles locally were given reduced rates of 35%. Japanese and Korean auto manufacturers in particular claimed that the moves put them at a considerable disadvantage since, not having local plants, they were not able to compete even with other non- member suppliers. These types of local content rules prompted several multi-nationals to set up automobile 18 data, and reduces the need for tariff averaging within headings-see next paragraph. Second, trade and tariff data match very well at the 6-digit level, because at this level the HS classification is universal across countries. At finer levels of disaggregation codes are country-specific."9 The tariff data were provided by UNCTAD and the MERCOSUR Secretariat-to whom we are grateful. Over the years 1989-1994 Brazil and Argentina defined their tariff data at HS 10-digits, while the Common External Tariff (CET) of 1995 and 1996, and the exceptions listed in the agreement of Ouro Preto Protocol, are defined at the HS-8 digit level. In order to concord the tariff and the price data we truncated the tariff codes up to the 6-digits and took simple averages. This averaging within the HS-6 level is not a serious problem because there is very little variation in tariffs within the HS-6 digit level. As an empirical exercise on the price effects of integration, a study of MERCOSUR is relatively problem-free. There are few problems of changes in quotas confounding price movements, since on signing of the Treaty of Asunci6n, all non-tariff barriers were to be removed for all trade including imports from non-members.2" Products having NTB measures before integration which could potentially affect prices over the series were plants within the MERCOSUR region. For details see Latin American Monitor-Brazil and Latin American Regional Report-Brazil, August (1996). '9 There is a slight discrepancy between the HS-6 digit codes in HS92 and HS96. Commodities have been deleted when such concordance problems arise between years. 20 See Laird (1997) and Frischtak, Leipziger, Normand (1996). The abolition was not entirely clean in practice, however. There are some instances where quotas may have been used, particularly in textiles. Due to heavy losses and high unemployment in the Brazilian textile industry there was great pressure to impose quotas and high duties, especially against Southeast Asian countries. Quota protection and local content rules were threatened by Brazil in the automobile industry as a means to attract foreign direct investment, but 19 deleted from our sample altogether.2' Applied tariff rates are entirely ad valorem charged on the c.i.f. value of imports. There were no major prior associations between these countries and therefore changes in tariff preferences are defined by the Treaty of Asuncian and the Ouro Preto Protocol. The first shock comes at the beginning of the transition period at the very end of 1991, and the effects can be seen in 1992, and 1993. Then another major shock comes in 1995, when the CET is implemented with exceptions which tend to increase tariffs on non-members.22 Internal tariff rates were calculated as the m.f.n. rate multiplied by (1 - average reduction rate for that year). Since the reductions take place semi-annually (see above) we have to average them for each year to match the annual trade data. The following chart provides a typical transition for most commodities, although we have incorporated the exclusions to this rule included in the agreement of Ouro Preto Protocol in December 1994, which took effect in 1995, as well as the changes that occurred subsequent to this Protocol.23 after further negotiations with Argentina they were revised and ceased to be binding--see Latin American Monitor: Southern Cone Report, February 1996. 21 This list, obtained from UNCTAD, includes products under quantity control measures such as quotas, and voluntary export restraints. 22 Most of the applied m.f.n. tariff rates charged to non-members including exceptions were compiled by UNCTAD. We are grateful to Aki Kuwahara of UNCTAD and Jerzy Rosanski of the World Bank for their help in obtaining them. Detailed information can be obtained in United Nations Conference on Trade and Development (UNCTAD) "A User's Manual for TRAINS", 1996. The internal tariff rates are estimated using these m.f.n. rates and the Treaty of Asunci6n's time path. Brazil's detailed import and export data disaggregated by source country were also provided by Aki Kuwahara. Argentina's trade data, which was used in the intermediate stages of our research, was provided by Tony Estevadeordal and Raphael Comejo of the Inter-American Development Bank to whom we are also grateful. 23 This list was provided by the MERCOSUR Secretariat. 20 m.f.n rate Internal rate t89 t89 t9o t9o t91 t91 t92 t92*(1-0.61 ) t93 t93*(1-0.75) t94 t94Z(1e .89) t95 Zero t96 Zero As an illustration of the evolution of tariffs, we have tabulated the tariffs charged to USA (m.f.n.) and Argentina (partner) and the preference margin in Table 1.24 These are HS 6- digit tariffs truncated up to 2-digits and then averaged (unweighted) across the nine categories specified in Appendix II. Some notable features are evident even at this aggregated level. First, although the m.f.n. rates are generally falling after 1991, there are also some increases in 1995 and 1996 as a result of Ouro Preto--in HS Chapters 16-27 (prepared foodstuffs), 41-63 (which includes textiles), 64-83 (which includes footwear, headgear, glass etc.,) 86-89 (which includes vehicles, aircraft, vessels, transportation equipment, etc.) and 93-96 (which includes toys). The increases in 1995 and 1996 were within Brazil's overall binding commitments at the WTO. Second, while m.f.n. rates decline from 1991 to approximately 1994 and then stabilize or rise, the tariffs on partners continue to fall until 1995. Thus member and non- member tariffs are not perfectly correlated, which greatly facilitates the identification of 4 This table is confirmed by Laird (1997), but unlike Laird, who averages all tariff data available, we provide the average tariffs only for the commodities for which US export price data are available over the years 1991-1996, since these are the tariff rates used in the estimation for USA export pricing behavior in the following section. 21 separate effects econometrically. Third, preference margins did not rise monotonically as MERCOSUR was implemented. Finally, member and non-member wage rates or labor costs could not be obtained at the industry level and certainly not at the commodity level over the time perioil necessary in this analysis. Thus in order to obtain data and also to recognize a wider range of inputs than just labor, we used GDP deflators to proxy export country costs (using aggregate export weights to Brazil to construct non-member costs). These variables could easily be converted into the currency of the importer.25 For the aggregate price index in Brazil and RoW we employed GDP deflators. 4. RESULTS 4.1 (A) Relative Import Prices in Brazil Our main results appear in Tables 2 through 6. As well as pooling all commodities, these also consider 9 sub-groups of commodities. The disaggregation allows scope for some variability in the degrees of competition and product substitutability (differentiation) across sectors. In every panel all variables are expressed in natural logs and as deviations from commodity-specific means. This is equivalent to allowing commodity-specific fixed effects. We also corrected for heteroskedasticity by collecting the residuals from the 25 The GDP deflator for the world in dollar terms was taken to be an export weighted average of the GD:P deflators of supplying countries, with weights coming from the International Monetary Fund, Direction of Trade Statistics: Yearbook (1996, 1997). The representative countries included in the weighted average are: 22 estimated unweighted equations and reweighting each of the variables by the inverse of the estimated commodity-specific residual standard deviations.26 This procedure improves the efficiency of our estimates and permits more accurate inference. First we examine the prices of Brazil's imports from Argentina relative to a series of non-member countries, equation (5).27 To try to isolate the effects of most interest, we have separated out the tariff effects.28 These initial estimates appeared to suffer very seriously from multicollinearity. This seemed traceable to the coefficients of the real income terms (Y/Q), which regularly had variance inflation factors above 20 and frequently much higher. The problem is three-fold. First, Brazil's measured real income was rather stable over 1989-96 so that there was little identifying power in the series. Second, with inflation reaching 2308 % in 1994, it was unclear whether deflated nominal income is really very informative anyway. Third, all the explanatory data except tariffs refer to macroeconomic variables (the exchange rate, costs, aggregate prices and incomes) which are invariant over commodities. Thus in effect we are seeking to identify three effects with eight observations. Belgium, Bolivia, Canada, Chile, China, Colombia, Denmark, France, England, Germany, Indonesia, Italy, Korea, Mexico, Malaysia, Netherlands, Peru, Philippines, Singapore, USA, Venezuela. 26 The homoskedasticity assumption was tested by using the log-likelihood ratio test and the null was always strongly rejected. The procedure adopted is a two step Feasible Generalized Least Squares (FGLS) estimation, which is unbiased. The coefficient estimates in the first stage regressions were quite similar to the cross commodity heteroskedasticity corrected set and can be obtained from the authors on request. The uncorrected estimations tended to yield very low R-squares, however. 27 Brazil is used as the reporter country for the data used in Table 2A and 2B, and therefore the data run from 1989-1996, with the exception of Germany which Brazil only reports from 1991-1996. The countries represented in Table 2 make up most of the imports to the Brazilian market. 23 We have adopted two approaches to the multicollinearity problem. In estimate (A) we have assumed that 2A = *,V and dropped the real income term. Strictly this implies that for each good, the Argentinian and non-member varieties have the same income elasticities of demand, but it is better thought of as merely as indicating that we have insufficient information to identify different elasticities. In estimate (B) we have swept out the macroeconomic effects with time dummies for each year, leaving the tariff effects as the only explanatory variables. Essentially relative Argentinian and non-member prices comprise a time-related component, which we isolate and ignore in these equations, and a commodity-specific component related to the two tariff rates. With some exceptions, the estimates of the tariff effects--our variables of interest--are similar between the two approaches. Tables 2(A) and 2(B) report the results from the overall pooled samples. They display a number of interesting features. First, tariffs matter for firms' pricing decisions. Both member and non-member tariffs are strongly statistically significant in explaining the relative prices of imports within the Brazilian market. Nearly all of the overall results are highly significant, have the correct signs and have reasonable magnitudes according to our discussion above. Second, Brazil's tariff factor on Argentinian imports (T*) affects relative member/non-member prices less than proportionately in ten out of the twelve cases. With the exception of Mexico and Japan, the member's tariff coefficients are less than one in 2S The results of equation (5) with the tariffs combined with the rest of costs are shown in the Appendix, Table Al. 24 Table 2A and not significantly above in Table 2B. The remaining estimates range from 0.282 for Korea to 0.884 for France, and all are statistically significantly different from one. These latter results reflect some convex combination of (a) Argentinian firms passing only part of the tariff cut onto consumers (partial pass-through) and non-members holding their prices constant (8o*=0), and (b) Argentinian firms passing the tariff cut through fully (P,*=1) and non-member firms partially following iheir prices down (0<68* r 77)-(p M-P.)(p -YqUp )- Two reduced form pricing equations for the non-member and member firms which are also homogeneous of degree one in the costs, general price and income, are shown here as equations (10), and are analogous to (3a) and (3b). 38 pi =I.-2,I+6 i2 +, .y +a, *Q, (lOa) s*' - *5 ') ( ~A) , ~(09Y -Y*'7y)(9p*r p7)-(9 -Y)(9; -Y*" a1] = 1-J- 81 ;- 21 = ± 2,+B;2; +2;*+a;.Q, (lOb) AA ,L((P-r v7p)(O -my) -(Op - r l7p)(0- 1 = 1 - a} -fi1 -i; 39 To simplify these unwieldy elasticities, assume that the marginal costs of both member and non-member firms are fixed, y, y*=O. Then the elasticities can be neatly defined as: TPT0 9. O- 9; ' 0 . OjZ P ppPp p p Assuming the denominator is positive, the signs of these elasticities depend on the signs of the elasticity of an exporter's 'marginal revenue' with respect to its own price, and its rival's price. The denominator being positive merely implies that "own" effects cn marginal revenue are greater than that of the "cross" effects. The elasticity of marginal revenue with respect to own price is, op = dnP=,- p2I97) l A( p) p 4 m z7 t a7 tl+J 11p Its sign only depends on the sensitivity of the own price elasticity to changes in its own price: -V p -- ~2 - __ ( P)= PP + P - 2XP = PPP + P (1- P ) = PP + xP (I1-U7p), which is negative given that demand is not too convex. For instance, given a linear demand curve, raising the price would reduce the price elasticity of demand (higher absolute number, i.e., more elastic). This implies that P > 0, but it is also notable that it is 40 possible to have 31 > 1 when firms behave in a strategic manner even when you have the normal case, (7p/,) < O .32 The sign of the elasticity of 'marginal revenue' with respect to its rival's price (op. ) is essential in determining strategic effects on prices. in p* pp' *p ( O1 O ldp. p p 0 m l 7Pm O4,) 1+lp) '4, li7p where the sign is only dependent on the sensitivity of the own price elasticity to a change in the rival's price, (*pA I( dx a)1 t p) x x@s-U . =-x .-Qx. T'he slope of the 'perceived' price elasticity of demand with respect to the rival's price is positive if the products involved are substitutes, xp. > 0, and the magnitude of xpp. is small. The strategic effect, 8,*, is then also positive.33 Symmetric results will be found for its rival's variables. 3 This is a distinction from Feenstra (1989), since in his outcome the 'normal' case is such that the pass- through (,B) is between 0 and 1. " This result can be expressed more elegantly by using the framework of Bulow, Geanakoplos, and Klemperer (1985) and recognizing that price competition in a Bertrand model is usually considered 'strategic complements', i.e., d' r >O by definition. Differentiating equation (1) by p, and obtaining e, l H(Pl XP;-) =oX where H(p, p`,...)= p( + Il_ iC wT1 as in (la), it is then apparent that ri 14;1 ~ where OX, e, the cross derivative is: d2 n1 el a H.(pH,p')+el d2x, H(p,,p,)>O and so HP.(p,,p0) 0. 'Strategic substitutes' would imply the opposite sign. 41 Let's consider a shift in the member's tariffs, hence a change in the member's price (p*). Since we have assumed that marginal costs are fixed, a shock that shifts this exogenous marginal costs such as a tariff change, will alter its marginal revenue. A decline in the member's tariffs will reduce the landed price, p*, of the member country's product, The non-member will alter his price depending on the effect it has on its marginal revenue. We first begin with the case that is more likely. If a reduction in the price causes the non- member's demand to become more elastic, (or,p /lc) > 0, then the optimal response is tc reduce price (p), where the elasticity is defined here so that it is negative and that more elastic implies that lp is a larger negative number. On the other hand, the less likely outcome which is also possible is that if the reduction in p* causes the non-member's demand to become less elastic, i.e., (977p/@ ) < 0, then it is optimal for this firm to raise its price (p). Both signs are theoretically possible when we are concerned with the price effect due to shifts in the rival's costs. 42 Appendix II: HS-2 Sub-Group Description 01-15 Live Animals, Animal Products,Vegetable Products, Animal or Animal Fats and Oils 16-27 Prepared Foodstuffs, Beverages, Tobacco and Tobacco Substitutes Mineral Products 28-38 Products of Chemicals and Allied Industries, Organic and Inorganic Chemicals Fertilizers, Pharmaceuticals, Perfumery Photographic and Cinematographic Goods 39-40 Rubber and Plastics 41-63 Raw Hides and Skins, Leather, Furskins, Travel Goods, Handbags Wood and Articles of Wood, Manufactures of Straw Textiles and Articles of Textiles 64-83 Footwear, Headgear, Umbrellas, Walking Sticks, Articles of Human Hair Articles of Stone, Plaster, Cement, Mica or similar Materials,Ceramics, Glass and Glassware Natural or Cultured Pearls, Precious Stones, Precious Metals, Jewelry Base Metals, Articles of Base Metals, Iron,Steel, Aluminum, Zinc, Lead, Tin, Copper, Nickel 84-85 Machinery and Mechanical Appliances, Electrical Equipment and Parts Sound Recorders and Reproducers Nuclear Reactors, Television Image and Sound Recorders 86-92 Vehicles, Aircraft, Vessels and Associated Transport Equipment Optical, Photographic, Cinematographic, Measuring, Precision Medical Instruments Clocks, Watches, Musical Instruments, 93-96 Arms and Ammunition Miscellaneous Manufactured Articles, Furnitures, Bedding, Mattresses Works of Art 43 Figure 1: Average relative price to Brazil, and the rest of the world. USA Korea 1356 commodities 99 commodities 0.05 1----------------- - -------- --------- --- ------------- ---- -------- --------- ---- ~~ 0. - ------------------------------- ------------------------------------ --- ------------------------- -- 0 4 - - --19 40 .0 5 - - - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 12 1_ 199193-- 4 1995 1996 0 -0.05 > - _ ~~>_~~~~ ~ ~~~~~~~~~~~~~~~~~~~~~~ ~ ~~ -0.05 199 _ 1992 1993 1994 19Q5 1996 -0.15 - -0 - - -- -- -0. 2 -0. -0.25 -… -- -0.2 -- - - - - - . 3 JAPAN ARGENTINA 580 commodides 686 commodities 0 .05 -- -- - -- - - - --- ---- - -- -- - - - - ---- - - - - - -- - - - - -- - - - - --- - - - -- - --- 0 .1 4 - --- - - - - --- - -- -- -- - - -- - --- -- - - - --- -- --- - -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --- 0.05-._ _ O .12 - - - ------- - - -0.14 -- ----- 1 99 - 1139 1994 1995 1996 0.12 -0.05 - ---. ------- - -- - -…-0------------------------ -0.05 - ---- ----- ----_--O-.04--------- - - ---- --- ----------------------0- -6 - -------------------I----------- - - - ---- ---- ---------------- ::;; -0.1-.-.--.-- - ---.~ - - --~~ - ~ ~ ~ ~ ~ ~ ~~00 - - - - - -0.02-,- - - - -0215 - -_-v 44 -O.O89 9-3- --~-- ---- -- --- 1994 1995 1996 0 . 2 - - - - - - - - - - - - - - 0 ~~~~~~~~~~~~~~0 4 0 - - - - - - - - - - - - - - - - - - - - - - - - - - - 7 7 - - - - - - - - - - - - - - - 44 Figure 2: Average absolute export prices to Brazil and to the non-MERCOSUR world. USA Korea 0.15 .- -----0.1 O0.1 -t0 5_ 9/ 0.05 - -A----- - - - - -- - - -- -------- - ° A 1990 199~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 1992 1993 1994 1995 1996 0 ''''.6 0.04 3 1991 1993 1994 1995 1996 1 -0.05 -- --- -------- ---- ----- --- ---- - -0.1 I; -0 . 1 ----- - -- - - - - - - - - --- - - - --- - --- - - -0 .1 5 -- - - -- - - -- - - -- - - -- - - -0.15 - - -- ---- - ----------- -- - --------- --- -- -- - - --------- ------- --------------- -0.2 --- ----- ------ -- ------ -------- - ------------ -- ----- --- + Brazil -C.World -+Brazil -<>World Ja pan Argo ntina 0 .3 - -- - - -- - - - - - - - -- - - -- -- - - - - 0 0 O .5 . - - - - - - - - - - - - - - , / O .0 3 .~.0 - -- - -- -- -- -- -- -- -- - -- -- -- -- - -- - - -- -- -- - 1990 1991 1992 1993 1994 1995 1996 -0.02 - - -- - ----- ---------------------------- - -- ------- - -- -- - --- - --.,,,+.-Brazil --World +Brazil -<>-World 45 Figure 3: Average relative price of Argentina/rest of the world (RoW), in the Brazilian market. World Germany 415 commodities 324 commodities 0.35 5 . - - --- - -- - - --. 0.25 - - - - - --- - - - --- 0.52 - - -----0.- 0- , - --- _ -\- 0.15 ----- - ----------- - - 0.2 ------- --------- - 0.1 - ---0 - --~ - A / - --- 0 1 - - -- - - -- -- - 0 - -- -- -- P- -- - - - - - - - - - - - - - - - - - -0 1 - - - - - - - -- --- -- --- 0.05 - - - - 2 -- - - ---- - -- - --- - -- -- -- - . .- -- 0 - ° * * -.1 1991 1992 1993 1994 1995 X -0.05 199 . 191 1992 1993 1994 95 1996 N. -0.15 1 - _0.32 - - - - - - V USA Japan 323 commodities 155 commodities 0-4 - .. 0 .4 ---- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 0~~~~~~~~~~~~~~~~~~~~. 0 .2 -.IV_ 199~~~~~~~~~~~~~~~~~~x ~~~~~~~~0.1 - -,----- ----- - - --------- 1 t 19 t-- + 1994 N1995 1996 0 - -0.2-------------- --- _____________________________ 0~~ . 2-.X -----°t- *-x- - - - - - - 1990 _9.21 1992 - - 1993 1994 ^ 1995 0 6 - 0 .-- - 1 - -. -0 .4 - - - - - - - - - - - - - - - - - - - - - - - --- - - - - - - - - - -- -4- - - - - - - - - - - - - - - - ---- - - - - - - ---- - - - - -0.6 - -0 .8 -- -- - -- - -- -- - -- 46 Figure 4: The effect of a PTA on member and non-member prices. P I ~~~~rf'(T*2) rf*(r 1) . ......... .... . ............ .......X N /'45 / /_ __ p A p p rf* r f* r*. Irf \450 L _ __ _ __ _ __450_ __ _ __ _ p p B C 47 Table 1: HS-6 tariff average (unweighted) for non-member, member and preference margin by sub-group and by year. HS-2^ YEAR M.F.N. PARTNER PREF. MARGIN4 01-15 1991 16.7 16.7 0.00 (55)' 1992 11.5 4.5 6.56 1993 8.3 2.1 6.08 1994 7.4 0.8 6.55 1995 7.6 0.0 7.63 1996 7.8 0.0 7.78 16-27 1991 28.7 28.7 0.00 (61)' 1992 22.5 8.8 11.86 1993 9.3 2.3 6.68 1994 8.3 0.9 7.30 1995 11.2 0.0 11.17 1996 11.5 0.0 11.53 28-38 1991 19.2 19.2 0.00 (340)' 1992 15.3 6.0 8.62 1993 11.8 3.0 8.55 1994 7.4 0.8 6.54 1995 8.0 0.0 8.00 1996 8.0 0.0 8.04 39-40 1991 26.4 26.4 0.00 (107)- 1992 22.4 8.7 12.40 1993 13.8 3.5 9.97 1994 12.2 1.3 10.69 1995 12.2 0.0 12.20 1996 12.1 0.0 12.05 41-63 1991 26.4 26.4 0.00 (141)^ 1992 20.6 8.0 11.37 1993 14.4 3.6 10.39 1994 13.1 1.4 11.50 1995 14.9 0.0 14.95 1996 14.2 0.0 14.17 64-83 1991 18.9 18.9 0.00 (150)' 1992 15.9 6.2 8.99 1993 11.4 2.9 8.24 1994 10.4 1.1 9.08 1995 12.2 0.0 12.21 1996 12.7 0.0 12.66 84-85 1991 30.8 30.8 0.00 (363)' 1992 26.1 10.2 14.33 1993 19.5 4.9 13.88 1994 19.3 2.1 16.76 1995 17.0 0.0 17.04 1996 17.2 0.0 17.17 86-92 1991 36.6 36.6 0.00 (110)* 1992 29.7 11.6 15.94 1993 20.9 5.2 14.80 1994 20.5 2.3 17.78 1995 16.4 0.0 16.42 1996 22.2 0.0 22.17 93-96 1991 48.3 48.3 0.00 (29)- 1992 40.6 15.8 20.98 1993 20.0 5.0 14.24 1994 17.8 2.0 15.50 1995 18.2 0.0 18.21 1996 19.9 0.0 19.93 l he parenthesis under the sub-group headmg is the number of commodities available. * The preference margin is calculated at the commodity level using {[(l+tmfn)/(l+tpartner)l-l l 100. 48 Table 2A: Estimation results of equation (5) over all commodities.** COUNTRY T SE I* SE w/e,Q, SE w*/e,*Q, SE R2 EDF CANADA 4.692 0.133 0.478 0.093 0.490 0.037 -0.239 0.039 0.399 1178 CHILE -0.242 0.096 0.601 0.065 -0.060 0.041 0,300 0.023 0.232 1138 CHINA -0.739 0.041 0.470 0.039 -0.344 0.022 0.631 0.038 0.403 1029 FRANCE -1.136 0.201 0.884 0.141 0.226 0.097 -0.147 0.064 0.032 2278 UK -0.680 0.152 0.417 0.093 0.245 0.041 0.084 0.033 0.075 2800 GERMANY. -0.570 0.111 0.338 0.063 -0.104 0.022 0.318 0.028 0.091 4076 ITALY -0.465 0.120 0.754 0.076 -0.151 0.020 0.361 0.027 0.058 3901 JAPAN -0.690 0.095 1.636 0.059 0.041 0.003 0.183 0.010 0.873 2836 KOREA -1.200 0.120 0.282 0.073 1.024 0.102 -0.299 0.065 0.299 1276 MEXICO -0.648 0.163 1.429 0.116 0.225 0.034 0.393 0.042 0.741 943 USA -0.822 0.129 0.636 0.094 -0.052 0.044 0.066 0.035 0.012 4699 WORLD -0.915 0.038 0.332 0.026 -0.019 0.012 -0.032 0.011 0.092 9049 Table 2B: Estimation results of equation (5) over aUl commodities with year time dummies.** COUNTRY X SE r* SE R2 EDF CANADA 0.968 0.226 1.195 0.149 0.195 1172 CHILE -0.876 0.213 1.073 0.139 0.275 1132 CHINA 0.482 0.116 0.087 0.140 0.203 1023 FRANCE -0.948 0.234 0.894 0.185 0.091 2272 UK -1.090 0.227 0.916 0.160 0.055 2794 GERMANY, -0.076 0.159 0.110 0.105 0.070 4072 ITALY -0.886 0.161 0.768 0.116 0.102 3895 JAPAN -0.776 0.178 1.455 0.128 0.198 2830 KOREA -0.765 0.169 0.525 0.118 0.051 1270 MEXICO -0.389 0.199 1.288 0.149 0.270 937 USA -0.446 0.110 0.329 0.093 0.025 4693 WORLD -0.558 0.079 0.092 0.057 0.031 9043 ** Estimates are in bold and standard errors SE are beside the estirnates; Data used is reported by Brazil therefore unit values are reported as c.i.f.; all variables represented above are in natural logs. The 'WORLD' represents the non-MERCOSUR world as an aggregate. .b Germany's data period runs from 1991-1996. All others 1989-1996. 49 Table 3A: Estimation results for equation (5), by 9 commodity groups.** HS-2 COUNTRY * SE e SE wle,Q1 SE w/le,*Q1 SE R2 EDF 01-15 fra -0.816 0.995 0.506 0.616 -0.274 0.250 -0.2i6 0.203 0.106 134 gbr -0.782 1.458 1.845 0.939 -0.585 0.265 0.682 0.227 0.137 84 ger4 -1.4i0 0.640 0.489 0.358 -0.789 0.127 0.851 0.137 0.290 162 ita -0.098 1.341 0.255 1.054 -0.433 0.035 0.775 0.167 0.605 100 usa -1.613 0.412 0.948 0.314 0.084 0.114 0.289 0.089 0.141 328 wid 0.078 0.328 -0.007 0.256 0.222 0.059 0.053 0.054 0.209 931 16-27 fra -2.835 0.661 1.952 0.488 0.905 0.425 -0.696 0.283 0.169 136 gbr -2.635 0.551 0.647 0.445 1.289 0.271 -0.867 0.267 0.223 140 ger. 0.547 0.195 -0.055 0.269 -1.983 0.460 1.695 0.332 0.219 187 ita -1.167 0.391 0.515 0.269 0.196 0.160 -0.099 0.150 0.035 253 usa -1.131 0.525 0.111 0.372 0.022 0.190 -0.026 0.166 0.066 289 wid -2.339 0.206 1.371 0.177 0.079 0.076 0.157 0.072 0.567 634 28-38 fra -1.605 0.461 1.024 0.368 0.403 0.219 -0.281 0.159 0.031 552 gbr -0.283 0.460 -0.235 0.367 0.339 0.120 -0.359 0.122 0.014 677 ger& -0.750 0.099 0.338 0.158 0.296 0.050 0.381 0.042 0.990 922 Ita 0.697 0.369 -0.983 0.252 0.494 0.067 -0.480 0.077 0.126 526 usa 0.043 0.259 0.347 0.162 0.099 0.086 0.160 0.068 0.123 905 wvd -0.834 0.203 0.362 0.174 0.290 0.023 -0.102 0.024 0.718 1394 39-40 fra 0.304 0.814 -0.692 0.519 0.527 0.377 -0.147 0.253 0.017 284 gbr -0.762 0.806 -0.112 0.578 0.840 0.281 -0.077 0.240 0.088 333 ger -2.036 0.680 0.826 0.462 0.075 0.490 0.645 0.402 0.087 408 Ita -1.142 0.727 1.902 0.491 -0.500 0.179 0.843 0.176 0.068 400 usa -1.363 0.626 0.519 0.382 0.250 0.203 0.038 0.156 0.023 497 wid -1.420 0.448 0.844 0.300 0.039 0.147 0.154 0.109 0.026 643 41-63 fra 2.431 0.893 -1.935 0.679 0.696 0.365 0.129 0.245 0.101 164 gbr 2.487 1.159 -0.470 0.847 -0.282 0.205 0.460 0.177 0.044 247 gera. 2.914 0.789 -3.729 0.468 -0.853 0.312 1.U60 0.274 0.849 338 ita 0.574 0.629 1.543 0.445 -1.052 0.112 1.576 0.113 0.488 429 usa -0.901 0.740 0.674 0.514 -0.225 0.150 0.338 0.125 0.028 521 wid 0.512 0.316 -0.987 0.212 -0.179 0.069 0.077 0.061 0.096 1378 64-83 fra -2.468 0.786 2.099 0.440 0.138 0.388 -0.104 0.241 0.102 269 gbr -2.765 0.772 1.973 0.530 -0.360 0.237 0.631 0.211 0.088 361 ger4 -1.996 0.516 1.208 0.168 -0.686 0.257 0.333 0.190 0.194 657 ita -3.090 0.084 1.013 0.164 0.622 0.102 -0.495 0.062 0.754 547 usa -2.222 0.203 0.516 0.286 0.076 0.142 -0.443 0.130 0.185 621 wld -2.326 0.348 1.549 0.223 -0.594 0.097 -0.043 0.087 0.950 1337 84-85 fra -0.154 0.697 0.868 0.369 0.041 0.285 -0.072 0.180 0.059 560 gbr 0.071 0.589 0.435 0.435 0.474 0.248 -0.238 0.194 0.066 729 ger4. -0.323 0.309 0.673 0.105 -1.268 0.149 1.235 0.108 0.459 1076 ita -0.060 0.374 0.814 0.252 -0.160 0.106 0.118 0.096 0.046 1219 usa -0.571 0.346 0.865 0.196 -0.198 0.097 -0.003 0.070 0.041 1135 wid 0.104 0.199 0.089 0.115 -0.182 0.074 0.092 0.052 0.004 1942 86-92 fra 0.522 0.532 1.032 0.539 -1.081 0.599 0.183 0.408 0.111 104 gbr -0.739 0.386 -0.014 0.406 0.180 0.382 0.043 0.294 0.033 152 ger4 -0.843 0.134 -0.759 0.076 8.097 0.275 -7.755 0.243 0.969 204 Ita -1.451 0.286 2.219 0.358 -1.016 0.155 1.458 0.121 0.466 260 usa -0.013 0.299 0.811 0.241 -0.803 0.228 0.481 0.167 0.071 224 wid -0.578 0.173 0.368 0.189 -0.543 0.158 0.259 0.108 0.076 452 93-96 fra -4.027 2.126 2.647 1.585 0.361 1.238 0.556 0.775 0.097 43 gbr -0.282 1.405 -0.690 1.167 0.186 1.088 1.818 0.830 0.320 45 ger4 2.085 0.612 -0.722 0.717 -1.221 1.105 1.355 0.834 0.176 90 ita 0.530 0.680 1.501 0.641 -0.264 0.051 -0.614 0.122 0.416 145 usa -0.966 0.515 0.503 0.429 -0.344 0.477 0.436 0.438 0.037 147 wid -0.470 0.416 -1.235 0.305 0.219 0.265 -0.744 0.214 0.690 306 "*Estimates are in bold and standard errors are beside the estimates; all variables are in natural logs. Countries represented are France (fra), Great Britain (gbr), Germany (ger), Italy (ita), USA (usa), and the non-MERCOSUR world as an aggregate (wid). 46Germany's data period runs from 1991-96. All others are from 1989-1996. 50 Table 3B: Estimation with Time Dummies by 9 commodity groups.** HS-2 COUNTRY T SE E SE R2 EDF 01-1S fra -1.277 1.002 0.537 0.639 0.135 128 gbr -2.316 1.677 2.166 1.057 0.189 78 ger* -1.866 0.581 0.504 0.378 0.199 158 ita 0.000 1.931 0.075 1.525 0.162 94 usa -1.372 0.478 0.840 0.367 0.242 322 wld 0.087 0.379 0.001 0.301 0.032 925 16-27 fra -0.831 1.399 0.943 1.070 0.265 130 gbr -2.762 0.510 2.008 0.459 0.315 134 ger4 1.326 0.473 -0.377 0.507 0.105 183 ita -0.198 0.451 0.173 0.325 0.128 247 usa 0.139 0.743 -0.686 0.529 0.097 283 wvd -1.721 0.439 1.052 0.321 0.074 628 28-38 fra -0.841 0.560 0.698 0.508 0.122 546 gbr -0.849 0.652 0.550 0.486 0.049 671 ger* -0.557 0.353 0.308 0.320 0.057 918 ita 0.034 0.679 -0.380 0.528 0.082 520 usa 0.176 0.394 0.124 0.335 0.030 899 wid -0.789 0.279 0.554 0.236 0.022 1388 39-40 fra 1.248 1.258 -1.027 0.958 0.120 278 gbr -0.173 1.249 -0.126 0.985 0.070 327 ger# -2.379 1.113 1.804 0.930 0.205 404 ita -0.773 1.108 0.889 0.816 0.158 394 usa -0.932 1.065 0.224 0.816 0.054 491 wid -1.021 0.686 0.508 0.553 0.054 637 41-63 fra 1.231 1.132 -1.469 0.804 0.175 158 gbr 3.179 1.301 -1.326 0.880 0.129 241 ger* 1.060 0.826 -1.655 0.529 0.302 334 ita -1.460 0.929 1.665 0.609 0.175 423 usa -0.776 0.896 0.492 0.621 0.037 515 wvd 0.968 0.300 -1.438 0.204 0.100 1372 64-83 fra -1.477 1.035 0.769 0.641 0.361 263 gbr -3.218 1.117 1.873 0.750 0.112 355 gera 0.461 0.769 -0.734 0.473 0.051 653 ita -2.616 0.495 1.713 0.387 0.229 541 usa -2.045 0.417 0.399 0.354 0.151 615 wvd -1.345 0.391 0.554 0.262 0.066 1331 84-865 fra -0.675 1.004 1.173 0.598 0.077 554 gbr -0.763 0.840 1.130 0.534 0.142 723 ger+ -0.409 0.440 0.816 0.261 0.066 1072 Ita -0.227 0.525 0.619 0.320 0.033 1213 usa -0.297 0.435 0.586 0.300 0.040 1129 wid 0.526 0.246 -0.394 0.167 0.020 1936 86-92 fra 1.245 0.749 0.630 0.589 0.644 98 gbr -1.417 0.482 -0.456 0.478 0.086 146 gera -0.040 0.599 -2.265 0.560 0.519 200 ita 0.124 0.423 0.017 0.356 0.669 244 usa -0.068 0.395 0.409 0.386 0.078 218 wid 0.303 0.251 -0.403 0.241 0.170 446 93-96 fra -4.085 2.327 3.440 1.276 0.709 37 gbr 2.877 1.710 -2.691 1.638 0.464 39 ger4 2.558 0.792 -0.542 1.066 0.190 86 Ita 1.265 0.584 0.198 0.872 0.452 139 usa -1.124 0.617 0.221 0.522 0.106 141 wvd 0.452 0.411 -1.824 0.311 0.339 300 * t Estimates are in bold and standard errors are beside the estimates; all variables are in natural logs. The countries represented are France (fra), Great Britain (gbr), Germany (ger), Italy (ita), USA (usa), and the non-MERCOSUR world as an aggregate (wid). 4 Gernany's data period runs from 1991-96. All others 1989-96. 51 Table 4A: Estimated coefficients of equation (6) over all commodities.** COUNTRY (years) X SE T* SE w/e,Q, SE w/e2Q2 SE w*le1*Q, SE R2 EDF CHILE (91-96) 1.353 0.10 0.127 0.08 0.828 0.13 -0.895 0.17 0.091 0.13 0.89 1042 GERMANY (91-96) 0.737 0.09 0.447 0.08 1.081 0.08 -1.280 0.17 -0.033 0.08 0.61 4959 JAPAN (89-96) 1.071 0.09 0.168 0.07 1.083 0.03 -1.055 0.05 0.015 0.02 0.72 2764 KOREA (89-96) 0.184 0.07 0.360 0.06 1.385 0.05 -0.073 0.12 -0.145 0.03 0.75 1372 USA (91-96) 0.883 0.08 0.445 0.08 0.779 0.16 -0.843 0.25 0.379 0.16 0.60 5463 Table 4B: Estimated coefficients of equation (6) over all commodities with time dummies.** COUNTRY T SE l* SE R2 EDF CHILE 1.126 0.13 0.711 0.12 0.84 1039 GERMANY 0.650 0.10 0.827 0.10 0.59 4956 JAPAN 1.029 0.11 0.370 0.09 0.70 2749 KOREA 0.373 0.13 0.838 0.11 0.64 1367 USA 0.881 0.10 0.495 0.09 0.58 5460 ** Estimates are in bold and standard errors SE are besides the estimates, all variables m in natural logs. The parthesis next to the country is the mnge of the data. The unit values used here are f.o.b. since we are using the exporters as reporters here. 52 Table 5A: Estimated coefficients of equation (6), by 9 commodity groups.** HS-2 T SE * SE w/eIQI SE wIe2Q2 SE wI/el*QI SE R2 EDF t 01-15 chl 1.384 0.22 -0.190 0.18 1.279 0.27 -1.261 0.37 -0.332 0.28 0.87 378 ger 0.231 0.33 * -0.633 0.21 1.706 0.10 -2.224 0.13 -0.926 0.11 0.94 184 jpn - - - - - - - - - - - 10 kor - - - - - - - - - - - 2 usa 0.090 0.50 * 0.127 0.42 1.743 0.46 -0.531 0.61 -0.703 0.49 0.75 279 16-27 chl 0.833 0.21 0.281 0.17 * 0.945 0.32 -1.912 0.45 -0.242 0.33 0.87 181 ger 0.749 0.53 1.248 0.44 * -1.518 0.75 2.819 1.38 2.027 0.67 0.46 160 jpn 0.033 0.40 * 0.996 0.30 * 1.456 0.16 -1.189 0.25 -0.061 0.12 0.85 66 kor 0.124 0.52 * 0.386 0.37 1.600 0.30 0.385 0.35 0.085 0.15 0.90 57 usa 0.545 0.30 0.830 0.31 4 0.121 0.85 -2.652 1.14 1.120 0.87 0.66 301 28-38 chl 3.826 0.53 -1.655 0.44 1.065 0.54 -2.902 0.78 -1.025 0.60 0.78 139 ger 0.316 0.19 * 0.283 0.17 4 0.800 0.14 -1.524 0.26 0.173 0.13 0.67 933 jpn 0.343 0.32 # 0.642 0.23 * 1.091 0.09 -1.157 0.14 -0.074 0.06 0.76 452 kor -0.224 0.78 -0.640 0.55 1.616 0.36 0.906 0.68 -0.173 0.21 0.57 86 usa 0.762 0.23 0.639 0.22 * 0.641 0.30 -0.425 0.43 0.422 0.32 0.62 1300 39-40 chl -0.655 2.03 2.636 1.17 4 -1.704 2.01 1.401 1.81 1.845 1.71 0.56 49 ger 0.916 0.40 0.246 0.33 0.813 0.33 -0.224 0.81 0.533 0.32 0.69 422 jpn 0.889 0.45 0.544 0.32 4 1.359 0.15 -1.338 0.25 0.109 0.10 0.69 270 kor 0.432 0.60 0.120 0.44 2.147 0.38 -1.015 0.55 -0.780 0.28 0.65 142 usa 0.354 0.39 * 0.118 0.35 2.065 0.63 -1.908 0.66 -0.610 0.62 0.76 475 41-63 chli 2.566 0.60 0.200 0.41 0.677 0.43 -0.400 0.60 0.480 0.45 0.72 152 ger 0.423 0.47 1.159 0.37 4 0.840 0.25 0.458 0.52 0.140 0.23 0.55 348 jpn 3.546 0.56 -1.084 0.40 0.459 0.15 -0.538 0.25 0.396 0.12 0.63 150 kor -0.710 0.37 ' 1.245 0.23 4 0.904 0.23 0.098 0.35 0.245 0.12 0.65 385 usa 0.757 0.42 0.607 0.32 * 0.853 0.43 0.224 0.76 0.457 0.45 0.58 633 64-83 chl 1.311 0.42 0.775 0.28 4 0.024 0.61 -0.541 0.86 0.702 0.63 0.89 78 ger 0.604 0.38 0.717 0.33 4 1.280 0.22 -2.196 0.56 -0.349 0.21 0.57 937 jpn 0.612 0.39 0.495 0.28 4 1.033 0.13 -0.899 0.19 0.036 0.09 0.54 471 kor 2.810 0.91 -0.334 0.65 1.749 0.40 -2.631 0.73 -0.432 0.24 0.55 147 usa 1.372 0.52 -0.460 0.48 0.877 0.76 -0.180 1.03 0.348 0.77 0.34 637 84-85 chl 0.900 1.88 -0.238 1.16 1.449 2.64 1.415 3.37 0.486 2.65 0.51 22 ger 0.909 0.23 0.669 0.20 * 1.122 0.19 -0.804 0.51 0.076 0.19 0.62 1579 jpn 1.148 0.22 -0.319 0.15 1.274 0.07 -1.442 0.13 -0.121 0.05 0.58 1044 kor 0.570 0.19 * 0.347 0.19 * 1.420 0.15 -0.345 0.31 -0.174 0.08 0.70 312 usa 1.177 0.38 0.629 0.37 4 0.297 0.82 -1.944 0.91 0.725 0.81 0.29 1464 86-92 chl - - - - - - - 3 ger 1.789 0.43 -0.450 0.46 2.681 0.60 -4.461 1.54 -1.587 0.60 0.54 269 Jpn 1.362 0.32 0.559 0.20 * 0.986 0.13 -0.368 0.24 0.095 0.09 0.71 206 Kor 0.955 0.14 0.606 0.40 0.346 0.52 -1.526 0.69 -0.153 0.32 0.64 90 Usa 0.739 0.40 0.077 0.47 1.969 1.87 1.275 2.45 -0.388 1.85 0.48 183 93-96 chli - - - - - - - - - - - I Ger 0.669 0.77 -1.572 0.73 6.498 1.52 -13.371 3.00 -5.725 1.35 0.45 87 Jpn 2.515 0.51 -0.437 0.52 0.369 0.42 0.653 0.82 0.291 0.28 0.52 45 Kor -0.078 0.44 * 0.094 0.36 1.244 0.43 0.844 1.43 -0.015 0.22 0.25 111 Usa 0.792 0.80 -1.126 1.06 6.903 3.97 1.749 3.75 -5.176 3.93 0.30 151 * Estimates are in bold and standard errors SE are besides the estimates; all variables listed above are in natural logs. To the right of the SE we have indicated 4 if the estimate is less than one with 95% confidence, and 4 if the estimate on the rival's tariff are greater than zero at the same level of confidence. t Missing values are assigned only to those estimates with very small error degrees of freedom (EDF) as shown. 53 Table 5B: Estimation with time dummies by 9 commodity groups.** HS-2 COUNTRY T SE T* SE R2 EDF t 01-15 chl 0.442 0.35 . 1.486 0.37 * 0.84 375 ger 0.418 0.34 # -0.812 0.28 0.60 181 jpn - - - - - 5 kor - - - - - 0 usa 0.565 0.67 -0.159 0.56 0.64 276 16-27 chl 1.049 0.25 0.274 0.23 0.83 178 ger 2.044 0.72 0.227 0.62 0.49 157 jpn -0.284 0.53 # 1.298 0.43 + 0.84 61 kor 0.074 0.67 1.068 0.54 + 0.78 52 usa 0.953 0.45 0.740 0.41 4 0.68 298 28-38 chl 3.989 0.58 -1.579 0.56 0.73 136 ger 0.358 0.22 t 0.392 0.19 # 0.65 930 jpn 0.019 0.42 # 1.139 0.34 * 0.73 447 kor -0.660 1.12 0.376 0.81 0.42 81 usa 0.858 0.24 0.597 0.22 # 0.62 1297 39-40 chl -0.778 2.21 2.669 1.59 # 0.59 46 ger 0.800 0.52 0.959 0.55 * 0.67 419 jpn 0.903 0.60 0.660 0.52 0.68 265 kor 0.503 0.92 0.920 0.70 0.61 137 usa 1.302 0.54 -0.743 0.55 0.76 472 41-63 chl 1.477 0.66 1.392 0.50 * 0.61 149 ger -0.255 0.50 . 1.899 0.39 # 0.55 345 jpn 2.773 0.78 -0.474 0.58 0.60 145 kor -0.352 0.91 1.516 0.60 * 0.58 380 usa 0.288 0.51 0.965 0.38 * 0.57 630 64-83 chl 0.726 0.71 1.524 0.60 # 0.72 75 ger 0.146 0.41 . 2.110 0.39 * 0.57 934 jpn 0.756 0.57 . 0.564 0.49 0.54 466 kor 2.843 1.31 0.287 0.89 0.47 142 usa 1.560 0.64 -0.530 0.60 0.32 634 84-85 chl 1.093 1.93 1.083 1.69 0.49 19 ger 0.919 0.27 0.968 0.32 # 0.60 1576 jpn 0.908 0.29 0.238 0.26 0.58 1039 kor 0.391 0.30 * 1.498 0.24 4 0.64 307 usa 0.915 0.43 1.107 0.45 * 0.29 1461 86-92 chl - - - - - I ger 2.070 0.45 0.488 0.59 0.54 266 jpn 1.556 0.31 0.309 0.27 0.66 201 kor 0.912 0.24 0.354 0.33 0.68 85 usa 0.445 0.46 0.242 0.58 0.49 180 93-96 chl - - - - - 0 ger 1.099 0.89 -1.471 0.96 0.39 84 jpn 3.305 0.99 -0.964 0.79 0.56 40 kor -0.334 0.52 . 0.424 0.45 0.26 106 usa -0.420 1.04 -0.880 1.31 0.34 148 ** Estimates are in bold and standard errors SE are besides the estimates; all variables listed above are in natural logs. To the right of the SE we have indicated # if the estimate is less than one with 95% confidence, and 4 if the estimate on the rival's tariffs are greater than zero at the same level of confidence. t Missing values are assigned only to those estimates with very small error degrees of freedom (EDF) as shown. 54 Table 6A: Estimation results of equation (7).** HS-2 T SE w*T*/e1*Q1 SE (w*le2*)1Q2 SE (wiel)1Q SE R2 EDF 01-15 0.378 0.230 0.887 0.444 2.926 0.642 -0.265 0.466 0.794 327 16-27 4.028 0.404 3.383 0.828 -2.215 1.353 -2.794 0.886 0.486 183 28-38 -0.581 0.336 1.883 0.507 -0.180 0.693 -0.876 0.530 0.747 495 39-40 -0.581 0.635 4.575 0.347 -0.721 0.474 -3.852 0.373 0.938 239 41-63 1.905 0.622 OA76 1.032 5.345 1.430 0.076 1.117 0.466 307 64-83 -0.745 0.503 1.363 0.871 -1.170 1.081 -0.347 0.947 0.653 275 84-85 0.226 0.552 -1.240 0.979 -0.078 1.453 1.864 1.105 0.083 583 86-92 0.413 0.270 0.043 1.342 -4.064 2.792 1.550 1.542 0.427 183 93-96 0.124 0.913 -5.429 3.202 -7.324 6.408 8.615 3.636 0.425 59 ALL 0.245 0.086 0.202 0.173 0.808 0.287 0.671 0.185 0.689 2691 * * The estimates are in bold and standard errors are besides the estimates. All variables are in natural logs. The member tariff factor has been rolled into the real exchange rate variable due to lack of time series in Argentina data. The unit values used here are in f.o.b. since we are using the exporter as the reporter. 55 Table 6B: Estimation with Time Dummies.** HS-2 _ SE w*¶*/e.Q, SE R2 EDF 01-15 0.383 0.224 -0.226 0.073 0.114 326 16-27 2.199 0.334 -0.396 0.075 0.797 182 28-38 -0.201 0.255 0.200 0.065 0.139 494 39-40 -0.509 0.840 0.215 0.126 0.107 238 41-63 1.811 0.568 -0.237 0.171 0.113 306 64-83 -1.305 0.434 0.173 0.122 0.104 274 84-85 0.137 0.528 -0.486 0.141 0.101 582 86-92 0.443 0.274 0.476 0.271 0.061 182 93-96 0.172 0.862 0.957 0.725 0.109 58 ALL 0.188 0.083 -0.055 0.030 0.025 2690 ** The estimnates are in bold and standard errors are besides the estimates. All variables are in natural logs. The unit values used here are in f.o.b. since we are using the exporters as the reporter. 56 Table 7: Total 1991 Exports to Brazil Terms of Trade Losses ($ million).** COUNTRY EXPORTS TOTAL EXPORT REVENUE LOSSES 4A 5A 4A* 5A* CHILE 524.4 -17.3 -25.7 -40.4 -51.2 GERMANY 2,030.0 -236.0 -198.8 -169.4 -165.2 JAPAN 1,349.6 -58.8 -13.1 -70.6 -20.8 KOREA 146.7 -13.7 -19.1 1.2 -8.3 USA 5,395.5 -624.1 -690.5 -545.3 -556.8 SUM 9,446.2 -950.0 -947.2 -824.4 -802.3 ** Revenue losses were calculated using the elasticities of the rival's tariffs from Table 4A and 5A. 4A* and 5A* also incorporates the own tariff effects due to MEN reductions. 57 Policy Research Working Paper Series Contact Title Author Date for paper WPS2136 An Empirical Analysis of Competition, Scott J. Wallsten June 1999 P. Sintim-Aboagye Privatization, and Regulation in 38526 Telecommunications Markets in Africa and Latin America WPS2137 Globalization and National Andres Solimano June 1999 D. Cortijo Development at the End of the 84005 20h Century: Tensions and Challenges WPS2138 Multilateral Disciplines for Bernard Hoekman June 1999 L. Tabada Investment-Related Policies Kamal Saggi 36896 WPS2139 Small States, Small Problems? William Easterly June 1999 K. Labrie Aart Kraay 31001 WPS2140 Gender Bias in China, the Republic Monica Das Gupta June 1999 M. Das Gupta Of Korea. and India 1920-90: Li Shuzhuo 31983 Effects of War, Famine, and Fertility Decline WPS2141 Capital Flows, Macroeconomic Oya Celasun July 1999 L. Nathaniel Management, and the Financial Cevdet Denizer 89569 System: Turkey, 1989-97 Dong He WPS2142 Adjusting to Trade Policy Reform Steven J. Matusz July 1999 L. Tabada David Tarr 36896 WPS2143 Bank-Based and Market-Based Asli Demirguc-Kunt July 1999 K. Labrie Financial Systems: Cross-Country Ross Levine 31001 Comparisons WPS2144 Aid Dependence Reconsidered Jean-Paul Azam July 1999 H. Sladovich Shantayanan Devarajan 37698 Stephen A. O'Connell WPS2145 Assessing the Impact oi Micro-credit Hassan Zaman July 1999 B. Mekuria on Poverty and Vulnerability in 82756 Bangladesh WPS2146 A New Database on Financial Thorsten Beck July 1999 K. Labrie Development and Structure As11 Demirguc-Kunt 31001 Ross Levine WPS2147 Developing Country Goals and Constantine Michalopoulos July 1999 L. Tabada Strategies for the Millennium Round 36896 WPS2148 Social Capital, Household Welfare, Christiaan Grootaert July 1999 G. Ochieng And Poverty in Indonesia 31123 Policy Research Working Paper Series Contact Title Author Date for paper WPS2149 Income Gains to the Poor from Jyotsna Jalan July 1999 P, Sader Workfare: Estimates for Argentina's Martin Ravallion 33902 Trabajar Program WPS2150 Who Wants to Redistribute? Russia's Martin Ravallion July 1999 P. Sader Tunnel Effect in the 1990s Michael Lokshin 33902 WPS2151 A Few Things Transport Regulators Ian Alexander July 1999 G. Chenet- Smith Should Know about Risk and the Cost Antonio Estache 36370 Of Capital Adele Oliveri WPS2152 Comparing the Performance of Public Antonio Estache July 1999 C-. Chenet Smith and Private Water Companies in the Martin A. Rossi 36370 Asia and Pacific Region: What a Stochastic Costs Frontier Shows WPS2153 The Mystery of the Vanishing Martin Ravallion July 1999 P. Sader Benefits: Ms. Speedy Analyst's 33902 Introduction to Evaluation WPS2154 Inter-Industry Labor Mobility in Howard Pack August 1999 H. Sladovi,h Taiwan. China Christina Paxson 37698 WPS2155 Lending Booms. Reserves, and the Barry Eichengreen August 1999 S. Kpundel Sustainability of Short-Term Debt: Ashoka Mody 39591 Inferences from the Pricing of Syndicated Bank Loans WPS2156 How Has Regionalism in the 1990s Isidro Soloaga August 1999 L. Tabada Affected Trade? L. Alan Winters 36896