POLICY RESEARCH WORKING PAPER 143-8 Trade Reform, Efficiency, and a fiunctioningrr arkcets-and a and Growth diersirfied production structure benefit more than' other countries from the Ejaz Gbani productivity gains through Carl Jayarajab :- - tradereFoml. The World Bank Eastern Africa Department Country Operations Division and Operations Evaluation Departmeent Country Policy, Industry, and Finance Division March 1995 POLIcY RESEARCH WORKING PAPER 1438 Summary findings The main objective of trade reform is to make markets the level of diversification in production at the timc more competitive and, by introducing competition reform is begun. among previously protected domestic firms, to change Their findings confirm the link between trade reform the behavior and performance of firms. Efficiency gains and efficiency gains. Reduced average tariffs and are achieved through increased productivity - more quantitative restrictions on imports arc associated with efficient use of resources - and a shift in resources from increased output growth for a given level of investment inefficient to efficient sectors. As a result of incrcased and capacity use. But the extent to which trade reform efficicncy, output grows. helps a country reflects the initial conditions prevailing But the transition from a restrictivc to an opcn tradc in the country. regime can impose short-term adjustment costs for Ghana, Indonesia, and Turkey began their trade industries newly exposed to external competition. This reform programs under different conditions. Indonesia can be compounded by efforts to restore macroeconomic and Turkey had a more diversified production structure stabilization, such as reductions in fiscal deficits that and a better functioning market than Ghana. All three could hurt the country's infrastructure. countries carried out intensive trade reform, but Ghani and jayarajah examine the impact of trade Indonesia and Turkey benefited more than Ghana did. reform on productivity and GDP growth, export groo th, In short, countries with well-functioning markets and a the diversification of exports, and the trade balance. better human rcsource base bencfit more from They alsc examine whether trade reform affects different productivity gains resulting from trade reform than reforming countries diffcrently - whether its outcome is countries with less well-functioning markets do. related to such factors as the functioning of markets or This paper - a joint product of the Country Operations Division, Eastern Africa Dcpartment, and the Country Policy, Industry, and Finance Division, Operations Evaluation Department- is part of a largereffort in the Bank to identify sustainable conditions for growth. Copies of the paperare available free fromtheWorldBank, 1818 H StrectNW, Washington, DC 20433. Pleasecontact Afsar Nokhostin, room J1O-285, extension 34150 (17 pages). March 1995. Ihc Poey Reearch Workig Paper Series dissenates the findings of work un progress to cowuage the exchange of ideas about deelopment ssues. An obyectve of the series tszo get thefings out quckly, even if the presentations me less than flly polished. The papers camy theu names of the asthors and sh be used and cited accordingly. The fndings, interpraaions, and condssions are the authors'Gum and shosdd not be attributed to vt; World Bank, its Executive Board of Directors. or any of its member countrib Produced by the Policy Rcscarch Dissemination Centcr Trade Policy Reform, Efficiency, and Growth Ejaz Ghani and Carl Jayarajah The World Bank " The authors are grateful to several people for their comments including late Bela Belassa, William Branson, Kazi Matin, John Nash, late Alexander Nowicki, Lawrence Summers, and Sulaiman Wasty. The authors are responsible for any remaining errors. Sanliang Yue provided research support. Trade PoRicy Reform, Effriency, and Growth A. Introduction 1. The central objective of trade policy reform is to make markets more competitve and, thereby, change the behavior and performance of firms, i.e., to introduce greater extemal competition into the previously protected domestic markets and increase economic efficiency at the level of individinld firms.y The efficiency gains are achieved both through increased productivity-more efficient use of existing resources in response to increased competition-and a shift in resources from inefficient to efficient sectors, i.e., gains from freer trade. Improved efficiency, in turn, contributes to increased output growth. The transition from a restrictive to an open trade regime, however, can impose short- run adjustment costs in industries newly exposed to external competition. This may be further compounded by efforts to restore macroeconomic stabilization, such as across-the-board reductions in fiscal deficits that could adverse!y affect the physical and hluman inEastructure of the country. This paper empirically examines the impact of trade reform programs for a group of trade refoming countries. In particular, it examines the impact of trade reform on productivity and GDP growth, export growth, the diversification of exports, and the trade balance. This paper also examines if the trade reforms had dissimilar effects across the reforming countries, i.e., whether the outcome of trade reform was related to the initial conditions such as the existing level of diversification of the production structure or the functioning of markets. 2. This paper confirms the link between trade policy reform and efficiency gains. Reductions in average tariff levels on imports are associated with increased output growth, for a given level of investment and capacity utilization. However, the extet to which trade policy reform has a beneficial impact is related to the initial conditions prevailing in the country. Ghana, Indonesia, and Turkey started their trade reform programs with different initial conditions. Indonesia and Turkey had a more diversified production strucutre and a relatively well fimctioning market compared to Ghana. Both groups carried out intensive trade reform- The beneficial impact of trade reform was more significant in Indonesia and Turkey compared to Ghana. This confirms that counties with well functioning markets benefit more from productivity-enhancing trade reforms. I' See R. Harris, "Market Structure and Trade Liberalization: A General Equilibrium Assessment,' in T.N- Srinvasan and J. Whalley, eds., General Equilibrium Trade Policy Modelling (Cambridge, Mass.: MIT Press, 1986); W.J. Baumol, J.C. Panzar and RD. Willig, Contestable Markets and the Theory of Industry Stm cture (California: Harcourt Brace Jovanovich, 1988); J. Vickers and G. Yarrow, Economic Perspectives on Privatization (Oxford University, 1990). - 2 - B. Trade Reform, Productivity and Output Growth 3. The two key variables that can have a major influence on growth include the availability of resources and the government policies that affect the efficiency of resource use. Both variables can determine the level and the rate of growth of output.2 In general, the level of investment declined or remained stagnant in the trade adjusting countries in the 1980s during the adjustment period; this pause in investment activity is well recognized by now and is known to be related to uncertainty arising from changes in policiesY Any output growth that took place, therefore, had to come from efficiency improvements. There are several channels through which trade reform can improve (finn-level) efficiency and, thus, oetput gro .. ih. First, import competition can force domestic finms to increase productivity 4 Domestic firms can raise productivity by increasmg scale-efficiency (producing at minimm efficient scale) and by improving technical efficiency (ahieuving the maximum possible output from a given bundle of inputs). In effect, import competition changes the structure of the market and, thus, its performance. Second, trade liberalization can also generate beneficial externalities. If exteal economies, such as technology diffusion, can be transmitted world-wide and are not nation specific, then trade liberalization could benefit technological developments in the adjusting countriesY. The following sections evaluate the impact of trade liberaization on productivity and GDP growth for nine trade adjusfing countries. 4. Productivity Gains. Ideally, data on factor productivity are required at the finm level to assess the linhge between trade reform and productivity;- they are, however, not available for the group of adjusters reviewed here. This paper uses data on labor productivity at the aggrege level. Table I reports the annual data on real output per worker in the m in sector for five trade adjusters for the period 1978-86 (Colombia, Indonesia, Mexico, Pakistan and Turkey).2' For the five adjusters as a group annual output per worker increased from 103 in 1978-82 to 130 in 198346 (1980= 100). a' For a survey, see W.R. Easterly and D.L. Wetzel, Policy Detenninants of Growth: Survey of Theorv and Evidence, PRE WP 343, The World Bank, December '989; R.J. Barro, A cross- country studv of rowth. savings. and government, NBER Working Paper No. 2855 (Cambridge, MA, 1989). 3 See W.H. Branson and S. Schwartz, 'Investment Efficiency and the Financial Sector," August 1989, prepared for the Rerort on Ad& ent Lending II. 4 The efficiency resource costs associated with restrictive trade policies, of course, would be less when trade restrictions are not binding, e.g., smuggling can continue to provide competition to local firms despite import restrictions. 5 For an altemative argument, see F. Stewart and E. Ghani, "Do Externalities Mater for Development?," World Development. Pergamon Press, May 1991. ! For example, see J. Tybout, J. de Melo and V. Corbo, The Effects of Trade Reforms on Scale and Technical Efficiency: New evidence from Chile, PRE Working Paper No. 481, 1990. 71 Time series data on capital productivity for the tade adjusting counties are not available. Table 1: LABOR PIROJCflVWTY IN THU t4ANUFACIURINC SECTOR Clsl - 100) Red Ouq AV-ag per Worker 197I3 199 190 293 KW 23 1914 29,5 1936 1972 190-3 Cokmt 94.9 INA M0.0 105.8 101.3 17.4 114.5 22I1 1373 101.4 221.3 bScaiz 36. 961 200.0 113.1 117.4 1287 133.0 1413 15.3 QL.6 139.7 Makdw 99.3 IILI 100.0 99A 97.1 0.0 1113 112A 106.7 9.7 10.3 PAktn 7L4 IL2 2100.0 I2L4 12Z.5 1360 1393 144. 152.6 99.1 1432 T- qy 112.0 95.6 100.0 119.2 26.0 127.3 131.9 133L9 257.9 110.6 139.1 Avcea 94.2 97.0 10.0 209.2 113.0 IZ16 126.0 133.1 142.1 2aL7 130A to 1991 is xueatnm period ah 198336 is Saj ped. sca: World Table. Wold Bl 5. The issue is whether the improvement in labor productivity can be linked to trade liberalization, iLe., whether output per worker increased in response to reductions in the level of import protection. This is verified by examining if labor productivity is systematically rented to the level of import protection. Import protection is measured by the average tariff rate (ratio of total import duties collected to the CIF value of import) since most trade adjusting countries had managed to reduce QRs on imports'i Figure I plots the relationship between labor productivity and the average tariff rate based on a simple regression equation for labor productivity where the tariff rate appears as the explanatory variable. It shows a negative relationship between the tariff rate and labor productivity, i.e., an increase in labor productivity is associated with a reduction in the tariff rate. When the sample is restricted to include only the intensive adjusters (Idonesia, Mexico and Turkey),!' the explanatory power of the equation increases and the tariff rate becomes more signifcant (equation 2 under Figure 1). The evidence confirms the linkage between miceased iLmport competition and improvements in productivity. The following sections examine the linkage between import liberalization and the GDP growth rate. 6. Output Growth. Table 2 shows the data on annual real GDP growth rate for the nine trade adjusters. For the adjusters as a group, the average anmual real GDP growth rate increased from 3.6% in 1978-82 to 4% in 1986-89. The trade adjustment program, therefore, is associated with improved growth performance. The table also compares the average GDP growth rate for the intensive and less intensive trade adjusters. The intensive adjusters increased their annual real GDP growth rate from 3.8% in 197842 to 4.7% in 1986-89; the growth rate for the less intensive adjusters did not change significantly. The intensive adjusters, thus, benefitted more from the adjustment programs than did the less intensive adjusters. There were, however, significant differences within the group. While Indonesia's economy expanded more rapidly compared to the other intensive adjusters, Cote d'Ivoire's economy contracted amongst the less intensive trade adjusters. u See Thomas, V, and J. Nash, Best Practices in Trade Policy Reform, OUP, and Trade Policy Reforms Under Adjustment Programs, OED, 1991. 9' The intensive trade adjusters reduced QRs as well as the average tariff levels on imports; the less intensive adjusters are classified as those which managed to reduce QRs only but not the tariff levels. Figure 1: Import Competition -and Labor Productivity Labor productivity 140 130 _- - _.. 120 _ -. 1 10 *-- -- - - - - 100 _.- 9 0 . -I 12.5 is 17.5 20 22.5 Import tariff rate Equation: Labor productivity - aO + al A&tr Notes: 1. Figure 5.1 is based on equation 1. 2. Equation 1: Five adjustars (Colombia, Izdoousia, Mxico, Pakistan and Tureky) Labor productivity - 155.3 - 2.6ATZ (6.7) (-L.7) R-uquared 0.18, DW: 0.44 3. Equation 2: Intensive adjusters (Tndonasia, Mexico and Turkey) Labor productivity - 138.3 - 2.3AAR (17.4) (-5.6) R-squared 0.48, DW: 0.78 3. Sample ranga: 1978-1986. a-u-l data 4. Labor productivity is given by real outpur. per worker In the Manufacturing sector (1980-100) and A3Z is import duties collectad as a percentage of value of CIF imports. 5. Data used are the average for the group of adjusters. Sources: World Tables, World Bank; IFS and CEM. T&a1e 2. REAL GDP OROWTH RATES (in 19 US DoOma) Rcal GDP Anal A%Vea Growih rae 1978 1979 19RD 1931 1912 1983 1954 1931 1936 19817 19 1919 1971W2 19W5 1911689 iA) Imemive rade adimut (P'= . 9.3 -1.7 0.5 -2.9 -6. 42 S.5 4.5 4.3 5.3 6.3 -02 3.0 5.4 ldna * 7.7 6.2 7.9 7.4 -0A 3.3 6.1 2.5 4.0 3.4 16.0 7.4 5.3 3.9 7.7 mcxkm 8.2 9.3 8.4 H.8 -0.6 -42 3.6 2.6 3.8 1.5 16 2.9 6.8 0.7 05 Mro * 2.9 4.5 9.1 -2.E 9.3 4.3 43 63 84 -2.6 10.4 3.5 -4.6 35 j.j 1uek1' * 3.3 409 40.7 4.2 49 3.8 5.9 49 3o 7.2 3 1.2 2.1 4.9 5.1 A 'aa 6.4 3.5 5.0 2.9 1.3 4.3 5.7 4.2 4.3 2.3 7.5 4.2 3.S 3.2 4.7 (l) Lsa Ineemiwke adk miudens Cdauhbk - 8.4 5.4 4.1 2.1 1.0 1.6 3.5 3.3 6.1 5.3 3.7 3.3 4.2 2.3 4.6 CIte d1vobIe 13.9 3. 40.8 4.4 1.5 -13 -4.4 93 16 -2.9 4.4 -1.1 4. 1.2 -1.7 boni a 0.4 -2.1 .6D 2.2 -103 15.1 -1.1 -50 23 5.5 0.3 4.5 -3.1 3.0 3.2 Pakian aW 8.1 3.7 10.4 7.9 6.5 6.3 5.e 7.5 5.6 6.5 7.4 5.6 7.3 6.4 6.3 AVece 7.7 2.5 1.9 4.2 -03 5.5 0.3s 3.1 4.4 3.6 1.3 3.1 32 3.4 3.1 (C) All adiunrm Are ' 7.0 3.0 3.6 3.5 0.6 2.3 3.5 4.0 4.4 3.1 4.7 3.7 3.6 3.3 4. e * indt, kighe real a_anu GDP gawib rave iLrcued. - 197832 a p-w pem 19W3-5 is adjusneo and 199 is piat-elte piod. v Fern ycawL YS Sarc CE± Y 9n-PAIC (FYIII4. e Woald Bak Dabase (BESTh Nalkid Aeam). 7. The link between GDP growth and the trade policy reform is verified by estimating standard output growth equations for the two groups of trade adjusters (intensive and less intensive) using pooled data. The dependent variable in the equation is the growth rate in real GDP, and the explanatory vaiables are the average import tariff rate, the real exchange rate, the investmenttGDP ratio and a measure of capacity utilization. It is expected that tariff liberalization, for a given level of investment and capacity utilization, would increase output by inTroving the efficieacy of reso..-c.. use. The average import tariff rate has several advantages as an indicator of trade liberalization. First, it is the major variable that influenced the level of import protection. Second, the average tariff rate is a direct measure of trade policy unlike the measures based on trade flows (e.g., export/GDP ratio). Indicators of trade policy based on trade flows can be misleading since trade intensity can vary across countries for reasons unrelated to the trade policy, e.g., geographical location of the country.1' Second, the tariff rate does w A comparison of export to GDP ratios for Mexico and Jamaica shows that Jamaica is less trade restrictive than Mexico, contrary to the trade reform experience. In general, in cross-country comparisons based on the tradetGDP ratios, small economies turn out to be more open compared to large economies. - 6 - not suffer from the subjective nature of the indices constructed nn trade policy distortions.l'l Some caution, however, needs to be exercised in interpreting the changes over time in the tariff rate as measured here. An elimination of duty exemption or a change in the composition of imports can increase the average tariff rate independent of any changes in the statutory tariff rate. 8. Thie regression results for output growth are reported in Table 3. The estimate for the group of intensive adjusters shows that the coefficients on the exchange rate and the import tariff rate are significant but the investment to GDP ratio is not. A devaluation of the currency (a fail in the index, 1980=100) has a positive effect on the GDP growth rate. A reduction in the import tariff rate is associated with increased GDP growth rate. When a measure of capacity utilization is included, the coefficient turns out to be significant and positive. Increased capacity utilization, therefore, also played an important roie in facilitating growth, at least in the short-run. The evidence, thus, suggests that a reduction in the import tariff rate, for a given level of investmnent, is associated with an increased GDP growth rate. This finding is consistent with the data on changes in labor productivity for the trade adjusting countries. 9. In the case of less intensive trade adjusters, the coefficient on the tariff rate, although it has the right sign, is not significant. The coefficient on capacity utilization is highly significant. To test whether the regression results merely reflect a trend, a tine trend is included in the regression; the coefficient on the time trend turned out to be insignificant. The low explanatory power of the equation, however, suggests that other variables, not included in the regression, may have also contributed to output growth. 10. The estimated output equation for the adjusters as a group constrains the coefficient on the tariff variable to be the same across countries. Trade reform, however, may not have a similar effect across countries. The output growth equations were, therefore, re-estimated individually for the trade adjusters. The regression results are reported in Table 4 They highlight the differential impact of trade reform on output growth. A cross-country comparison of the estimated coefficients on the tariff rate shows that it has a significant effect on output growth in Mexico and Turkey but not in Ghana. Similarly, a devaluation had dissimilar effects on output. The estimated coefficient on the real exchange rate for Turkey is almost twice that of Ghana. The findings here support the view that the impact of trade policy reform is also dependent on the economic structure of the adjusting country.' Countries with better human infrastructures (i.e., a higher level of education and literacy) and more diversified production tend to benefit more from productivity-enhancing trade reform. More micro studies, however, would be necessary to establish this proposition more rigorously. ' See L. Pritchett, Measuring Outward Orientation in Developiig Countries: Can it Be Done?, PRE WPS 566, January 1991, for a detailed discussion on this subject; see also E. Leamer, "Mcasures of Openness," in R. Baldwin, ed., Trade Policv Issues and Empirical Analysis (Chicago: University of Chicago, 1988). Iv For r=ent empirical findings that support a structuUist view of adjustment, see R. Faini and J. de Melo, Adjustment. Investment, and theReal Exchange Rate in Developing Countries, PRE Working Paper No. 473. The World Bank, August 1990. See also H. Chenery, "The Structuralist Approach to Development Policy," American Economic Review, 1975; and H. Chenery, S. Robinson and M. Syrqmin, Industrialization and Growth, OUP, 1986. - 7- Table 3: TRADE UDERALIZATION AND GROW1H: 1910-19g (Dekd vwrbblk: ra GDP groth) 11 Capacity PAM RER ATR(t-l) GDP'Ot-I) Wtilinti T C of abo. R2 S.E.R D.W F-assstsd (A) Insrsve iade adiusten -D.03 -0.07 0.05 0.07 45 0.15 0.02 1.93 3.6- (LOD), (2.9?)' (1.22) (.64) (Ghan. Indies-. Mexico. Macco sa Turkey) -0.2 4.06 0.04 (3L1 0.06 45 °2 0-c2 1.56 4.13 (1.49) (2.89? (1.06) (1.71) (2.07) -0.04 -(.LU 0.05 40.1 0.10 45 0.15 o.m 195 2.99 Q.22T) 0.M)4) (1.27) (0.92) (55) (13) les intemmwe adijsern -0.04 -010 -0.38 0.12 36 0.9 o.m 2.27 2.19 (0.71) (1.25 (1.79% (104) (Colombia. cGhe d-lv_ire. in]ac and Pakista)o -04.2 0.10 -0.44 1.02 0.09 36 0.44 0.0 1.75 7.79 (0.58) (1.7) l.85), (4.0l (1. is) .0.03 -0.10 4.38 9.0 0.12 36 0.06 0.02 2.27 1.59 (0.45) (1.42) (1.W0- M.05 (C61 (C) All aditnes -0 36 0.07 0.93 0.26 S3 -0.03 e.46 0. 14 0.93 (1.29) (0.12) (1a4) (0.53) -0.36 0.07 0.93 -023 01 Sl 1 0.46 OL13 0.69 (1.22) QxlII) (133) (0.06) (OA9 -0.36 0.07 0.93 042813 0.26 8 04M 0.46 OL14 0.69 (1.10) (0.12) (li) OLm) (037) Eq- ion: (ogGDP&-IoZgDPt-i1- aG + at ogRERI + a2ATaz-1 + .3 hlODft-I + a4 CUt + aS T 1. S fn: 19018: aa_tl dam 2. Absolute vale f t-sttistic in prremee ad - shows aigtficance at S% kwl. 3. Edmat by OLS usiag pod dam: HCeekmdticicY-r-iSurt CnariaiG Matrix acne iaia 4. R2 adjwuted for dege sf fieedon. S. S.E.R: Staard Enr of Regssiu. Defrhuitons of variabks: 1. Redl GDP (in 1930 USS): BESD NPaioml ArrsnL 2. RER: Real ExchaWe RItes (1980-10): a fall of RER is deprccatic IMF. 3. ATR: AvengeTariffRafts(S).gireebytcl npondutiescoll-eedna raioof value of ip I--nstp dties: vriousCEMsadIFSdocunes: ce oriwona IFS. 4. IIGDP1t-1): Real gres demnc i-est.n- t as a rtio of real GDP oF die previous yr: World Ba Datbase (atial Aceosa. 5. Capacity utilization: slc diltresn between nil lagGDP al Broed vaus of ogGDP. wher foetd vales acr estimatd by LtGiDP -c+Titme Trenax. 6. When oneyn lagged RER was added to sl equstn. slu ekeftim - it wa rna aignrua. (1) Inive wade adjourn cigGDF-ogCGDPZ-1) - 0.06+0.035 RERt-0.021 ATR(t-1)+0.004 lI(GDP(t-I))1O+O1 RER(tl) (1.350 (-1.903) (1.191) M.0M (2) Les ieve adjustrs (logGDPt4cgGDPt-1) - 0.075+0.09 RERt-0.021 ATR(t-l)-0.0l8 U (GDP(t-1).D.439 R30-1) (0.747) -1.094) (-2301) (.0.16%) 3) All adiers (lqgGDPx-tlaGDGtM-) - 0.082+0.37 RERt-0A002 AT14t-l)-.0.0 IflGDP(tO-0002 RtR(1t-1) (1-37) (4.191) (4137) (-0.D69 - a - Table 4: TRADE LIBERALIZATION AND GROWTH: 19s393 (Detnent vwiable: el GOP prwdi) lOY caadty No. RER ATR(u-l) GDP(g-l) UghlIngla T C or &A. R2 S.E.R D.W P4taIs& (A) Integslwe Tade Adfmers harm .0.04 0.2 40.6 0.10 9 0.54 0.02 1.66 4.23 0.41? (0.35) (1.06) PAO) *0.03 e0i 0.6S -0.23 .0m 9 0.50 oL 1.95 3.02 (2.63r (037) (0.74) (0.33) AM17) <.00 4)Ve -0.33 -0.002 0.19 9 0.47 0o.2 1M. 2.79 (19) W) (0.236 (0.93) (0.74) Inkesia 4106 0.18 -0.34 0.16 9 .021 0.02 2.00 0.36 0t75) (0.20) (0.36) (1.00 403 0.44 .0.65 1.10 0.17 9 0.87 0.01 3.01 14.47 (3l93) (1.03) (.3l5 (9.3lr (353) 4X21 -0.68 0.33 40.01 0.46 9 022 o 0.02 225 0.64 (2.47) (0.66) (0.72) (1.6 (1.59) Mike 405 -0.70 9.82 4.02 9 -0.19 0.01 250 Il.6 (1D51) (4.50) (2.77r (016) -0.05 -0.70 9.73 40M 42m 9 0.74 0.01 2.53 6.4 (0.41) (4.l1r 0-34r 0.12 (0@15) -0.03 -.59 9.91 oo0 4-0.03 9 0.74 0.01 25S9 6.34 (05 (1.81) (4.41? 0156) (.4) mm 0.03 -0.13 2.16 -O.44 1 058 0.03 1.96 4.68 (0.09) (t25) (3.00 (1.09) 40.12 0.14 0.97 1.14 0.02 9 0.68 0.01 1.66 5.37 (0.61) (035) (153) 0.24? (0.07) 1.03 .145 157 0.02 -233 9 ."4 0.01 1.62 4s50 (1.29) O8- (1.55) (1.44) (1.66) Tniuky -0.06 0.0 3.97 0.01 9 0.79 0O=5 2.56 10.94 (2.90) (633? (1.71) 9109) 40.07 405 4.85 -L02 4000 9 0.74 0.01 2.47 6.80 (3.m02)r 022 (1.43) 042) (0.03) .0.37 40.07 3.51 4.003 024 9 0.81 04.O 2.14 9.47 (3.41? (5.3ir (1.55 (1.75) (1.65) -9- Table 4 (acu'3d I()! CaicIly No. RER ATRh(.1) GDPD-2) Utilllntim T C r f oba. R2 S.H.R D.W P-stadmic (11) Lms Intnlmai Trade Adluelera Cioabla 4.07 4213 -1.07 0.22 9 0.59 0.004 2.02 4.91 PO.9) P0.94) AM22 GA?) 4.07 4.17 -1.72 016 0.24 9 0.63 0.0)4 2.46 4.45 (2.26) (1.61) t0.41) (1.37) (4.23) 0.01 4214 -1.43 400. 0.30 9 0.69 O.C04 2.49 5.49 (2.03) (2.32) (0tn) (1A7) 1 t5.17) Cbs dilrafe 4.010 .0.24 0.23 0.25 9 .0.034 D.02 2.00 0.91 .PAD) (2. 15) P.24) PP-SM -0.22 -a10 O.6s 0.u 0.24 9 0.55 0.01 1.99 3.42 (0.93) (P8U4) (1.22) (4.16)' (I.-C) -0.21 472 3.59 0.02 0.35 9 0.35 .m L99 Z.09 (2.2z) a59 (2.26r (2.4r (1.33) Januta -0.06 0.18 2.50 40.19 9 .0.11 0.03 2.118 0.74 (0.37) 047) (1.L3) 0.52) 10.B 0.004 1.97 1.10 4.09 9 0.35 0.03 2.17 2.07 0.01) P02°) (1.39) (3-M7 0.31) 0.01 04 2.14 0.04 .0.30 9 .0.37 0.06 2.35 0.46 0.02) to05) BS) (0.24) (0.44) Pakrun 0.06 0.04 .4.15 0.01 9 4.42 0.01 2.57 0.21 0.64 3 O34) (1.07) p.O) 0.04 0.03 -5.44 40.59 0.01 9 -0.67 0.01 2.10 0.20 P54) (P261 P.72) (0.35) (0-06) -.OS 0.07 0.0 40.O04 0.14 9 0.25 0.02 2.76 1.67 (0@9 (0P) (O."5) C2M) P0.52) Eqa (logGDPGDP'-k DPt-1)- t0 + 3l sgRERE- + a2 ATRh-I + d ILIGDPR-I + *4 CUt + aS T Notes: 1. S- rnq Nl. 2. AotlLte valtu orf z;dst inL psaren and I so_ signiane 5% keve. 3. EalesS by OS eai anwml dat; HencrladaateuyCanhmas Cot tflaeW e Mbix to eatimate ittica. 4. R2 a4ured kr degroe t*edef I. 5. S.Et Swhrd Eirer of Regrssiorn 6 D.W.: Dubin-Wason atitle. Definidn of wiables: 1. Real GDP (in 2910 $US): EuSD Natual Aon. 2. RER Real EFd Ra (1930-1; a fall fRER isdqr at: IMP. 3. ATJ : Aver Tariff Ram (S). vmter bl lkWn duties ealeted loS ofCIF waltd of ipom; lntda*ies: wk CEMs and IFS documnents; vba of tin: IFs. 4. IIGDP(t-1): Real pea dofmfa i_estorm as a ratio of real GDP or ime prvo4 year: Wodd Baivc Debas (tNeo Acna). 5. CUV Capaity ulidaion is tbe diiercte between ata 2ogGDP ad fitned walhs of legGDP. whe fitn valh s ae esinautd by L4GDP - c+Ttne Trend. - 10 - C. Export Perfornance I. Export growth. Table 5 shows the data on real export growth rates. The average annual export growth rate for the nine trade adjusters increased from 6% in 1978-82 to 9.3% in 198649. The intensive adjusters achieved higher growth rates compared to the less intensive adjusters. Ghana, an intensive trade adjuster, and Cote d'lvoire, a less intensive adjuster, make an interesting comparison. Both Ghana and C6te d'Ivoire had a similar production structure prior to the reforms. Ghana achieved an export growth rate of 24% per annum in 1986-89 compared to the negative growth (-1.3%) for Cote d'Ivoire. This is explained by the differences in the design of the trade adjustment programs for the two countries. Ghana devalued whereas Cote d'lvoire was not able to do so. In CBte d'lvoire, the lack of currency devaluation was further compounded by the relatively high cost of domestic and imported inputs. The price of electricity and petroleum, as a result of the divestiture efforts, increased in Cote d'Ivoire and is much higher than in Ghana; this reduced the international price competitiveness of Cote d'Ivoire's exports. This is an example of an adjustment program where price liberalization of public enterprises came into conflict with the objectives of export growth in the short-run. Table 5: GROWTH RATE OF REAL EXKIR1S m 199D USS AmsI Awc:g 1979 1980D 193 1982 298 19U1 IPli 19S6 2987 19818 1989 1979-1Z 1913-134 19849 (A) Inmeniuw adjuster Glum -1.7 -13.6 -L9 15.2 45.7 95 6.5 41.6 29.4 79 185 -22 -9.9 243 IrAeesia 2.3 5.5 -18O -9.0 4.7 0.6 4.0 14.4 16.8 -103 9.1 -41 -0.9 7.5 Mexico 12.1 6.1 11.6 21.5 13.6 5.7 -4.4 32 9.S 3.0 1.4 m2 4.9 4.4 mocw 05 3.6 0.2 4.9 8.1 2.8 4.2 2.2 10.1 182 4.6 2.3 .l 5.5 Twkey -93 4.1 R5.1 40.1 13.7 19.1 12.3 -1.5 27.3 19.7 4.7 3D.0 153 12.5 Avore 0.5 1.2 14.0 14.6 -1.1 7.7 2.1 32.0 1S7 7.7 5.0 7.6 2.9 101 CB) Less intive adiusiers CC'ubia li4 5.1 -l.l -1.6 4-9 103 14.4 2.7 54 3.5 6.S 0.0 7.9 91 Ce d'lwire 2.3 12.7 6.9 14A -.0 30.3 -0.9 -3.4 -5.2 -16.9 205 59 0.5 -1.3 Janiica 2.0 -15 4.0 -12.4 9.4 17.7 112 6.7 12.0 2.0 9.2 -2.0 17.8 7.5 Palaun 13.3 19-5 1B2 -60 24.6 -3.7 -0.4 327 32.3 -1.6 11.5 113 6.5 13.8 AVeg6 5 9.0 4.3 -4.6 63 16 6.1 14.2 6.9 -13 IZO 31 7.0 7.5 (C) All adpustn Averg 33 4.6 9.7 6.1 22 5.3 3.9 13.0 13A 2 8.1 5.9 4.7 9.3 '1979-t2 pe-reform period. 103-85 is a4junent and 1916S9 is Pet-eform peidL Soee: Word Bank Dalba 03ESD. 12. Export Diversification. The diversification of exports was an important objective of the trade program; it was aimed at reducing the vulnerability of the trade balance to external shocks. Table 6 shows that the average share of manufactured exports in GDP for the nine adjusters increased from 4% of GDP in 1978-82 to 6.4% in 1986-88. Both Indonesia and Mexico were highly dependent on oil exports for foreign exchange earnings prior to the adjustment proggrams. Both managed to diversify exports and reduce their reliance on oil exports as a result of the adjustment program. In Indonesia, - 12 - manufactured exports increased from less than 1% of GDP in 1978-82 to more than 5% in 1986-89, while in Mexico, it increased from 1.2% to 6.4% during the same period. Ghana is the only intensive trade adjuster where the share of mamnfactured exports in GDP did not increase (0.6% of GDP). Table 6 MANUF&CrURED EXPORTS AS A PERCENrAGE OF GD? 1973 19 ism3 1931 iP l293 1934 193S 19t6 1937 1938 197B-92 t- 19364 (A) lIrembe tade adhutem Gna 0.7 0.5 0.6 0.1 OA 0.4 02 0.2 0.4 0.6 7 0.6 0.3 0.6 lusl1nea51 0.4 0.9 07 01 0.9 1.9 2.5 2.8 3.6 5.7 6.6 0.7 2.4 s3 Ma. 1.6 1.3 1. 1L 1.0 2.2 2.4 2.5 5.5 7.0 67 12 2.4 6.4 M 2.6 2.9 3.0 43 4.6 517 6.9 6.2 7.2 8.2 3.5 6.5 7.2 Tusrk I0 CS IA 3.0 4.9 5.6 L3 9.8 7.7 9.9 10.6 2.2 7.9 9.4 Avge 1.2 1.3 1.4 2.0 2.4 3.2 4.1 44 4.7 6.1 6.6 1.7 3.9 LB (1B) Lan ileniveSns Colubia 2.3 2.6 2.4 2.3 .9 15 1.6 i.9 2.3 3. 29 2.3 17 27 Cae diwire 2.0 23 2.8 3.1 33 3Is 3.0 3. 3.1 2.7 2-9 2.7 .5 2.9 _aLi 17.1 17.7 2.9 23.8 25.0 14.1 72 17.5 16.2 14* 13.3 19.3 17.3 15.0 Pakistan 4.B 5.9 5.4 53 4.6 611 5.1 5.6 7.2 9.1 82 52 6.1 8.1 Ave- 65 7.1 SA 8L6 6.2 6.5 7.6 7.2 7.2 7.4 6.9 7.4 7.1 7.2 AveageC 3.0 3.6 3.5 3.6 3.3 4.0 3.5 3.3 4.2 4.9 4.6 3.4 3.7 4.61 (C) AU Siyns A eW 3.6 3.9 4.5 4.9 4.1 4.7 5.t 5.7 5* 6.7 6.7 42 53 6.4 No- 1. Maeufrssrd ecM ind.&Ishatcal andre.nAI. basic mencm . acshitay tutteeQL. alr luaanrdl ased Sopds neo dssMe4 cdaEagf nn-ea nab. 2. - csdoirs _umic. 3. J9734f2 h Is prem pmnot 193345 is adjutanti ad 19B6U n pn rmfima period. Semn: LVort (O) in cawn u in basseS US dsiln: Wed Tables; GDP at o prim in haut US ddl: IEC Mtinal AaS. 13. Indonesia's dependence on oil was substantialy reduced by the expansion of traditional manufactured exports of textiles and plywood products. While these exports were important, the non- traditional manufactured exports also increased, e.g., plastics, ceramics, basic metal products, glass, paper, rubber products and footwear. The growth in non-traditional exports was accompanied by a shift in the export destination-away from industrial countes markets to other developing counties. The share of export to industrial countries decreased from an average of 75.1% during 1978-82 to 73.3% during 1986-89, while the share of developing countries increased from 22.8% to 26% during the same period. 14. In Mexico, the growth in non-oil exports was also outanding. Prior to the trade reform, export earnings were mostly concentrated on natural resources - agriculture, mining, oil and tourism. Trde reform encouraged the exports of manufactured goods. These exports grew by 33% in volume - 12 - terms during 1985-87. By 1987, manufactured exports (excluding exports from "maquiladoras") were the single most important source of export revenue, representing 36% of exports of goods and non- factor services. Mining exports generated only 33% of export revenues and agricultural exports about 6%. Even within manufacturing, naural resource based industries (food, beverages and tobacco, wood and paper products) continued to decline in importance - with the notable exception of glass, cement and other non-metal mineral products, whose exports increased significantly during the 1980s. Exports of labor intensive products (textiles and apparel), by contrast, grew significantly during 1985-87, reversing the poor export performance in the 1970s and early 1980s. Among the other metal industries, transport equipment, metal products and machinery, and basic metal industries (including iron and steel) substantially increased their share in total manufacturing exports. The share of exports to the industrial countries increased consistently, reaching almost 90 % in 1989, while the shares to developing countries and non-market socialist economies declined. Particularly, the share of exports to developing countries nearly halved from 15% in 1978 to 8.8% in 1989. In conclusion, Indonesia and Mexico succeeded in diversifying both export structure and export markets. This helped to reduce the vulnerability of their trade balance to exernal shocks. 15. In contrast, in Pakistan, although manufactured exports increased from 5 % of GDP in 1978- 82 to 8% in 1986-89, the export base remained relatively narrow. Export growth came primarily from the growth in traditional exports, all of which are primary commodity-based, e.g., taditional cotton, cotton textiles and rice exports, which account for two-thirds of merchandise export earnings. The increase in traditional exports also accounts for the increased importance of developed countries' market for its exports. At the same time, the share of exports to developing countries declined. The foreign exchange eanings from the traditional exports, however, fluctuated substantially. 16. Exports and the Excbhane Rate. This section verifies whether the improved export performance can be attributed to the exchange rate policies adopted by the adjusing countries.fl Export equatons are esmated in which the explanatory varables include the real exchange rate, uncertainty in the exchange rate and the world demand for exports. - The estimated export equation, using pooled data for the nine countries, had a poor fit although the real exchange rate had the right sign, i.e, currency devaluation encourages exports and it is statistically significant. 17. The export equation is also estimated for individual adjusters and the results are reported in Table 7. The coefficient on the real exchange rate is significant for most adjusters and it has the correct sign. A cross-country comparison reveals the differential impact of the exchange rate on exports. Both Turkey and Ghana substantially devalued their currency compared to the other adiusters.1l5 Turkey's coefficient on the real exchange rate is twice that of Ghana, i.e., Turkey's export response to currency devaluation was greater than Ghana's. The differential impact of exchange rate on exports is consistent with the previous finding on the differential impact of the exchange rate on GDP growth. 3W See also B. Balassa, Incentive Policies and Agricultural Performance in Sub-Saharan Africa, WPS 77, The World Bank, August, 1988. This study found a high export response to depreciation. 1 The coefficient on exchange rate uncertainty is significant for both Ghana and Turkey. - 13 - Table 7: tMPACr OF T1E EXCHAE GE RATE ON EXPOR1S -Icinia Tdablc meal epU Cda*ia COle diwire Gba Idesih Jamaica Mexic Morocc Pablan Turie Rel eoxbowe raie) -05 4.7 -1.151 -0.540 .377 0.357 0. 4.946 -2.497 (141ir (LI217 (2.4 (2.35) (1333 (1.64 (DM3 (1.33) M.65)* Ezgbaw nexetuaelay (OOS 0.016 .0.2 -0.3S -Q 012 40.013 0014 -0.O3 -0.044 (2.35) (1.093 (4.3W (1.79) (0.6) P.95) (1.46) 10.05) (391' Wued dCma.d(3 0.n15 410117 -2Z212 -1.643 1.409 0.4116 2545 0391 -1.010 (2.1) z m2r (3.06r (13) (LSiD aim ([.32 w n02 (0.54) Cana= 1.663 10.72 2114.97 26.75 -3.26 05ZI -15.92 4.475 26.15 (M63) (33 (335) p.26) (3) tO1M (02) (1.293 R-sq4A_ 0.993 0.629 Q.715 0.63S 0967 0.323 0941 0.93 J 961 &E.R 0014 0.075 am 0o.7 ons 0.97 41049 w.sW9 0.0c7 D.W. 2.496 1.73 2256 1315 2.167 1.3 1.2OS 2.99 1.711 F-zdsetics 641.2 2.M 3.113 2351 39.57 6S12 2131 17.67 32.7 Rel eama c(1) -0.750 4036 4.aN9 -.6103 017 -m3t5 0291 0A4123 -2.3U3 (7.43) (Z.42r (035 (.93) (11.44) L16r (0.59) t092) (1.75) WVeld d_ml(t) 0ll .*0597 0369 -LIII 1.974 164 2.29 16 1.-0 (0.29 (1.493 0.13) (1.9)) (6.5Br aAg) CL42)' (1.61) (0.49) ccogm 7.97S 12.54 5.11W 31.92 9.131 -320 -12.19 -1123 7.405 *7.06) (35) (CLIP) L 9) (DAN (I.15 P9))) (310) R.u i 0.9l3 0.534 0.152 0319 0Q959 0.737 03 o.111I 0.43 S.EJt. om7 0.069 0.315 n09 0O" 0.109 0.46 0.0AM 0.248 DW. 2.099 1.709 0.98 1.44UZ 14 0.10 1.367 Z.37 1.49L F-srdea 242.5 3.440 0.531 1.912 71.07 11.10 4263 23.90 16.20 (1) IX - aD +41 augRE + A2 IcEltUt + 43 I%VfDr 3Z) gX - al0 + *1 oRERM + 2 logWDi 1. S*kIamle w 9113. sm.i. a absltSe VtEof IWt a jatetl amn - haws ' :lii at5S leve. 2. EP-dam by 0O5; _ CovGaim Mazrh an ge cote a ient. 3. S.E.R SatIad Ew df Rcrozion. 4. D.W.* Durbia-Wem ie Derinitn or vriables: 1. RER.: Rel em (191-100): a fli it d0kd adum 2. Real esgm ae re vub is _dmnkd by twywr amlg 3. Weld dcund is apparxuted by OECD GDP at com pnoCs. 4. Rodl q t Valueof d ia 119 c0 ns. Sowrc. IF5, BES.D585 i (N I Am). D. rade Reform snd the Trade Balance 18. Since most trade adjustment programs were imitiated from an initial position of large macroeconomic imbalance, one concern is whether trade libeaizafion would conflict with the balance of payments objectives. Table 8 reports the data on the trade balance before and after trade reform. Th1 trade adjusters, as a group, improved their trade balance from a deficit of 3 % of GDP in 1978-82 to a surplus of 1% of GDP in 1986-89. A comparison of intensive and less intensive trade adjusters shows that the itensive adjusters succeAded in achieving a higher trade surplus in 1986-89 thm in 1983- - 14 - 85. There are, of course, substantial differences between countries within the group. Among the less intensive adjusters, the trade balance for Jamaica deteriorated sharply in 1986-89, while that of Colombia improved substantially during the same period. 16' In general, trade reform did not cause the trade balance to deteriorate despitz the increases in imports resulting from import liberalization. The improvements in the trade balance is explained by the greater increases in exports in comparison to imports and reduced domestic absorption. Tablt S TRADE BALANCE AS A RATIO OF GDP [ie l911 Ui) A~cg 1973 1579 1910 19B1 19Q 1933 14 1'M 193 1937 133 1959 19M82 19%- 19649 (A) Ilaetui uas ads. Gb 04 LA4 -0.7 -02 4.4 0O5 0.7 4.2 1.L 03 -042 QJI I 0.4 0.7 lad1s 16.1 13.9 US 0.0 -16 -2m 2.9 0.6 1.4 4.0 7.0 6.7 LI 03 4.8 Uesic 1.7 0.3 -23 -3.1 4.7 9.9 9.4 7.7 95 10.6 03 6.9 03 9.0 La MG= -14.4 -14.7 -10R -11.3 -10.3 -5.9 *61 -5.0 -59 -62 -3.3 4.4 -122 -5.7 -5.4 Turkey -&a -7J -7.1 -44 -D 1.6 -32 0.0 -5. -4.0 0.0 -44 -SJ -16 -24 A%aW -09 -1.4 -1.7 -3.3 -1.0 QO 0n3 0.7 04 1. 2.3 15 -17 05 13 *13) L1s labiie ads Cd1a 1.3 2.4 06 -2.0 -335 -to 0.0 2Q 53 53 6.1 6.3 -02 03 62 CMk d'lwsc -9.7 -75 -6.1 -.6 2.9 3.1 9.7 10.4 9.1 &.1 7.6 143 4.4 7.7 93 *3in 4-63 -15 -21 -33 410 -6.9 -5 -12.4 -53 -7A4 47 -21.4 -42 -93 -13.7 PFWioz -9.3 -14.1 -11.6 -40 4.4 -3.1 4.6 -51 -IA 02 -13 -0.1 -33 -45 046 A-MP -61 -52 48 -2.7 -33 -2 -0.9 -12 L.9 1.7 -1S 401- -44 -4A 4 (C) AN aiusim A%CW -32 -30 -3.1 -3 -20 -.0 0.0 O42 1.1 13 0.5 0 -2.9 404 09 T'ade RHahn IewptNFS mom iqowe. 1978-2 lis pe-efrm ped: 1910-45 is amjsu sad 19319 i pwt periodL 3n Weld B11 DBan (BIED Nshmui Amat). E. Exhange Rate and the Labor Market 19- The interaction between exchange rate and the labor market is central to the export supply response- Usually, currency devaluation provides an export supply response by reducing real wages and improving the international price competitiveness of exports. It has been argued, however, that a f' The deterioration in the trade balance in Jamaica, partly, was a result of Hurricane Gilbert. MI Some adjusting countries were forced to reduce the level of external imbalances before the adjustment process began; this was caused by the change in the international environment for commercial lending. The data for Mexico shows that it increased the surplus on trade balance from 1983 onwards. - 15 - cut in real wage can result in output losses by adversely affecting the incentive to work.' A currency devaluation aimed at reducing real wages can be offset by productivity losses. Table 9 compares the data on real wage and the ratio of output to real wage per worker for five adjusters for which data are available. For the adjusters as a group, real wage per worker increased from 102 in 1978-82 to 111 in 1983-86. The output to wage ratio, however, increased from 100 to 122 (1980=100) during the same period. In Turkey, real wages fell from 111 in 1978 to 98 in 1986 (1980=100), but the output to wage ratio increased from 101 to 161 during the same period. In general, the impact of exchange rate adjustment on the labor market appears to be consistent with the export objectives of the tade programs. Tabc 9: REAL WAGE AND UN LAOR Cf IN THE MANCIrtG SECIVH OF TM AItfN COUNrR1S (19-I-) RNl Wqt Averwc Per Wkmr 1973 1V79 1910 1931 12 120 19U 1i 19J6 19W6 1963- CIdWmba 109.3 1653 lO. 101.3 1 106.S 117.3 11I9 1511m 169 g0 tnsk 923 93.6 IO0.0 103 123.2 1264 127.5 1392 143A 103.4 1.6 Mcziw L73.4 ILU 100.0 103.0 193.6 74.3 n.6 9S BSA 101.9 11.2 Pikw 59.4 962 IO.0 99J. 11139 119.3 1D0 124.2 130.4 97.9 1242 110 6 11015 1000 1043 97.3 96.0 K4 3L4 963 1016 00. A-. 100.9 101.6 100.0 2033 196.3 13.3 105.0 IO3 tZ.6 IO3 110.3 3m] Oipv/ Rtal WSWe 1973 199 190 1931 1912 1933 1916 1i5 1M I97W 1933-3 cdoma 363 995 1.o 104.4 9"1 93. 97.6 1105 19. 97.3 99.0 TNk=!da 93.3 127 1m. 104A 953 M103 1013 la" ICIL7 90. 1W7 ''--io g 96.6 997 10. 96.5 96.6 12 133 27 lfl.9 97.9 1345 1'akira 17.6 39.6 10. IOIL6 117.9 114.0 3.2 116.6 117.0 1.7 1151 1Tar1zy 1013 365 1mo 1143 123L9 133 156 170.6 100.7 16O2 2552 Av:nW 93.2 95.6 1000 105.6 107.4 116.9 124 6 119.7 1M4 121.5 197W mispu.efti and 190416 aOm.n Smu : World TaM. Wdrld Bamek 20. It has been argued that a devaluation can be neutralized by an increase in wages. The estimates of the real wage equaion, with the real excmange rate and the employment level included as explanatory variables, are reported in Table 10. The data are ken fom World Bank tables. The estimate for the five adjusters, using pooled data, shows a significant correlaton between real devaluation and the real wage. A real devaluation is associated with a reduction in the real wage level after controlling for the effect of employment. The elasticity of wages, with respect to devaluation, is 17 See J. Stiglitz, "The Causes and Consequences of the Dependence of Quality on Price," Journal of Economic Litature, March 1987. - 16 - about 0.5. Adjustment in the real wage was, thus, a major channel through which the adjusfing countries improved the interational price compettiveness of their exports.-L Tb 10t EFFECT OF EXBCAMJ RATE DEVALUATION ON WAGES Olpxdfs wibb: ad wmug) RER(1-I) EzIDfl-I) C No. droW 32 SY-IL D.W 1Fabtis (A) Five diuses 0.5 0.3n2 Q2D 30 0269 0.46 IS.IS 720 oar (o1l3r rsm3 (B) Inteeivc ude a.Ljsh hmesia eCItIZ 0.473 0.59 6 0.4 2 1.747 3.1106 (0.43 a-=0) (L Maxi LB1 -ISi.5 2.752 6 O9tJ 0.017 ZAGI 31-C06 (L451)r (L17r OMP Tay IJ09 2.3 -5953 6 0.1129 2.54 3mA n3s6>, ¢Lr ¢us} (C) Lan hasive ude aduusa Caxitla 0.631 -2.6 7.6I3 6 a.im 02i6 2.311 15.066 (1.36) M.17s5r (7AW PAbm 0.0113 13 -4.Itl 6 ODil 0.06 1.774 6&796 --M a -or 0-2ae bg Rod Wage - t + al hgft-1 + ii hgEuiloyt-I 1. Sak :w n ISS--itl. maul da alu e ia s iuface: SS Int 2. Eatnamd by OLS; CwlCaur Mari to e cias. 3. SYER: Stanar Enr orRnasae 4. D.W.: DuabinWm D dfmiim .fv rbbL 1. Reat Wag It! egnin per Iae in d-me nunab ribg seewr (U-lUtl: WOld This. WOdd Pa 2. RER R:sI Pd of (1933-lOD a fa1 iaa reinda 3. EumpIer Epkomn bab. mm wn _ c macb (1911-1)- Weld Tab. Woldd fL Ls The regression results for the individual adjusters suggest a far closer relationship between exchange rates and wages for Turkey, Mexico and Colombia as compared to that of Pakistan and Indonesia. - 17 - F. Conclusions 21. While the benefits of trade policy reforms have gained widespread recognition, several concemns persist regarding the policy design and impact of these programs. This paper examined the impact of trade reform on efficiency and growth in Dine trade adjustimg countries: Colombia, Cote d'Ivoire, Ghana, Indonesia, Jamaica, Mexico, Pakistan, and Turkey. Evidence shows that all trade- adjusting countries reduced non-tariff barriers, while some actually managed to proceed further by reducing tariff levels on imports. 22. The countries that managed to reduce the tariff levels on imports experienced a greater increase in output growth in compaison to the countries that merely managed to reduce non-tariff barriers. Similarly, the more intensive adjusters experienced a higher export growth in comparison to the less- intensive adjustrs. Increased import competition intrcduced by itensive trade adjustment forced domestic firms to improve their efficiency. Simple regression estimates suggest that a reduction in the import tariff level is associated with increased labor productivity. At an aggregate level, the reduction in import tariff level, for a given level of investment, is also associated with increased GDP growth rates. 23. A comparison of the individual adjusters highlights the differential impact of productivity- enhancing trade reform. Indonesia and Ghana were intensive trade adjusters. Both traditionally relied on resource-based exports. Indonesia made substatial progress in increasing manufactured export and thus diversifying the export base. In comparison, Ghana's share of manufaured exports in GDP did not change substantially after trade reforL This suggests that countries with better human infratucture and well-fmnetioning markets tend to benefit more from producivitty-enhancing trade reforms. Improvements in supply response in the least developed countries would, therefore, require investment in human and capitl infrastructure. Policy Research Working Paper Series Contact Title Author Date for paper WPS1425 On the Intersectoral Migration of Donald Larson February 1995 J. Jacobson Agricultural Labor Yair Mundlak 33710 WPS1426 Russian Unemployment Its Simon Commander February 1995 V. Reid Magnitude, Characteristics, and Ruslan Yemtsov 35195 Regional Dimensions WPS1427 Corporate Govemance and Equity Stijn Claessens February 1995 F. Hatab Prices: Evidence from the Czech 35835 and Slovak Republics WPS1428 Short-Term Supply Response to a Bruno Boccara February 1995 M. Pfeiffenberger Devaluation: A Model's Implications Fabien Nsengiyumva 34963 for Primary Commodity-Exporling Developing Countries WPS1429 The World Trade Organization's Bemard M. Hoekman March 1995 F. Hatab Agreement on Govemment Petros C. Mavroidis 38535 Procurement: Expanding Disciplines, Declining Membership? WPS1430 Intergovemmental Fiscal Relations Richard M. Bird March 1995 G. Coward and Poverty Alleviation in Viet Nam Jennie I. Utvack 80494 M. Govinda Rao WPS1431 The Industrial Pollution Projection Hemamala Hettige March 1995 A. Williams System Paul Martin 37176 Manjula Singh David Wheeler WPS1432 Using Financial Futures in Trading Ignacio Mas March 1995 K Binidey and Risk Management Jes0s Saa-Requejo 81143 WPS1433 Enterprise Restructuring in Eastem Bemard Hoekman March 1995 F. Hatab Europe: How Much? How Fast? Gerhard Pohl 35835 Where? Preliminary Evidence from Trade Data WPS1434 Govemment Credit Policy and Charles W. Calomiris March 1995 P. Infante Industrial Perforrnance (Japanese Charles P. Himmelberg 37642 Machine Tool Producers, 1963-91) WPS1435 The Political Economy of Formal Arup Banerji March 1995 D. Evans Sector Pay and Employment in J. Edgardo Campos 37496 Developing Countries Richard H. Sabot WPS1436 An Empirical Model of Sunk Costs Mark J. Roberts March 1995 J. Ngaine and the Decision to Export James R. Tybout 37959 Policy Research Working Paper Series Contact TlhU Author Date for paper WPS1437 Travel Mode Subsfttution in SAo Joffre Swait March 1995 C. Jones Paulo: Estimates and tmplications Gunnar S. Eskeland 37699 for Air Pollution Control WPS1438 Trade Reform, Efficiency, and Growth Ejaz Ghani March 1995 A. Nokhostin Cad Jayarajah 34150 WPS1439 Nontariff Baniers Africa Faces: What Azita Amjadi March 1995 S. Lipscomb Did the Uruguay Round Accomplish, Alexander Yeats 33718 and What Remains to Be Done? WPS1440 Poverty and Social Transfers in Christiaan Grootaert March 1995 N. Sachdeva Poland 82717