Peru - Toward a more open economy

48012 UNDP-WORLD BANK TRADE EXPANSION PROGRAM COUNTRY REPORT 10 PERU Economy This country report is a product of the joint UNDPIWorld Bank Trade Expansion Program which provides technical and policy advice to countries intending to reform their trade regimes. The views contained herein are those of the authors and do not nkcessarily reflect those of the United Nations Development Program or the World Bank.  PERU Toward a More Open Economy Trade Policy Division Country Economics Department World Bank October 1992 Washington, D.C.  This report was prepared for the government of Peru in response to its request for assistance from the United Nations Development ProgramlWorld Bank Trade Expansion Program. The report was directed by Dominique Hachette (Consultant) with the participation of Roberto Abusada (Consultant), Federico Changanaqui (World Bank), Michael Connolly (Consultant), Patricio Cortk (Consultant), Geoffrey Shepherd (World Bank), Adolfo Stunenegger (Consultant), and Wendy Takacs (Consultant). The main mission visited Peru from February 25 to March 9, 1991. The office of the UNDP under the direction of M. Pierre den Baas provided excellent support to the mission. Discussions with different representatives of the government and of the private sector were extremely useful and stimulating particularly those with Ministers Carlos Bolo& and Victor Joy, and Javier Galvan and Char Pehanda from the M i t r y of Economy and Finance. The views and recommendations of the report are those of the authors and do not necessarily reflect those of the United Nations Development Program of the World Bank. Executive Summary Chapter 1. Tendencies and Open Questions Brief retrospective Situations in August 1990 Main issues Chapter 2. The Macroeconomic Framework After August 1990 The stabilization program of August 1990 and subsequent measures The impact of stabilization and adjustment programs The real exchange rate Prospects for the real exchange rate Other policies that affect competitive capacity Chapter 3. Iiberalization of Imports Tariff regulations Nontariff barriers The treatment of agricultural imports Imports of used goods Integration and the Andean Pact Chapter 4. Exports Export behavior Export promotion policies of the 1980s Anti-export policy bias in the 1980s Chapter 5. Agriculture Characteristics and evolution of the agricultural sector Evolution of agricultural pricing policies Evaluation of agricultural policies Chapter 6. Anticipated Effects of Trade Iiberalization Impact on industrial production and employment Fiscal impact Impact on specific industrial sectors Impact on the agricultural sector Chapter 7. What Remains to be Done? a Trf ifs Antidumping legislation and reference prices Nontariff barriers Remaining obstacles for exports Proposals for better use of agricultural potential Making liberalization irreversible References Appendixes  LIST OF FIGURES -Figure2.1 Net domestic credit to the public sector, 1985-91 Figure 2.2 Real money and real liquidity Figure 2.3 The nominal exchange rate Figure 2.4 Theoretical behavior of the real exchange rate after a monetary shock (Case 1) (Case 2 ) Figure 2.5 The real exchange rate in Peru Figure A5.1 Economics of the motor vehicle industry Figure A5.2 Negative effective protection to motor vehicle assembly Figure A5.3 Estimate of effective rates of protection for vehicle assembly, June 1990, August 1990, and March 1991  L I n OF TABLES Paee No. Table 1.1 Growth in production and expenditure in Peru, 1950-87 .Table 1.2 Economic indicators for Peru, 1950-1990 Table 2.1 Fiscal and quasi-fiscal deficit 1985-90 Table 2.2 Exchange rates and prices Table 2.3 Sources of change in the money supply Table 2.4 Interest rates in Peru Table 2.5 Daily exchange rates, December 17, 1990 to February 7, 1991 Table 2.6 Exchange rates and prices, monthly averages, January 1990 to February 1991 Table 3.1 Number of rates and average tariffs, April 1990 to March 1991 Table 3.2 Number of tariff items according to rate and code, September 1990 Table 3.3 Fiscal cost of import duty exemptions Table 3.4 Calendar for tariff reduction, July 1991 to January 1995 Table 3.5 Tariff structure, 1989 and 1991 Table 3.6 Formal nontariff restrictions to imports, by number of tariff lines, selected periods since 1981 Table 3.7 Principal sectors with more than 40 percent of their production covered by import prohibitions, up to D ece mber 1989 Table 3.8 Imported food products and inputs with compensatory surcharges, March 1991 Table 3.9 Specific duties applied to imported food products and inputs as of May 1991 Table 4.1 Export evolution: 1970-1990 Table 4.2 Evolution of CERTEX r a tes,selected periods 1981-90 Table 4.3 CERTEX: Subsidies and nontraditional exports 1980-90 Table 4.4 Tax on traditional exports, 1980-90 Table A4.1 Peru: composition of exports, 1970-90 Table 5. 1 Production, exports, imports, and growth rates for the principal agricultural products, 1989 and 1966-87 Table 5.2 Nominal direct protection for agricultural importables and inputs Table 6.1 Mustrial sector protection, employment, and production before the 1990-91trade reforms Table 6.2 Sectors likely to gain or lose from from trade liberalization Table 6.3 Effective protection of industry before and a f k the reforms of August 1991 and March 1991 Table 6.4 Tax revenue from foreign trade: 1989-91 Table 6.5 Effects on government revenues of a tariff reduction of 25 percent and 50 percent Table 6.6 The Peruvian automotive industry, 1965-69 to 1990 Table A4.1 Protection in Andean group countries: distribution of tariff universe, January 1991 Table A4.2 Nominal protection for the industrial sector, 1991 Table A4.3 Failures in the trade and foreign tariff liberalization program to April 26, 1991 Table A5.1 Institutional structure of the automotive industry in Peru Table A5.2 Production structure of the automotive industry in Peru Table AS.3 Basic data for the calculation of effective rates Table AS.4 Estimate of effective rates of protection for vehicle assembly June 1990, August 1990, and March 1991 Table A7.1 Annual ,rate of growth of overall GDP and of the manufacturing GDP in selected periods Table A7.2 Evolution of the nominal tariff structure Table A7.3 Nontariff restrictions on imports, 1979-86 Table A7.4 Survey of business persons w ith respect to government activity Table A7.5 Imports, Ja~I~ary-May 1991  Executive Summary Over the last thirty years Peru has experienced wide swings in government - from a liberal government that opened Peru to the world, through the autarchic socialism of military governments to the extreme populism of President Garcia. With the abandonment of the liberal model toward the end of the 1950s, Peru was afflicted by numerous problems that profoundly affected its social fabric and jeopardized its future. Notable among them was a loss of dynamism in the economy, especially in exports, a growing incoherence ib macroeconomic policy, and an ever more problematic dependence on foreign markets, all accompanied by increasing migration toward the cities. By 1989, hyperinflation was well under way, brought on by the extreme macroeconomic disequilibria provoked by incoherent economic policies. For 1989, inflation was 2,800 percent (7,600 percent in 1990). and it was accompanied by unprecedented declines in production (19 percent) and real salaries (60 percent) during 1988-89. Fiscal revenue fell from 14 percent of GNP in 1985 to 3.5 percent in 1989, while the money stock fell from 20 percent of GNP in 1985 to 6 percent in 1989. President Fujimori confronted three major challenges when he took office in July 1990, each linked to a structural problem: equilibrating and strengthening the macroemnomy, increasing the potential for growth, and stimulating institutional efficiency and flexibility, especially in the public sector. Macroeconomic f'ramework To stabilize the economy and reduce inefficiency, the new administration introduced a restrictive fiscal policy and coherent monetary policy and liberalized and unified the exchange rate. The authorities later broadened the program through measures designed to effect structural change, most significantly by opening the Peruvian economy to foreign trade and increasing the efficiency of factor h markets. To a large extent, the success of t s opening process will depend on the behavior of the real i exchange rate and the other factors that affect the competitiveness of foreign trade. Price deregulation, especially the sharp increases in the price of the goods offered by public companies, initially kept inflation high (400 percent in August 1990), but it later dropped substantially, though it was still in the three-digit range as of t h s report's preparation. In August the i combination of a high rate of inflation and a proportionally smaller increase in the money supply and liquidity under the stabilization plan generated a significant monetary shock from which the economy had not totally m e r e d by Febmary 1991, when real liquidity was still 50 percent lower than twelve months earlier.  The restrictive monetary policy meant a drastic reduction in the monetary financing of the fscal deficit. The real supply of credit extended to the public sector by the banking system fell from 210 billion intis in July 1990 to 58 billion intis in January 1991. Real domestic credit fell from its level in 1989, and net sales of foreign currency represented the most important source of monetary growth during the last quarter of 1990. Control of domestic credit will be decisive in controlling the money supply and inflation. However, if the monetary authorities continue to intervene in the exchange market in order to maintain a 'high' nominal exchange rate, they will lose control of both the money supply growth and inflation. High real interest rates have slowed the monetary adjustment and depressed investment and demand for durable goods. Since August 1990, when the exchange rate was unified, the government has followed a policy of managed flotation of the exchange rate. The authorities intervened heavily in the foreign exchange mar ket during the last quarter of 1990, apparently to maintain a high exchange rate for the overvalued inti. Government intervention has fallen off considerably since the final months of 1990. The real exchange rate has evolved in a worrisome fashion since August 1990, its systematic appreciation punctuated by several minor surges in early 1991. P art of the reason for the undershooting of the real exchange rate was the strong real contraction of liquidity together with an increase in demand for intis and in real intenst rates. Tbe same phenomenon was observed in the latest stabilizationprograms in Brazil and Argentina, among others. Other factors also help explain the appreciation. Tbe significant increase in the prices of state-produced goods (wntradables) with a significant weight in production costs, is one-gasoline and diesel fuel prices rose twenty four-fold in August 1990. Tbe sale of dollars obtained from rising cocaine exports, which could be putting downward pressure on the dollar, may be another. And the opening of the economy itself, with the expected increase in the demand for cash, has been clearly insufficient to counteract the effects of previous policies. Tbis is surely l i i to the recession, itself a consequence of the adjustment process and the weakened demand for imports. What happens to the real exchange rate will of course depend on what happens to the fundamentals. The end of the undershooting linked to the monetary shock, recovery of economic activity, new reductions in tariffs and remaining import controls, and tht gradual adjustment in tho prices of public goods toward their lower, 'normal' levels, will tend to generate a red devaluation of the exchange rate. That tendency could be jeopardized, however, if real herest rates remain high and the flow of narco-dollars continues t o grow. In any went, a policy of devaluation (or artificial maintenance of the nominal exchange rate) will be ineffective in maintaining or raising the real exchange rate. At most it can achieve only a short-term real devaluation. - Commercial liberalization imports and exports The new administration decided to drastically simplify trade restrictions, to redress the strong anti- export bias of the existing system and to release the growth potential of the export sector. The number of tariff rates was reduced from 33 in April 1990 to 3 in March 1991, with a maximum tariff of 25 percent (down from 84 percent) and a minimum tariff of 10 percent, although a few exceptions still favored steel producers. Consumer goods and capital and intermediate goods already in production were at the higher tariff, with most other goods subject to the lower tariff. A schedule of further reductions was announced for arriving at a single tariff rate of 15 percent by January 1995. Nontariff barriers-the principal instmment used by the previous government to restrict imports-were eliminated as well. Import prohibitions, which had protected more than 22 percent of GDP by the end of 1989 at a fiscal cost equivalent to 1.4 percent of GDP, were eliminated. Multiple .. exchange rates, prior licensing for imports, and registers and other admlnlrtratve barriers were removed; quality certification for the health and industrial sectors was simplified; the foreign trade functions of the national boards for rice, coffee, wheat, and milk products were abolished; and state import monopolies were eliminated. The majority of tariff exemptions were also rescinded-these had covered nearly 50 percent of imports in August 1990, reducing tariff income by 50 percent in effect. As is the case in many developed countries, agriculture is the major exception to these liberalization measures. Certain food products and agricultural inputs are protected through specific duties applied to minimum duty values. The resulting implicit nominal protection reached more than 180 percent for certain products (corn, sugar) in 1991, with obvious harm to consumers. The fundamental changes in trade policy also included the export regime. Exports had been affected not only by an increase in the apparent antiexport b ias of trade policy in the 1980s, but also by a growing discrimination among export activities. A aeries of measures taken since March 1991has eliminated, at least on paper, the greater part of this discrimination in trade policy. The new policy ought to reverse the declining export t r d of the 1980s, while the growth in nontraditional manufactured exports should compensate for the contractive effect of import liberalization on industrial activity.  Despite the salutary effects, export growth, particularly among traditional products, is likely to be limited because of other types of regulations (especially labor laws), continuing public sector ownership (especially in the mining and fishing sectors), fiscal discrimination faced by large exporters, and the inadequacy of physical infrastructure. The agricultural sector Agricultural land is scarce in Peru, with high opportunity costs-although its export potential is also high. Most of current production is tradable and the rest (potatoes, corn, beans) has a strong substitution correlation with some tradables, which increases its sensitivity to the openness of foreign trade. Agricultural production has stalled. Per capita agricultural product and the exportable surplus fell during 1966-87, with only rice, coffee, and yellow corn production increasing during that period. Including coca in these calculations changes them dramatically, since coca exports were equivalent to more than 50 percent of all legal exports in 1989. Agricultural price policies have traditionally left export production disprotected (taxed), with subsidies of agricultural inputs, particularly for irrigation and fertilizers, providing some minimal compensation. The interest groups favoring cheap food-urban consumers and agmindustrialists-prevailed over the interests of producers, remlting in urstable treatment of agricultural imports as the two sides jockeyed for influence, and in net negative protection for the sector. Substantial subsidies for mme inputs, such as water and fertilizer, have not compensated for the distortions caused by other government interventions but ratber have caused gross distortions @oth in their use and in the choice of products to be subsidized) and harmful ecological effects (salinization of arable land, among others) and constituted a considerable fiscal drain. Trade liberalization has significantly reduced antiexport b ias withii the agricultural sector and between that sector and the rest of the economy. Liberalization has also involved the progressive privathation of the institutions engaged in foreign and domestic transaction in agricultural produce and inputs, in order to stimulate efficiency through competition by e l i i g monopoly control of imports. Efforts have also been made to equalize incentives among diemu kinds of importable products and to eliminate discretionary elements of the trade regime l i e d to administrative, quality, and technical controls and other nontariff obstacles, thus increasing transparency d simplicity in foreign trade. Several recommendations for the f ?shape of agricultural policy are suggested by the undesirable effects of the policies in place before August 1990: the need to analyze the agricultural sector within  an economy-wide framework that considers, among other problem, the appropriateness of self- sufficiency in the production of some goods; to simplify and reduce the number of incentives; to assure greater policy stability in order to stimulate investment and induce the changes in production required by a more efficientallocation of agricultural and other resources; and to avoid using pricing policy as a redistributive tool, given its low efficiency in that role. Integration efforts: the Andean Pact After a long period of stasis, the Andean Pact has begun to attract new interest as an integration scheme, now that its members have adopted liberal trade policies. All tariffs on intra-pact trade were to be removed by December 31, 1991, and dismantling of the lists of exceptions was to be accelerated. All other lists were to be eliminated by the end of 1995, when the Andean Tariff Union is to be fully established. The common minimum foreign tariff has been reduced from an average of 28.5 percent to 17.9 percent, and the number of rates has been lowered from 19 to 7, with a minimum of 0 percent and a maximum of 50 percent. This implies a substantial reduction in protectionist measures at the intraregional level, but it is premature to evaluate the real e f f h of these decisions on intraregional trade. Intraregional exports grew an estimated 24 percent, but they still constituted only 4.4 percent of total Andean sales that year. Two-thirds of trade among the Andean Pact countries is still restricted . by the lists of exceptions and by violations of agreements, especially with respect to the common minimum foreign tariff. For Peru, the changes planned for the Andean integrationist trade structures do not appear to promise significant benefits. Peru is likely to derive far greater benefits from its policy of greater overall openness along the lines laid down in recent months and, eventually, from a possible free trade agreement with the United States, within the fimework of the Enterprise for the Americas Initiative. The impact of trade liberalization The trade reform begun in August 1990 reduced the average effective rate of protection of industry from 82 percent in June 1990 to 24 percent in March 1991, while the reforms of March 1991 increased the homogeneity of the effective protection levels. However, high levels of protection still remain for some activities, such as the garment and footwear industries.  The fiscal impact of trade liberalization is expected to be positive. If the general sales tax (IGV) and the selection consumption tax (ISC) are included, tax revenues for 1991 were expected to be US$ 796.4 million, a 100 percent increase over 1990. Liberalization will benefit sectors that together represent 60 percent of industrial product and more than two-thirds of the industrial work force. Sectors not favored by liberalization absorb only 3.3 percent of the national work force, and jobs in those sectors need not be totally lost if the sectors adapt to the new, competitive conditions. While that adaptation implies resource reallocation, it must be remembered that the effects of liberalization on production and employment in other sectors will be positive. The trade reforms have affected industrial sectors unequally because the tariff and nontariff incentives facing them were different and because not all nontariff barriers have been removed. Three cases illustrate this point: Effective protection in the automotive sector dropped from 79-141 percent to 17-18 percent. This implies that requirements for domestic content and obligatory exports should be abolished. Rationalization of the sector should follow, along witb reduction in its size. Automobile assembly will probably disappear, while truck assembly will probably survive after considerable adjustment, as will diesel motor production and some motor parts production. Some companies will become distributors. State intervention in the ferdlizer sector has been enomus. Fertilizer production and distriiution have been basically a state enterprise. Distribution was subsidized at irrational levels in recent years Q7 percent in 1988). At best, national production is only marginally competitive. The elimination of subsidies on agricultural inputs and products eliminated the extraord'i distortions such subsidies brought about. Now a 134 percent tax on petroleum-based inputs is causing a new distortion by eliminating the small remaining competitive margin of FERTISA (a sulphate and ammonium-nitrate producer) and imports may replace domestic production. Because socioeconomic evaluation suggests that domestic production should be maimined, elimination of the tax on petroleum-based inputs ought to be considered. Not all quality or regulatory controls have been dropped, although timy have been substdally reduced. Pesticide imports pose the quality control problem in extreme terms. These controls are costly, slow, and arbitrary, and include obligatory licensing. A special effortis required to make the system efficient.  In the aggregate, agricultural production will probably increase, as protection for the sector increases (or rather, its degree of disprotection decreases) at the expense of the rest of the economy. Production of exportables will increase, as will that of some importables (yellow corn, wheat sorghum, and milk, among others), as long as international prices do not continue to drop. Production of meat products and of the main domestic products will remain the same or contract. Rice production will remain the same (or even decline) if elimination of the subsidy for fann irrigation becomes a ,fact. Exports of farm products should increase and agricultural imports will probably maintain their current levels, given the decrease in the relative price of meat products and a continuing decline in international prices. The distributive effects of the trade liberalization will be mixed. For lower-income groups the effects are likely to be regressive because, as the relative price of wheat (mainly domestic) rises, the price of the food basket could also rise because of high per capita consumption of wheat, rice, potatoes, and corn starch. In the rural sector, the distciiutive effects will vary according to production type. Small farmers of the coastal and jungle regions will benefit from changes in the relative prices of sugar, yellow corn, and some exportables (coffee, cotton, fruit, vegetables) and of wheat over rice (rice will suffer from the expected elimination of the farm irrigation subsidy, even though it is protected by a specific tariff). Small farmers of the mountain regions will benefit from the probable increase in the relative prices of their produce, although significantly higher transportation costs will wipe out some of the gains. What remains to be done? The current administration has taken significant and brave steps to open the economy to foreign . markets. These actions require complementary measures to reiafarce the process, to reduce the probability of reversal and eliminate remaining sources of inefficiency. Ta * instruments The authorities ought to consider bringing tariff duties down to 14 percent very soon. This move would lower the indirect tax on exports implied by tariffs; reduce the dispersal of effective protection among productive sectors, which the current tariff structure still allows; and exert upward pressure on the real exchange rate, a welcome incentive for exporters. To fortify the new miff structure against protection-seeking interests, it would be worthwhile to commit the country to the new structure by consolidating its tariffs in the multilateral negotiations of the Uruguay Round.  Nontan~measutes Reference prices (official minimums, used good, agricultural goods, and goods in noncommercial operations) pose a serious challenge to trade openness to the extent that they are discretionary instruments of protection. They risk stimulating antidumping reactions among Peru's trade partners, and they enwurage low-quality domestic production if applied unequally to inputs for exportables and .importables. The Andean Group has approved regulations (Decision 239) for preventing or correcting competitive distortions generated by dumping and subsidies. A project for a similar measure (a Supreme Decree)has been presented at the national level as well. Promulgation of the decree should be accompanied by Peruvian subscription to the GATT Antidumping Code and Subvention and Compensatory Measures Code, which would make it possible to resist pressures to adopt and multiply reference prices. Although legislation against unfair practices and subsidies contributes to the transparency of administrative procedures, such legislation does not cover all the cases to which reference prices currently may be applied, and it can still be used for protectionist purposes. To be effective, such legislation requires limited use, strict time limits, and full disclosure of potential prejudice. Review of the motivations for the law is also important. The proposal in t h s report to i restructure the customs system includes recommendations designed t o limit the proliferation of minimum custom values. Import registers have been eliminated. Any movement to reintroduce them should be resisted to avoid the possibility of a recurrence of hidden admhbtrative protection. &Pons The strong antiexport bias of the pre-August 1990period has been substantially reduced but has not been entirely eliminated. Implementation of additional measures, including the prompt e l i i o n of all remaining taxes on exports, is needed to further reduce that bias. Exports would benefit even more from speedy reforms in factor mark- and in port and transportation infrastructure and security. The government has taken legaI measures to increase the fleribility of land and labor markets, and the prompt, efficient, and broad application of these measures will encourage the resource reallocation necessary to stimulate export production, especiatly of agricultural products. The shedding of extensive state land holdings in the mining and W i n g sectors, both clearly export sectors, is also desirable for speeding adaptation to the changes brought about by trade liieraliin. Andean area With the decision to implement the Andean Free Trade Zone comes a more urgent need for greater harmonization of national incentive regimens for exports and of exchange rate, monetary, and fiscal policies. Agricultural senor The agricultural sector has long suffered from differential and, on balance, discriminatory treatment. To exploit its comparative advantages, Peru needs to establish a structure of neutral incentives for all its productive sectors. Equal treatment, both across and within sectors is also a powerfbl line of defense against sectoral pressures for increased protection. Equal treatment among products within agriculture is the best guarantee that the sector will develop the potential implied by Peru's enonnous diversity of agroecological zones, so that the high opportunity cost of land resources will be effectively covered for each production area and each agricultural region will exploit its potential comparative advantages, despite significant intenone differences in production costs. Achieving neutral incentives implies establishing tariffs similar to those applied in the industrial sector and eliminating any prohibitions and specific levies (as those on several milk products, cereals and some food products). To moderate the domestic impact of strong fluctuations in international prices for a few products for which such fluctuations are particularly large, price ranges could be established as an exceptional measure. Tbe stabilization mechanism (variable tariff&without intervention of procurement agencies) must, however, allow domestic prices to reflect international price variations. Other indirect measures could enanuage resource reallocation within Peruvian agriculture that would be more consistent with its comparative advantages. Tbese could include eliminating farm irrigation subsidies, replacing subsidies on farm produce transport and on specific crops with technological assistance, improving infrastructure, installing agmindustrial processing concerns for primary products, and compensating farmers (by d irc et redistributive fiscal measures) for the adverse effects of abandoning price policies d ire ctd toward income distribution objectives or toward the e promotion of substitutes for coca cultivation. Tbe government agricultural procurement agencies, ENCI and ECASA, ought to focus their regulatory activities on domestic and foreign commercialization through i n d i i actions designed to defend competition, to provide technological assistance for producers, and to promote the creation of  effective mechanisms for providing information on prices and establishing adequate trade infrastructure. Their trade-promotion functions ought to be privatized or dissolved. Ensuring the irreversibility of liberalization Macroeconomic aspects '. To a large extent, successful trade liberalization depends on a coherent exchange rate policy and on the degree of stability perceived by economic agents. It is the task of macroeconomic management to ensure stability and some degree of improvement in the real exchange rate. Managing the fiscal deficit is key. Also important is preventing the reappearance of multiple exchange rates, which would completely destroy the relative uniformity in the degree of protection that has been achieved for activities that are competitive with imports. The policies of a floating exchange rate and fiee convertability work to that end. On occasion the Central Bank has intervened in the money market and pressured groups to invest in order to affect the real exchange rate through the nominal exchange rate. Any resulting devaluations are counterproductive for the objectives of stabilization. First, the price of tradable goods will increase immediately. Second, the purchase of dollars at a higher price will inject intis into the monetary base, necessitating a contraction of domestic credit in order to achieve the goals of monetary growth. A system of managed flotation is preferable for preserving cohexence between the goals of monebuy growtb and the exchange rate. Moreover, the benefits of an increase la the i excbange rate would be short-lived, especially wth the rapid price readjustment (and overshooting) that have occurred in recent months.' Tbe authorities also run the risk of destabiliziig inflationary expectations and inflation itself. 'Ihe authorities should minhize interventions in the currency market, - acting instead to influence real variables that affect the real price of money reducing the sizu of the state, privatizing state companies through debt-for-equity swaps, reducing the price of public services to their long-term equilibrium levels, and leveliig tariffs to 15percent. Any reduction in fiscal revenue from trade liberalization can be counteracted by transforming nontariff measures into tariffk, eliminating exemptions, and increasing merage. A drop in the real exchange rate will have the opposite effect. Thus management of the real exchange rate has become a vital element for reenforcing trade liberalization efforts. Microeconomic issues The 5 percent tariff must be eliminated in order to eliminate pressures to include more and more items under this tariff, which would selectively increase the degree of effective protection. The proposal for fixing a deadline for adopting a uniform tariff of 15 percent should be ratified. That step would reinforce Peru's bargaining position with the Andean Group in its effort to adopt a common foreign tariff. Specific levies, reference prices, and compensatory duties are instruments that naturally attract the attention of interest groups. To help them withstand pressure in the future, these instruments must be legally and clearly defined as soon as possible and their modes of application must be transparent for all economic agents (see section on reference prices). To avoid any backtracking or reversals on prior import licenses, foreign exchange licenses, or the myriad nontariff measures suspended or eliminated since August 1990, numerous supreme decrees must be consolidated into a parliamentary law or . promulgated by the executive. This legal instrument should also increase the cost of implementing any new protectionist measures and should empower economic agents affected by those measures to denounce them and demand their elimination. Legislation is also required to protect the tariff structure against the arbitrary leakages caused by new exemptions. The legislation should make explicit any desire to favor a given sector by requiring ad hoc budget transfers rather than requests for exemptions, thus exposing the social cost of such requests.  Chapter 1. Tendencies and Open Questions Brief retrospective Peru has experienced wide swings in government and policy during the last thirty years. Peru's departure from the liberal model toward the end of the 1950s was accompanied by numerous problems that profoundly affected the country's social fabric and jeopardized its future development. Notable among those problems was a loss of economic dynamism (table 1.I), a growing incoherence in macroeconomic policy, an ever more problematic dependence on foreign markets, and increasing urbanization. Economic distortions, land reform, and large economic fluctuations brought about the systematic loss of economic dynamism (table 1.2). A strategy of import substitution was adopted in the 1960s, accompanied by implementation of tariff and nontariff measures (such as the Industrial Law of 1970) and the emergence of multiple exchange rates. These measures generated high levels of inefficiency and stimulated corruption without halting the decline of the industrial sector as an engine of growth (table 1.1). 'Ibat inefficiency was intensified by growing government immaation in foreign trade and production. Price controls multiplied erroneous production signals, particularly in the agricultural sector, where behavior has been erratic since the 1950s. and undercut the government's income redistribution efforts directed toward the urban population. Interventions in the labor market reduced work force mobility, slowed employment in the modern sector, and encouraged informal sector activity. In financial markets interest rate controls, compartmentalized credit, and state expropriation of banks (during the Garcia government) slowed the development of the capital market and distorted the placement of financial savings. The agrarian reforms of the 1970s resulted in the virtual elimination of the land market, creating a major obstacle for private investment in agriculture. Foreign investment, in turn, was intermpted by the nationalization of foreign companies and the introduction of strict, discriminatory regulations. Violent fluctuations in economic activity also disrupted the rhythm of economic growth, as has been faintly reflected in the variable and modest levels of investment since the 1950s (table 1.1). 'Ihe fluctuations were the result not only of exogenous forces as different as irregular variations in the terms of exchange (table 1.2) and the d i u s effects of El N i , but also of enonnous political  changes, which affected expectations and confidence, and erratic macroeconomic management, which was reflected in variable rates of inflation, fiscal deficit, and the real exchange rate. The incoherence in macroeconomic management - linked especially to the combination of fiscal deficits financed by the Central Bank and to fued (or slowly adjusted) exchange rates - was one of the m a in causes of the disequilibria observed in the current account, in the fluctuating and growing rates of inflation (7,639 percent in 1990), and in the accumulated arrears in foreign debt service, which accounted for half of the U S 2 0 billion in total accumulated debt as of December 1990. Tabk 11 Growth in production and apendin in Peru, 1950-87 . w=-w!-) GDP 5.2 5.5 3.7 1.4 Exports suffcfed in particular &om all these fhors, with growth turning negative a i k 1980 (table 1.1). Along with the declining trend in exports came a slowdown in the importation of the capital and intermediate goods necessary for investment. Several partial efli~rts of the 1970s to reverse the trend were unable to counteract the strong sraiexpor? b i a s created by the set of distortions, violent fluctuations, and direct interventions in foreign trade. The increasing migration to the cities that took place during t hs time was partly a response to increasing violence and rapid demographic growth and i m partly the influence of the i p oa substitution strategy. 'Lhe i m p o r t - s u b M o n m g y , in particular, encouraged the development of only a few urban centers by favoring the development around them of industrial and commercial activities that made intensive use of idkamuuure, specialized labor, telecommunications, and the other sesvices typical of modem cities. 2.TbsprotectioProfiappoltrub~recton~tht(hoy~~,~,~~.loarmtberolrdvepro~trbniy(d romctiveoeu) of 'UI the production of exportable .ad noatdablo pods.  Situation in August 1990 The severe macroeconomic disequilibria that resulted from the incoherent policy combinations eventually exploded into an exchange rate-wage-price spiral accompanied by recession that drove the Peruvian economy into hyperinflation. A World Bank study found that inflation was 2,800 percent for 1989, reaching 6,000 percent in the twelve months prior to August 1990. The hyperinflation was accompanied by an unprecedented drop in production and real salaries. GDP fell 19 percent during 1988-89 and salaries dropped 60 percent. Hyperinflation effectively destroyed formal financial intermediation and eroded the public sector's fiscal base. Fiscal income declined from 14 percent of GDP in 1985 to 3.5 percent in 1989, while the money supply fell from 20 percent of GDP in 1985 to 6 percent in 1989. The crisis was also devastating for social services and physical infrastructure. Poverty was increasing in both urban and rural areas, with social indicators showing rising malnutrition and infant mortality. Infrastructural deterioration was diminishing Peru's capacity to compete internationally. Conditions worsened during the first months of 1990. The monthly inflation rate was rising, reaching 63.2 percent in July; the fiscal base continued to erode, with heavily subsidized public services; and economic activity continued to decline. International reserves were at a minimum, and the foreign debt grew with the payments delays. The country lost access to international capital markets. Main issues In July 1989, President Fujimori inherited an economy suffering from at least three struchual maladies that strongly conditioned the immediate and mid-term future. F i t was t h e m macroeconomic dm-. Tbis has been an obstacle to reestablishing normal relations with the rest of the world, especially financial relations, and to developing an efficient domestic capital market. It has also impeded the recovery of efficient delivery of public goods and services and exerted continuing regressive pressure on income distriiution. Success would require not only a strong, coherent initial effort but also perseverance. Second, the combination of low rates of investment, both public and private, with infrastructural deterioration, strong distortions in relative prices, and weakened investor confidence in the face of heavy contraband tr-c and violence (and recently cholera) explained the country's w e d fir prowtk as the new government took office.  Third, fnstinrtional ineficiencv and inflexibilitv, especially in the public sector, had sharply reduced the government's room to maneuver and implied that changes would be difficult, given the lack of executing mechanisms and adequate human capital. Table 13 Economic indimtors for Peru, 1950-1990 Year U ~ t o p t m /X+M) T . o j Ir@nim Unempbymm Reafa&ange Public C~mmt in& GDP (% wdjbm) a&ange @emmuage) mu mcwr pclblsc recwr flP80-loo) (1~-IW) Wdt fi GDP) m a : $tatistical Ycllbook for .. 3Bt public Sector b e Latin Americrn - 1990 Hditiao; ~ hlih to f a ~ & w . . l h . .. rod  Chapter 2. The Macroeconomic Framework After August 1990 From the beginning the Fujimori government took measures to stabilize the economy and reduce inefficiency. A restrictive fiscal policy, coherent monetary policy, and exchange rate liberalization and unification have been the pillars of that effort. Once the stabilization program was consolidated, measures were directed toward structural change, notably a significant opening to foreign trade and increased efficiency in the functioning of factor markets. To a large extent, the success of renewed market openness will depend on the behavior of the real exchange rate and other factors that affect international competitiveness. The Stabilization Plan of August 1990 and the measures taken later to reinforce it and to initiate structural reforms are briefly described in this chapter. The behavior of the real exchange rate between August 1990 and January 1991 is also analyzed, particularly with respect to the effects of the stabilization program and other relevant variables in order to predict its possible course in the near future. The evolution of the monetary variables that determine the nominal and real exchange rates-money and inflation, fiscal deficit financing, and interest rates-is presented briefly. The principal changes in factor markets that can reinforce the efforts to open the economy to foreign trade are also discussed. The stabilization program of August 1990 and subsequent measures On August 8, 1990, the government initiated a program of economic stabilization. On the revenue side, measures included increases in the prices and rates of goods and public services, elimination of price controls on foods, and the imposition of several emergency taxes. To reduce fiscal expenditure, an administration committee was established to ensure strict control of expenses. Reform of the exchange rate regime included unification of the multiple exchange rates and adoption of a managed float. To reduce inefficiency in the trade system, import licenses were elimhted and customs duties were rationalized and simplified. F i l y , emergency measures were introduced to soften the short- tenn negative impact of the program. Between August 1990 and April 1991, the authorities consolidated the stabilization program and introduced complementary measures to bring about structural change (for more detail see appendix 1):  Public sector wages and other expenditures were frozen, subsidies for parastatal agencies dealing with oil (PETROPERU),rice (ECASA), and other imported foods (ENCI) were eliminated, and incentives were introduced to encourage staff to resign or retire. The bank and insurance sectors were liberalized, eliminating monopolies in those sectors, , . interest rates were freed and there was a drastic reduction in the rate of monetary growth. Most import prohibitions and other nontariff restrictions and exemptions were eliminated, tariffs were reduced to only two rates (25 percent and 15 percent) and antidumping norms were established. Rules were established to reduce state participation in productive activities; a new penal code was passed with sanctions for economic crime; private ownership of land was guaranteed by law; and an antimonopoly law was passed, affecting various public companies such as ENCI, ECASA, the Public Steel Company of Peru (SIDERPERU), and the National Railroad Company (ENAFER). The impact of stabilization and adjustment programs Stabilization and structural adjustment measures in Peru have had a range of effects: on monetary growth and inflation, the fiscal deficit, employment, institutional structure, foreign trade, the financial system, and the exchange rate. These are discussed in the following sectioas. lcyscal d@&, money and iflation Imprudent fiscal management-increasing real expenditure without proportional increases in revenuegenerated ever larger deficits in the 1980s. 'Ibe central bank financed a growing portion of those deficits because the domestic capital market bad limited absorption capacity and foreign sources were exhausted. ?be resultant growth in the money supply clearly had inflationary effects in the mid- and long terms. And as inflation accelerated, real income fell. Government tax income fell from 15 percent of GDP in 1985 to 5.5 percent in 1989 while demands for financing rose to 16 percent of GDP in 1988 (table 2.1). State enterprises, which bad already been hurt by price controls, also suffered a loss in real income as they received delayed inflation-reduced payments. This phenomenon became particularly serious as the economy moved into inflation in 1990. Table 2.1 F isa cl and quasi-fsca 1985-90 ldef~it, merit of GDP) Central government income Central governmalt Extend expenditures .intetd Central government deficit Deficit of NFPs includii public enterprise&' 0.4 15 1.0 4.7 . 18 0.7 Ccntral bank l o- 2.0 . 19 3 .8 4.4 . 03 0.0 Credit to agwicm bank 0.3 1 .5 1.6 . 16 . 23 O 1. Public oector indcbrodneu 5.9 9.3 13.0 16.0 10.1 6.6 Extend furancin& 5.2 3J 2.4 4.3 2.9 3.6 ~omcltic f~rancing 0.7 5.8 10.6 11.7 7.2 3.o Notes: &' Lcluda d interat accumulated on total &of foreign debt (iicludiig hre paymenu .nd interat on hre payments). k' Noo-fuuncid public rsaor. e' Wudos furancing of d a cc umulated foreign interat payamts. Source: World Bask data. The previous had financed the growing deficits through money creation and the accumulation of arrears on its foreign debt. The new government took the more fiscally sound approach of reducing deficits by reducing expenditures and raising taxes. The stabilization plan of August 1990 involved substantial h i in the prices of goods sold by state companies and a devaluation of currency. The government relaxed price controls, making subsidies unnecessary. The price of diesel fuel and gasoline rose 24-fold, the price of electricity more than 7-fold (table 2.2). As a result, tax revenue for 1991 was estimated at about 7 percent of GDP,a slight recovery from 6.4 percent in 1990. Although these measures reduced public sector deficits and the need for fiscal financing, the rate of inflation still reached 400 percent in August 1990. The consumer price index rose 397 percent that month and the nominal exchange rate fell 160 percent.  Table 2 2 Exchange rates and prices (data ue average for period. July=100) Exchange mtes antis Der US$) Ckdlne exchange cost@ m:el (0~1.84) Elecrricky lnurnst BwY- mte January Fdnray Match April &Y June July sourca: cmrd W o f P l a r . Nomiarl oxcluap nto multiplii by the d m y i n d u Y Rriy ach.nrr me: oomiarl exchmge nto, ~ u l t i p l iby . i i Qmertic * During this time, the real money supply (MI) grew by only 105 percent (175 percent in mminal tesms), rdecting the restrictive monetary policy consistent with the objective of drastically reducing the rate of inflation and monetary financing of the fiscal deficit. The real stock of credit extended to the public sector by the bank s - fell from 210 billion intis in July 1990 to 58 billion intis in 1991 (figure 2.1). Recently, the money supply growth rate has been about 4 percent a month. With inflation at 400 percent in August 1990, real money declined 59 percent and real liquidity 45 per- a strong real monetary shock (figure 2.2). Real monetary balances recovered between September and December 1990, due in part to increased real demand for money but then feIl again as growth in the mminal money supply slowed after November 1990. Part of the decline in real liquidity was the result of an a b ~ pdecline t in the money multiplier as reserve requirements on deposits in the commea-cial bank system were doubled. Figure 2.1 Net domestic credit to the public sector, 1985-91 - - Fi g ure 2 3 Reel m o w and real liquidity  Monthly inflation has moderated substantially, from an average of 38.4 percent in the first seven months of 1990, to an average of 13.4 percent from the end of August to the beginning of 1991. The supply of real money has recovered somewhat from the shock in response to lower current and anticipated interest rates. The real exchange rate has appreciated 51 percent since August. The increase in the money supply (the sum of net international reserves and domestic credit for the bank system) moderated substantially during 1990, with the exception of the fourth quarter. That result basically reflects the counterpart of the central bank dollar purchases; real net domestic credit for the public sector was also high, although it remained below 1989 rates (table 2.3). In January 1991, the flow rate was 160 billion intis, somewhat less than before, although net dollar sales caused the money base to contract by 32 billion 1989 intis.It will be crucial to keep the flow of net domestic credit extended by the central bank and others, such as the Banco de la Nacion, low in order to stabilize it successfully and permanently. With nearly'clean" exchange rate flotation, domestic credit will be the principal source of money supply growth. It should therefore be strictly controlled in order to achieve the money supply growth goal of 4 percent a month. Wlions o pbwrr (I) - Tabk 23 Soarea of change h the mney supply f 1989 intis) h m d N T . 111 ( ) I@ / ) (a) om) (a) Interest rates In inflationary situations, nominal interest rates tend to rise to incorporate relief from expected intlation so that lenden will not SUIT= losses on the principal. Borrowers am disposed m accept that rise because they expect to pay in intis with less buying power.' The "expectation' element can be clearly seen in the decline in nominal interest rates immediately after implementation of the stabilization plan in August 1990 (table 2.4). However, real rates of interest in terms of goods rose; levels of 7.5 percent (in January 1991) a n t 9 percent (February) were reached. Such high ex-post rates are explained by two factors: the significant real monetary contraction while the demand for intis was rising, and lower than expected inflation. Table 2.4 Interest rates in Peru (1988-1991) Nominal r' Real 1989 Dec 20.0 Mu 275 Jun 27J % 27.5 Dec 23.7 1990 Jan 25.0 -3.7 Fcb 29.7 -2.2 MU 34.9 -1.7 w 36.0 2.4 May 46.8 2.9 JUO 51.0 -7 .5 Jul 51.0 -69.6 h g 44.0 23 -0 %t ' 18.0 7.7 Oct 15.5 9.1 Nov 15.8 6.4 'Dcc 18.7 7.5 f A&vc race (Banm Mi). &'Nominal intaert rate leu inflation. Note: I o M Mtho avomge for tho paiod. Prior to A u p t 19#). mtm wao tho maximum authorized by tho Caotnl Bank of Pau and included tax on tht i - n soukc: M i i o n wtimatu. lhe nomind achange rare Dual exchange rates in Peru were united in August 1990 as part of the stabilization package. During the first two weeks of August, there was substantial depreciation of the nominal exchange rate. Since then, a policy of managed flotation has been implemented. The ocofia or street exchange rate was 172,000 intis to the U.S. dollar (to purchase) on August 1; by August 15, it had reached 310,000 intis, just below the official exchange rate of 317,700 intis. The exchange rate depreciated to 440,000 intis in mid-september 1990, stabilizing around that level until November 30, a two-and-one-half month period of apparent intervention (figure 2.3). By December 12, the exchange rate rose to 570,000 intis to the U.S. dollar, appreciating to 535,000 intis by the end of December, apparently because of a significant accumulation of cash (table 2.5). Further nominal appreciation occurred and in early February the nominal rate settled at around 450,000. During those months, the Central Bank announced that it was not intervening in the market except to purchase dollars to service the foreign debt. The exchange rate has declined gradually since then. However, movements in the exchange rate in early 1991 seem small for a fiefloating system. Do they actually reflect a degree of control? F i 23 The wminat exchange rate  Table 2.5 Daily achange rates, December 17,1990 to February 7,1991 (thousands of intis per dollar) Fa&l /.OEOna 1 Unified Erchanpe && Buy I990 17 Dec 580 la Dec 555 19 Dec 540 20Dec 540 21 Dec 530 24 Dec 550 26 Dec 540 2 7 k 535 28 Dec 535 31 Dec 535 1991 02 JM 03Jm WJM OlJm OSJM 09Jm 10 JM 11 JM 14 JM 15 JM 16 JM 17 JM 18 JM 21 JM 22 Jm 23 h a 24Jm SJM 28 JM 29 JM 3OJm 31 JM 01 Feb WFeb 05- Ob Feb 07-  There is little empirical support for the hypothesis of exchange rate control by private banks. As intermediaries, their profits come from the spread between purchase and sales prices. Any oligopolistic behavior should be reflected in high margins in foreign currency transactions. Given the strong competition in the street market, it is not surprising that margins have been rather low compared to the spreads that are common in most developing countries. In fact, spreads have fallen 'since unification of the exchange market, dropping to less than 2 percent in February 1991, suggesting that monopolistic behavior on the part of private banks is not a credible explanation for the apparent stability of the nominal exchange rate during the first three months of 1991 (table 2.5). The real exchange rate The real exchange rate plays an important role during trade liberalization. An appropriate real exchange rate can reinforce a favorable response by exports, while an appreciatinp exchange rate can raise the costs of trade liberalization by reducing the competitiveness of Peruvian exports in world markets and making imports more attractive. By any definition, the real exchange rate has been appreciating systematically in Peru since at least July 1990 (table 2.0, deflated by the traditional price index (parity) or by a salary index (costs). What happened during 1990 and early 1991 and what does t h at imply about the likely future behavior of the real exchange rate? In the long run, the real exchange rate is determined by real factors, including the foreign tenns of trade, tariffs and other trade barriers, current and capital accounts, the pattern of governmental spending, consumer preferences, tmnsfets and intementions in the nontradable market, and technological changes. In the short run, both monetary factors and disequilibria in any market (especially nontradable goods) can influence the process. The price liberalization of August 1990, together with a reduced rate of monetary growth, caused a strong real monetary contraction of approximately 50 percent (figure2.2). That contraction was accompanied by a sharp increase in the demand for intis and, consequently, an increase in red interest rates (table 2.4) resulting in an 'undershooting" of the nominal exchange rate and real appreciation of the inti (table 2.6). "Asset shock' on the reaO exchange rate are ftequedy associated with credible stabilization plans when exchange rate flotation is permitted. To the extent that real demand for money reamers and individuals are holding close to their desired levels of intis, the exchange rate reams to i ts (highest) level of long-tenn equilibrium.  When economic agents fell short of intis, they restructured their assets by buying dollars. This allowed them to take advantage of important differences between foreign and domestic interest rates resulting from the reduction of real money in the country and to maintain a relatively "normal" level of goods and services transactions. This led to dollarization of the economy in the short tend, as foreign currency flowed into Peru. With its growing abundance, the dollar temporarily dropped in .price, and the exchange rate adjusted, regardless of price flexibility (figure 2.4). However, a gradual depreciation should occur ceteris paribus, after a period of real appreciation, once financial stock adjustments have ended and the tendency reverses as the real supply of intis increases. Table 2.6 Exchange rates and prim, monthly averages, January 1990 to February 1991 Uuly =loo) - - - - m 1 Jm 88.7 11.9 39.4 52.8 2 5.7 29.8 Feb 77.9 95.2 34.9 42.6 7.7 31.7 30. 5 Mu 94.9 96.3 42.9 43.4 35.7 33.7 32 .6 85.7 103.9 39.3 47.6 4 7.4 39.4 373 &Y 101.3 110.0 473 513 393 353 32. 8 Juo 153.6 110.5 72.7 66.5 87.2 44.5 42 .6 Jul 118.0 151.9 54.7 70.4 6S.5 59.7 63. 2 87.0 76.8 42.4 37.4 158.5 ,422.4 391.0 Scp n.o 115 .8 46.5 45.8 37.4 24.1 13.8 Oa 1 86.0 43.0 43.9 2.1 4.8 9 .6 NW 82.3 83.2 40.5 413 -13 27 5.9 Dtc 73.9 7 9.2 38.0 10.7 23.2 28.5 23.7 1990 94.5 102.3 45.1 48.7 46.0 63.0 60.0 Jm 68.0 70.6 333 34.5 0.4 12.0 178 . Fell 9.4 Notc:Tbo~oftbod~pr~et~iDdaroftbom~~prdo.vcrrgoforthepaiodoftheCoofnl~ of Pau,bued on the bulra of curreaciar. An iocrsuo ia the i n h i a d i i red ~ ~ , v b i4 cbaaro l e h d i i ral mpp~iui00. 5.'Ibe&~of~pauVi.n~b~ubypainil.tioocuuodd~toaodo.~~mic~nvrvedo oftbe bti tbii dollariutiw wu incrmd by tbe mwemart of &ortarm lpecuttiveapitrl. - 15- Eigure 2.4 Theoretical behavior of the d exchange rate after a monetary shock Case 1 - - - a - E P P wp0tmt1-1 w u t m t mmmu~ngImtamm- fw*lgn erchrqm rat. a d c r r v I r e a t M J u t m n t r l t h m1C.v p r l ~ . . . --*fmotl W of tk r a t rate QCD~O . wlth nq.tl*. mmrr menu mt t*. b t Tlu! Cue2 a' - -1-1 MJY - w -1- -10 DTla m- a b4R V l a r ar(rrp.rmcm ma-. -1 = '\ : = rme rlth t -t#r -mt=Y.rr(*.tF. t- TIY  Simultaneously, other factors influencing the exchange rate were at work; the most i m p o m t among them being trade and price liberalization, which worked in opposite directions. Tariff reductions and the elimination of quantitative restrictions caused the real equilibrium exchange rate to depreciate, while the second factor (amounting to an increase in the price of nontradables) caused it to appreciate. Thus the net impact is ambiguous and difficult to estimate, especially during a period of important monetary and real adjustments. The expected net effect within a framework of plausible cross elasticities assumptions is an increase in the real exchange rate, defined as the relation between tradable goods prices and domestic goods prices. A theoretical exercise, presented in detail in appendix II, suggests that the expected effect of lowered tariffs and the elimination of nontariff barriers in Peru would be a depreciation of the real exchange rate of about 10 percent, a low figure.6 Working in the opposite direction is the increase in the dollar price of public sector goods and services such as diesel fuel, gasoline, and electricity and the dramatic increase in their relative price since August 1990; they have caused large-scale appreciation. Intuitively, it would seem that if the inti price of goods offered by the state rises much more than the inti price of the dollar or of exportable and importable goods, then the real exchange rate will fall. Another way to analyze that effect is to note that the sharp rise in the price of public goods and services also generated a transfer from the private to the public sector, where the demand for non-tradable goods is more intense. That transfer puts upward pressure on the prices of those goods and, consequently, downward pressure on the real exchange rate. Using the same type of economic model mentioned in the preceding paragraph (see description and estimate in appendix 2), it would seem that the real exchange rate has appreciated around 45 percent. Therefore, that effect would explain a high percentage of the drop in the real exchange rate observed between August, 1990 and February, 1991 (51 percent). However, the goods offered by state companies represent only a part of all nontradable goods so using only public goods and services exaggerates the impact of non-tradable goods prices on the real exchange rate.' The opening of the economy to foreign trade is likely to be accompanied by overshooting in domestic prices, which means that either other domestic prices will adjust (downward) or the overshooting will be only temporary, because of market distortions and reduced sales. Further, 7.L fact, if wrga w u c used u r proxy for the price of wntmdabk goods, 3 could be concluded that acmtdabk price movanaits tended to dqmehtc the arclunge rrte.  the low prices of public utilities before the adjustments described were accompanied by a proportional fiscal deficit, since costs of the utilities had to be covered by transfers. When prices were adjusted upward, they may have lowered the real exchange rate. But the corresponding elimination of the fiscal deficit (or part of it) would push the real exchange rate upward. Consequently, the net impact of the readjustment of public utility prices is certainly less than it would otherwise be, and perhaps even ambiguous. Finally, the reduced size of the government-an objective of the adjustment program-will tend to increase the real exchange rate. Thus the model overvalues the impact of the rise in public service prices even in the short run. It is often argued that the appreciation can be blamed on increased supplies of dollars from coca paste exports. But this is an unlikely explanation, for two reasons. First, those exports are not a new phenomenon and, therefore, have been affecting the real exchange rate for some time; only the greater value of those exports in 1990 could have the effect of appreciating the rate further, and the relevant information to estimate the magnitude of this is not available. Second, even if the value of cocaine exports for 1990 were greater than that of other years, its possible impact on the real exchange rate will only be proportional to the amount actually returned to Peru, an unknown, but certainly low amount. As demonstrated in figure 2.5 and table 2.6 the real exchange rate appreciated after August 1990, due to the strength of the inti after the monetary contraction of August 3 and the subsequent flotation of the exchange rate. In terms of export costs, the real appreciation is less than indicated in figure 2.5 and in column 2 of table 2.6. Prospeds for the real exchange rate The evolution of the real exchange rate will depend of course on the behavior of its determinants. The gradual disappearance of the undershooting associated with the mon- shock augurs well for an increase in the real exchange rate. The gradual accumulation of real monetary balances, accompanied by a drop in real interest rates, will reduce hreign aurency flows, which would lead to fewer dollars being held and an increase in the price of the dollar. Furthennore, as the ewaomy recovers, demand for imports will increase, putting upward pressure on the real exchange rate. An improvement in the fiscal situation could allow the g o v m e n t to reduce delays in making foreign debt payments, also putting upward pressure on the real exchange rate. A new reduction of tariffs and of other remaining controls would have a similar effect.  Figure 2.5 The real exchange rate in Peru 80 c19e5 - ~oul 70 60 50 40 90 20 10 0 Jan Fab Mar Apr May Jun Jul Aug Sop W Nov Dee Jan I lea I1Wll R R Expert Rat. RER Frw Fbt. N a Tho index of the red exchap nts u tbe index of the multilrterrl cxchmge rate. 8 v m p d for the paiod, crlculvod by tbo Cmtnl~cBMkofPauonthebaiuofmbadcstofcurreacia.Aabcrouemtboindexindiatared~urtiw,mdsasuc indiredrpprociuiw. The question is whether, l i e tariff reforms, adjustment of public prices has a permanent effect on the real exchange rate, that is, whether the prices achieved to date in the domestic goods m a rket are the nondistorted equilibrium prices. That some prices have reached levels very nnrch higher than those of their international substitutck suggests that they are not equilibrium prices, but that they are probably above their long-term levels. In addition, the steep drop in real salaries suggests that the domestic labor market is not at its long-term equilibrium, thus distorting current outcomes. This means that the real exchange rate will tend to devalue to the extent that those prices evolve toward their long-term equilibrium levels. If the authorities were to consider swaps of foreign debt for domestic capital, the p r i v a t i o n of public companies would also stimulate a real devaluation of the inti. Only an increase in income from drug traffic and repatriation of a substantial and stable part of  that income could imperil the observed tendency toward an increase in the real exchange rate over the medium term. Trying to fix the real exchange rate at a higher than equilibrium level is fruitless. Given that domestic prices adjust rapidly in a context of hyperinflation, an attempt to provoke a real devaluation not supported by fundamental conditions would tend to return hyperinflation to center stage. At best, 'only short-term real devaluations can be achieved. The key point in any devaluation policy is to maintain the goals for growth of the monetary base through a contraction of domestic credit, counteracting increases in the monetary base arising from the purchase of dollars. Another important parameter for the exchange rate is the evolution of the value of the dollar relative to other currencies. A 'strong' inti is, in part, the result of a 'weak' dollar. Appreciation of the dollar in terms of other important currencies allows for a greater depreciation of the inti. The crossed rates indicate that: w x SPBs dollars SDRs dollars, so that depreciation relative to the dollar is possible without changes in the inti price of the SDR if the price of the dollar in SDRs rises. Otherwise, the 'low' price of the dollar in intis reflects a low dollar price in SDRs. In this way, appreciation of the dollar relative to other cutrencies eases the Peruvian situation somewhat, given that dollar trading is M y Gtrong in Pea. Other policies that affect competitive capacity Trade openness achieved by exposing productive sectors to greater foreign competition forces them to adjust their production processes. Current rigidities in factor markets imply high domestic costs for producers and limit their capacity to adapt to changing conditions. Reducing those rigidities yjll facilitate factor mobility from the affected sectors to potentially more profitable areas. Some important advances have already been made toward liberalization in fibut important areas: the labor market, the exchange market, the money market, and the land market. The government may also need to adopt complementary measures to reduce adjustment costs and to minimhe its intervention in different markets. In the lpbor mar&, one of the principal obstacles to the opemess process was the labor stability law. The law's impact on the labor market has been considerably reduced by the unequivocal -20- definition (in March 1991) of the reasons for which a worker can be fired, thus providing greater clarity in the work relationship for contracting parties. Greater uniformity in pension and retirement laws was also introduced in March 1991, and workers can now receive their payment in a financial instrument and financial institution of their choice. The new labor regulations have introduced greater flexibility into the labor market. The land marker has been significantly liberalized since August 1991 by a measure making agricultural property more secure and freely tradable. Its objective is to encourage efficient exploitation of the country's scarce arable land so as to increase food production and agroindustrial processing activity. The measure facilitates the resource reallocation induced by trade openness and so promotes capital investment and the appropriate use of technology. The measure has the legal character of "extraordinary" until the General Agrarian Law is passed. It marks an important change in Peruvian agriculture because it transforms the system of property ownership and land management. Previously, land owners with titles granted under the land reform were not able to dispose of their land freely. Now all landowners can sell, rent, mortgage, divide, encumber or otherwise dispose of their land as permitted by law. Societies of capital such as open corporations and limited partnerships can now own and manage farm land as well. In May 1991, cooperative agricultural associations and companies were authorized to freely choose their business form, including forms of ownership that made farmers true landowners. Important changes are planned for the financial markg. A new bank law is designed to promote mechanisms for bank competition that will guarantee lower financial costs. Measures of financial liberalization and improvements in financial services have contributed to a reduction of those costs.  Chapter 3. Liberalization of Imports This chapter describes the principal measures taken between August 1990 and May 1991 to liberalize imports. It also briefly reviews recent efforts to revive the Andean Pact. Tariff regulations Recent trade liberalization pressures have substantially reduced discrimination in the treatment of productive sectors. In August 1990, tariff restructuring was begun as a first step toward reform. The maximum tariff was reduced from 84 percent to 50 percent (table 3.1). The minimum tariff was lowered from 15 percent to 10 percent and was applied to all goods with tariffs previously lower than 11 percent and to those that had been free of tariffs under 39 of the 139 exemption schemes. The measures reduced the number of tariff rates from 38 to 22 and lowered the simple average tariff from 45 percent to 38 percent. Table 3.l Number o f rates md average tuilfr, April 1990 to Murh 1991 - April 1990 lsusurt 1990 Much 1991 1990 In September 1990, a three-tier tariff structure was approved with ad valorem (cif) rates of 15 percent, 25 percent, and 50 percent and a temporary surcharge of 10 percent on imports at the 25 percent and 50 percent rates. Tbe simple average tariff dropped from 38 percent to 26 percent (table 3.2). Tbe 15 percent rate was applied to food and medicine (finished and inputs), sted products, capital goods for agriculture, medical instruments, machiery for the textile and metal-working industries, disassembled completely kaockeddown kits for automobiles aad transport equipment, parts aad pieces for assembly industry, tires, and inputs in general. Tbe 25 percent rate was applied mainly to capital goods, such as construction matesial, machinery fiK industry and commerce, r e p b e a t parts, boats, buses, trucks aad light trucks; some inputs aad amsumer goods, such as wood, papa  (except for newsprint), woven goods, sacks, automobile spare parts, and buttons; and capital goods, such as finishing equipment for construction. Also in September nearly all tariff exemptions were eliminated. These exemptions, which had covered about 50 percent of imports, reduced tariff income from 41 percent of imports to 21 percent-a substantial fiscal cost (table 3.3). .Table 3.2 Number of tariff items according to rate and code, September 1990 JoriffRarr Awmge nom&lal Table 3 3 Fisal eort of import duty aemptiooa (weishtcd bport Cuitl) Exceptions included donations, commit mu^ under immtional agreemeats and treaties, special exemptions granted to universities and cultural antea, industrial duty free zones, and Central Bank imports of gold, silvva, currency, and printing plrrco. In Noveanber exemptions were restored fu goods brought i nto the country by Pemvhs returning from diplomatic and officials posts or tours of duty in internatonal organhtbns. To reduce pressures to restore tariff exemptions, which b e p to be felt almost immediy, exemptions were defined in a positive list in the legal text promulgated in March 1991. In December 1990, the Govemnent announced its inrention to unifv the tariff rate at 15 percent over a period of bur years (table 3.4). . Table 3.4 Calendar for tarilC reduction, July 1991 to January 1995 Tariff rate I%) Period IS 25 50 'July 1991 15 23 45 January. 1992 15 23 40 July. 1992 15 22 35 January, 1993 15 21 30 July, 1993 15 20 25 January. 1994 15 20 25 July. 1994 15 18 18 January, 19% 15 15 15 Source: Supreme lkcrst No.322-mEF. This gradual approach permitted businesses receiving a high degree of protection to postpone adjustment and gave interest groups time to pressure for abandonment of the proposed goals and calendar. Indeed, by the end of January 1991, the first signs of an incipient reversal appeared in the form of a 5 percent ad valorem tariff for imports of steel-working inputs covered by 25 tariff lines. The tariff rate of 5 percent generated arbitrarily high levels of effective protection for the five steel companies-levels that would naturally be higher the lower the percentage of value added. Belief in the government's ability to stand firm was shaken. Producers and investors had reason to believe that the government would continue to bestow dis-ry tariff treatment on certain sectors. With the loss of credibility came a reluctance of economic agents to invest and to reallocate productive resources to sectors relatively more favored by liberalization objectives. To demonstrate decisive commitment to its policy of economic openness, the government announced in March 1991 that it was accelerating the dismantling of nontarB restrictions. Also, the 25 percent tariff rate was reduced to 15 percent and the 50 percent rate was e l i i e d , with tariffb on items in that category reduced to 15 percent or 25 percent. The 15 percent rate was applied to inputs, capital goods, and consumer goods for which the country does not possess a comparative advantage, such as automobiles. The result was a basically uniform tariff structure, with 81.5 percent of items at 15 percent (table 3.5). and a drop in the average tariff from 26 percent to 17 percent and in the variability coefficient from 50 percent to 23 percent.  Nontariff barriers In past trade liberalization episodes, the persistence of nontariff restrictions had impeded the efficient allocation of resources and drastically reduced fiscal income. These restrictions consisted primarily of import prohibitions and restrictions such as prior import licenses. Table 3.6 shows the number of tariff lines affected by these restrictions since 1981. These restrictions, together with the xystem of multiple exchange rates and exchange rate restrictions, reduced the welfare of the majority of Peruvians, while encouraging illicit methods of accumulating wealth and channeling profits to specific companies. Tabk 3.5 T M strpetun, 1989 and 1991 (P=-Wa) Previous situation New structure Goodr/ Source: Romid (1991). Table 3.6 Formal nontariff restrictions to imp*, by number of tariff lines, selected periods since 1981 Fret 5.008 5,136 4,757 3,259 0 0 4.192 4.721 5.256 5,269 Restricted 112 118 350 1.553 4.715 4,724 535 6 0 0 Prohibited 7 8 8 525 539 539 539 539 13 0 Temporarily prohibited and other 80 0 188 0 9 3 3 3 3 0 Total 5.37 5,262 5.303 5337 5,263 5,266 5.269 5.269 5,269 5,269 Source: Mission eaimat#. Import prohibitions and restnstnCtl'ons In the decade before August 1990, nontariff barriers constituted the main instrument of protection in Peru. In 1987 and 1988 these barriers were so pervasive that not a single tariff item could be imported freely-10 percent of these items were prohibited and 90 percent were subject to prior import licensing. During 1989, import licensing was relaxed so that by the end of the year import licensing affect only about 10 percent of tariff code provisions. Despite this partial liberalization, it is estimated (Abusada, July 1990) that at the end of 1989, 285 of the previous 539 prohibitions remained, protecting more than 40 percent of the production of 19 industries at the fourdigit level (Uniform Interdona1 Industrial Classification) (table 3.7), production representing about 22 percent of GDP. The fiscal cost of those prohibitions was about 1.4 percent of GDP. To eliminate the distortions created by import prohibitions, the new goveament first reduced the number of prohibited tariff items to 13 in August 1990, and then elimhkd all of them in September. In August, it also temporarily suspended all prior license requknents-whish covered 5040 percent imports-until the end of December 1990. The suspension w as extended after December, and license requirements were finally abolished in March, together with other registers, consular approvals, permits, and adminisaaive requirements (Supreme Decree 060-91-EF). The Decree of Marc i h 1991 achieved two objectives. Frs t,it put an end to all uncexbhty with respect to the government's intentions to permanently eliminate licensing. Second, it ended the de fado discrimination against imports fromother Andean Paa amatries that had arisen when prior Table 3.7 Principal rectors witb more than 40 percent of their production covered by import prohibitions, up to December 1989 IpU No. of Sof So f Simple Secwr knu VBP VIIC avrmge rorid (sb) 3112 Milk products 11 1.85 41 91 3113 Caned fruit 48 0.2 6 83 % 3114 Procucedfuh 14 0.5 6 90 108 3117 mproduar 3 0.65 100 % 3119 Cbocolrtcr 11 0.5 8 69 101 3131 Alwholicbcvctaga 13 0.2 6 88 110 3133 Bea 1 0.9 9 98 110 3211 Thread 47 6. 45 44 85 3212 W w a tcxtila 29 1.51 94 103 3213 Ksiasdtextila 22 1.51 68 106 3220 Clothing 29 3.1 7 95 110 3240 Footwcu 4 0.70 97 110 3320 Woodea furniture 8 1. 25 98 107 3523 S u q and cormctia 2 0.7 1 67 78 3559 Rubbcxproduccl 2 0.28 93 110 3832 Rdiaodtdwiri 3 0.28 62 86 3833 A p p l i 6 0.1 4 57 110 3843 Vohicl~ u I.n 53 91 3#)1 Jewelry 7 0.89 54 107 Subtow 285 1s All prohibii 539 23 sowcc: Abuod. (July 1990). licease rq&emen$ wae mspended, but reghmbn, the cauP5espart fix Andean P act imports, was not. Any attempt to renew the import registers ought to be resisted, evea if the argument is couched in terms of 'automatic' registers 'for statistical purposes only' to allow m v e agenrs to 'supenrise' imports to prevent dumping or emsure import quality. Any register of this type would slow the import process, impose additional costs, and iuhuduce a discretionary elemeat. Before the reform of March 1991, there were nearly twenty forms of regulation or similar measures impeding imports-sanitary regulations, technical norms and quality certificates,the no competition dedsion, reference prices, state monopolies, and diierse forms of bureaucratic p a p m r k . While their original intent may have been otbenvise, many of these measures had been bent to protectionist ends (Appendix III lists these restrictions).  In February 1991, about 300 products-both domestic and imported-were subject to obligatory national technical norms. These had been defined by a committee of the Industrial Technology and Technical Norms Research Institute (ITINTEC) representing producers, consumers, and technicians, and expressed as resolutions of the Ministry for Industry. Most related to the automotive, alcoholic beverage, cement, health, cooking oil, and security sectors. For imports, the process worked as follows: potential importers first would have to discover if norms covered the products they wished to import. When imports arrived in the country, the importer would prepare a declaration covering the size of the shipment, product type, and the technical norms that must be met. ITINTEC personnel would then go to the port to examine a sample of the shipment and verify that it met the norms. ITINTEC then had ten days in which to certify that the shipment met the norms or to issue a provisional certificate allowing the importer to store the goods in its warehouses, but not to sell them. Technical norms constitute import protection if they are applied differently to domestic and imported goods. Importers complained about the incoherence or protectionist slant of many of these norms, arguing that norms often favored the products of large national companies to the detriment of imports. For example, the Peruvian standard for hydraulic pumps requires adaptation for use at very high altitudes, even for those to be used on the coast; machetes must be of a certain dimension, even though world norms allow for a considerable variety of shapes. Inconsistent application of the norms is another source of dissatisfaction. While technical norms were the same for both national and imported goods, to the extent that procedures were applied more frequently or norms were enforced more strictly for imports than for domestic products, quality verification constituted a nontariff barrier to imports. In March 1991, the government completely overhauled the system of obligatory technical norms (Supreme Decree 006-91-ICI'I). Old norms were no longer mandatory. A new set of norms was developed, limited to products affecting life, health, and security. The certificate of compliance can be obtained from lTINTEC or from other national or foreign institutions approved by the M i fot Industry. It is not clear whether each shipment must be certified as before, or whether one-time certification is sufficient as long as a product's inputs and design remain the same. Requiring only one certification for a product (domestic or imported) and elimhathg the requirement for test samples from each shipment would reduce asymmetry in regulatory obligations for domestic and imported goods, and the distortions this causes.  The application of sanitary regulations and puality certification standards was also simplified in March 1991. Both national and foreign private institutions were allowed to assume responsibility for the required controls. Sanitary regulations for imports and exports of vegetable products were . modified so that sanitary permits for imports did not constitute import protection. Meat products regulations were similarly modified. Gold trade was also freed of trade-impeding quality regulations. .. The March on agricultural imports. Supreme Decree 1991 reform also eliminated State monopolie~ 060-91-EF further eliminated all nontariff restrictions, licenses, decisions, prior and consular authorizations, import registers, registers of importers and exporters, previous approvals, and any other condition affecting the importation or exportation of goods. The only exceptions were prohibitions established in the Unique Text of Export Prohibitions and restrictions derived fiom exchange rate norms, or established to protect the health of the population, to defend the cultural heritage and the environment, or to ensure internal order and security. In April, it was clearly announced that since all nontariff barriers to imports and exports were being eliminated, the foreign trade functions of the National Boards of Rice, Coffee, Wheat, and Milk Products and of the Foreign Trade of Hydrobiological Products were to be dissolved. Some pennit requirements were changed so they could be m a by permits issued in the country of origin of the import. In May, fbreign trade in the agricultural sector was further deregulated, with agricultural a - n pesticides and similar substances and v products requiring simply the presentation of a sworn statement to the M i for Agriculture. The treatment of agricultural imports8 A special regime covers agricultural imports,both to shelter domestic markets from subsidized prices of these products in world markets and to prevent large fluctuations in international prices'fiom affecting long-term agricultural production decisions. In March 1991, a compensatory surcharge per metric ton was applied to the cif value of certain imported food products and inputs to moderate the domestic impact of significant fluctuations in international prices (table 3.8)? The surcharge was applied to powdered skim milk, powdered whole milk, anhydride milk tat.wheat, hard yellow corn, hulled rice, sorghum, wheat flour, unrefined sugar, refined sugar, and.food pastes. The surcharges 9.Thc Mmirey for Agriculturejuatifta thir tool botb for that purpose and u M antidumping norm. Thc mmo meuurc ia riot m c a w d y approprimto for both of tbora objectives. affected all impom, even those from countries with which Peru has trade agreements granting tariff advantages. Table 3.8. Imported food products and inputs with compensatory surcharge, March 1991 W.S. dollars per metric ton) Conpemarory ~ndyct surcharge Skim p o w d d milk Whole powdered milk Anhydride milk fit Whut H u d yellow oom Polirhed lice sorghum Whut flour Unrefined sugar Refined sugar Food- In May 1991, the compensatory surcharge was replaced by w i f i c on a graduated scale, expressed in U.S. dollars per metric ton, in customs tables based on the product's fob price (the weekly average of the closing fob spot quotations) in the international market at the time the merchandise was shipped. The lower the international price, the higher the duty and, conversely, the higher the international price, the lower the duty, until the duty is 0 if the price rises above a level specified in the tables. Table 3.9 summarizes those customs tables, indicating the lowest fob reference price for each product with its corresponding specific duty. The specific duties are applied to imports from all countries, including those with which Peru has trade agreements granting tariff advautages. Imports of used goods In March 1991, imports of used goods, such as machinery, equipment for industrial use, motors, automobile mechanical assemblies and subassemblies, vehicles for cargo transport and mass transit, and passenger vehicles, were authorized. To resolve the value estimahn problem for duty and other tax payments posed by imports of that type of goods, linear depreciation of the value of similar new merchandise is applied. This measure permits maximum use of available capital f ix investment and broadens tecbnologic.1 alternatives and the selection of relatively more labor-intensive technologies.  Tabk 3.9. Specirk. duties applied to imported food products and inputs as of M a y 1991 (U.S. doll& per metric ton) Minimum fib Fob prlce Mmirmrrn Mininuan SPMC T i * rdermcc abow d c h pdJc W a c huyh W a c Pmdvet price rhrhuy huy huy umo/ T i huy LO torib 6) Z fi) Wheat 61 106 90 1 47.5-0.7 15 62.5-15.7 *our M d - food P- 61 106 118 2 93.4-1.5 15 108.4-165 Rict 100 215 138 1 38.04.5 15 53.G15.5 Hard yellow CO~O 30 98 82 1 173.3-1.0 15 188.3-16.0 SUP 80 256 212 1 165.04.4 15 180.Gl5.4 Whole powded lnilk -949 17W1749 959 55 6.5-1.013.3-3.2 15 215-16.Ot18.3-18.2 -powdaod mil& - 9 1-1449 960 55 60.0-47.914.1-3.9 15 75.0.62.9t19.1-18.9 Anhydride milk t.t 1000-1049 205@2099 1217 11 21.7-16.010.60.5 15 36.7-31.Otl5.615J Source: Supmas Docroe No.001691 AO. Integration and the Andean Pact After a lengthy h i , the Andean Group has received renewed aaention as a result of trade lberalization policies in member cowries that are providing new impetus to braregional openness. At the Fourth Andean Presideaid Council in La Paz in Nwember 1990, members agreed to accelerate intraregional tariff reduction so that all tariffk d d be elhinted by December 31, 1991. They also agreed to eliminate the a d m h h d trade regime, which established product quantities through bilateral negotiations; the reserve register, which listed products included in regional industrial planning programs; and the exemptions lists, which contained the tariff lines not included in the process of integration-all by the end of 1991. The common minimum external tariff was reduced from an average of 28.5 percent to 17.9 percent, with a minimum of 0 percent aad a maximum of 50 percent, implying a significant reduction of protection at the regional level. Dehition of the common minimum external tatiff was moved up to the end of 1991, with the objective of achieving the Andean Tariff Union by the end of 1995. Colombii Peru, and Venezuela will implement the common external tariff toward the end of 1993, Bolivia and Ecuador, by the end of 1995. Most mcmba countries will eliminate their exemption l i by the middle of 1991. Other lists are to be eliminated by the end of 1995.  It is premature to try to evaluate the impact of these decisions on intraregional trade. Intraregional trade for 1990 grew an estimated 24 percent (more than US$1,300 million) and represented 4.4 percent of total Andean region sales for that year-8 percent if fuel exports are excluded. Currently, two-thirds of the items in the tariff code are not traded freely because of the enduring effects of exemptions lists, reserve registers, and the difficulties arising from violations of agreements, 'especially those related to the common minimum external tariff. The decision to implement free trade among member countries has highlighted the urgency of achieving harmonization of national expon incentive regimens and greater harmony in the monetary, fiscal, and exchange rate areas. The planned changes in regional institutional structures are unlikely to bring significant benefits to Peru. Greater benefits are promised from a continuation and broadening of its policy of general openness, and eventually, from a free trade agreement with the United States, within the framework of the Enterprise for the Americas Initiative.  Chapter 4. Exports After a brief period of liberalization between 1979 and 1982 (see Nogues 1991), Peru's trade policy became progressively more interventionist during the 1980s. Exports were affected not only by greater antiexport bias in trade policy but also by growing discrimination among export activities. Much of that discrimination was eliminated, at least on paper, through a series of fundamental changes in trade policy introduced by the Fujimori government between August 1990 and March 1991. Any continuing antiexport bias seems to derive from the remaining customs tariffs (relatively low) and taxes on exports (most of them low). The new policy should reverse the declining trend in exports of the 1980s. Growth in nontraditional manufactured exports is expected to compensate for the contractionary effects of import liberalization on industrial activity. What is now likely to slow the growth of exports, particularly traditional exports, is domestic regulations (especially labor laws), the persistence of public ownership of property (especially in mining and fishing), the ad hoc fiscal discrimination faced by large exporters, and inadequate physical infrastructure. Export behavior Export behavior in the 1980s was poor, as was economic performance generally. From a peak of US$4 billion, or nearly a quarter of GDP h 1980 (when international prices of traditional exports were high and many export incentives were in place), total exports fell to less than one-tenth of of GDP (Table 4.1 and annex table A4.1). Even during the preceding period of export growth, from 1960 to 1980, Peruvian exports grew only sev-ld (in m m t US dollars), whiie exports grew 23 times in the global economy and 13 times in Latin America (National Society of Exporters, Annual 1990). Both traditional and nontraditional exports were sluggish, although nontraditional exports behaved slightly better. The share of traditional exports dropped from 78 percent of total exports in 1980 to 70 percent in 1990, a surprisingly small drop considering that nontraditional export prices, which closely reflect world manufacturing price trends, probably evolved more favorably than traditional export prices. State companies account for a large part of exports-more than half in 1990--especially in mining and fishing. Their share was probably smaller during the preceding decade, howevet.  Tabk 4.1. Export evolution: 1970.1990 (fob, millions of U.S. dollars) Tmdiriond Nm- T d To& fipont~ra 6bbl rrodidmd rob) imp= prmprrrmtof GDP Ibe largest unreported export is coca, both processed and unprocessed. Coca exports are estimated at about USS1 billion annually (equivalent to about 30 percent of legal exports); a large part of those proceeds are not returned to the country. Export promotion policies of the 1980s Peruvian governments instituted a series of export promotion activities during the 1980s. although the Garcia administration began to close off the economy after 1985. Multipk exchange rmcs Multiple exchange rates were established in 1978, allowing for discrhhatory treatment of productive activities. Beginning in 1985,through various cornplea and shifting mechanisms, exports generally enjoyed higher exchange rates than imports,and among exports, nontraditional items received the highest exchange rates.  With the imposition of measures closing off the economy in 1985, real exchange rates began to appreciate rapidly. Under-reporting of export values, as well as the existence of multiple exchange rates in 1987-89, may have somewhat cushioned the effects on exports of the decline in the real exchange rate. With the foreign exchange premium on the black market rising rapidly after 1985, reaching 240 percent in 1988, exporters may have taken advantage of the black market exchange rate 'by underreporting their export income. That would mean that the exchange rate regimen was offering better incentives than the official rate indicated. w o n subsidies Nontraditional exports enjoyed a variety of subsidies during the 1980s. The most important of these was the Certificate of Tax Return for Nontraditional Exports (CERTEX), created in 1976 and administered by the Trade Institute (ICE). The CERTEX mechanism was a freely negotiable tax certificate calculated as a percentage of fob export values. Its original purpose was to compensate export producers for the cost of taxes on imported inputs, but in practice the scheme became an indirect subsidy. From July 1988 to August 1990, the base subsidy rate varied from 15percent to 30 percent, according to the amount of value added and labor intensity; products and production regions that fulfilled their requirements received additional subsidies (table 4.2). Table 4.2. Evolutbn ofCERTEX rates, sdected periods 1981-90 &' A niogle CERTM mo of 35 pacsot mr rpplkd to aunuretprodpodr a d tho llrt of eligible produrn mr br#deood. ! Soufw: Aburd. (1990). Exports able to take advantage of the CERTEX scheme (around two-thii of nontraditional exports) enjoyed an average subsidy of about 25 percent in 1980. Tbis implies a subsidy for nontraditional exports as a group that averaged about 17 percent (table 4.3). These high subsidies ran afoul of countending duty laws in Peru's major export marketr. In 1982, the United States imposed compensatory duties of 40 percent on three categories of Peruvian textiles; as a result, Peru's exports  in those categories were suspended for two years until the government ended the CERTEX scheme. Exports of lemon essence oils, steel bars, and fresh flowers to Chile and Germany also experienced similar problems. Table 4 3 . CERTEX: Subsidies and nontraditional exports, 198890 1980 I981 1982 1983 1984 1985 1986 I987 ISW I989 IP90 Nontndiod =PO* (Us$) 845 701 762 565 726 714 645 709 747 980 974 CER'ZEX (US0fob) 1PZ 129 114 7 1 121 126 110 121 210 151 128 CERTEXupacent of expo* 22.7 18.4 15.0 12.6 16.7 17.6 17.1 17.1 18.6 15.4 13.1 The second m ost importaut type of export subsidy was the Credit for Nontraditional Export Promotion (FENT), managed by the Central Bank. Eligibility for FENT was restricted in January 1990 and again in March,and the program w as suspended in June of that year. With an effective annual inkrest rate of 4 percent on FENT credit, credit subsidies to nontraditional exports amounted to an estimated 11percent off the maximum interest rate of 15 percent established by the Central Bank for foreign currency credit (Abusada, May 1990). In addition to FENT, there were other mechanisms for financing export credit for capital goods (FONEX), for insuring nontraditional exporters (SECREX), and for s u b s i i g export trade promotion (FOPEX). SECREX, established in 1978 to promote nomxditional exports, was a small d commercial risks. Ficed by private and state fund (USf10 million) used to awer political a banks and hsurmce companies, the fund was controlled largely by the private sector. SECREX operated closely with the government during the sharp drop in exports during 1983 to mahtah FENT and save many exporters from bankruptcy. SECREX also managed a state fund for political risk. T - on q o m Peru has commonly taxed traditional exports, although not as heavily as other Latin American countries. Direct tax rates were high during the second half of the 197Os, when prices were high, but fell rapidly ( h m an m e of 20 percent) early in the 1980s. Toward the cnd of that d&e, the tax on traditional exports averaged only 2 perant, with 4 percent on coffee, 3 percent on fish meal, sind 8 percent on minerals (table 4.4).  The principal jndirect taxg applied both to imports and domestic production. They included the general sales tax (IGV); a value added tax (currently at 14 percent); a selective consumption tax (fluctuates between 10 percent and 100 percent) applied to 127 tariff items; and various complementary taxes on energy. The tax burden was not distributed equitably among products, however. The narrow and stable taxpayer base meant that tax avoidance was widespread. In addition, reimbursement of indirect taxes on exports was spotty and difficult to arrange. Table 4.4. Tax on trrdiiml cxporb, 1980-90 Note: The foUowiag ~ u ewa r alao in effect: (1) a 1 pacent d v.lonm tax on the fob value of expo- ftom April 13,1981to November 30.1990 (yielded USSl m'h in 1 988- 89);(2) a 1 paocot d valorem tax from Wnury 27.1964 m November 3 0 . 19#) (yiddod U S 1 m i ll= in 1989); a d (3) a 1 0 percent tax on coffee expo- from Fctmuy 7,1& ( y i i USSl million i n 1990). Souroe: CentrrJ Bank of Pau. There was a mechanism for reimbursing indirect taxes paid domestically by exporters, but its - complexity - it required invoices d d i n g all transactions involving taxes effectively limited its use to only a few of the largest exporters. There was no indirect reimbursement mechanism. Expo- were unable to obtain reimbursement for the indirect taxes and duties on imports paid by suppliers of locally purchased inputs. The mining industry suffered from a double diivantage. It was formally excluded from most of these reimbursement schemes and its burden of indiirect taxation was particularly heavy because of its large fuel consumption needs (energy taxes are high). Indirect taxes for the industry have been estimated at between 10 percent and 40 percent of sales. Two temporary admission schemes were implemented for nontraditional exports during the 1980s to exempt them from paying duty and IGV on imported inputs. The first, established in 1982 and .. admmmed by the Mrdstry for Industry, w as very restrictive. Based on a negative list designed to protect domestic industry, it was available only to a few nontraditional m a n u b e d goods eqmts.  The second scheme, also administered by the Ministry for Industry, allowed exporters to import capital goods duty free, subject to certain requirements enumerated in an Export Contract. lo The two temporary admission schemes had only a limited impact because of the use of a negative list, bureaucratic requirements of the Ministry for Industry, and the inefficiency of the customs service. And exporters frequently preferred fiscal incentives to temporary admission, since the value of the imports that enjoyed temporary admission was to be discounted from the value of exports before calculating fiscal benefits for exports (particularly CERTEX). At the end of the 1980s;the Garcia government passed a law creating a framework for free trade zones. The first duty free zone (ZOTAC) was recently created in the south, near the border with Chile. It is not yet being used for production because of the lack of infrastructure. That zone is controversial because it has become a conduit for contraband from Chile. Antiexport policy bias in the 1980s Ever stronger nontariff protection of imports and an appreciating real exchange rate (after 1985) created a strong bias against exports. Traditional exports were further disadvantaged by taxes, although they were low during much of the decade. High subsidies for nontraditional exports, managed principally through CERTEX, were designed to compensate for the antiexport bias of the trade regime, but variability of the CERTEX rates and their general uncertainty led some exporters to view them as unexpected profit rather than as an offset to the bias of the trade regime. Orher aport barriers One objective of the export restrictions in the Industry Law was to divest primary products first to the domestic market and only then to export markets. In practice, the principal affected markets have been wtton, alpaca wool, and minerals. Thus the textile industry in Peru has evolved to supply the domestic market, with exports of cotton fiber only a marginal activity. In fact, Peru should be exporting the greater part of its production of high-price long, highquality cotton fiber and importing shorter-fiber cotton for domestic consumption. Peru continues to have several fonns of obligatory export d c a t i o n . The most serious cost in this area is incurred by fish exports, which must be certified by the state company CERPER; while the state intervention may be justified for sanitary reasons, there are problems with the system. For one 10. The comprny mua sxpolt at W 10 p a a m of im production a taro you paiod. at a nluo of at lart tho rmouot of swpoaded dutia withim Iive yam. thing, CERPER used to be a successful private certification company. For another, the quality of state certification has fallen off to the point that some exporters are required to provide a second certificate. Mineral and agricultural exports are also subject to quality and sanitary certification requirements. The poor condition of transportation infrastructure constitutes another serious obstacle to Peruvian exports. The road network is badly maintained and falling apart. The railroad system suffers the effects of inadequate maintenance and poor administration. The ports are also poorly administered and - dominated by worker monopolies. As a result, port costs are high one exporters' association estimates them to be seven times higher than in Chile. Many activities have been intempted by terrorism, with mining especially hard hit. Other distortions in the economy have also hampered exports, as well as production for the domestic market. Excessive regulation of labor markets has made it difficult for employers to adapt their labor force to changing market conditions. Interventions in the financial markets, including controlled interest rates, state ownership of banks, and weak supervision have led to distortions in the allocation of resources and to high mediation costs. The absence of clear rights to private property in the rural sector has systematically slowed agricultural development. Qlrrnges in -on policy since August 1990 Many of the reforms introduced alter August 1990 have affected exports. The exchange rate was unified and freed and exporters were no longer required to return their foreign exchange earnings to the Central Bank. The CERTEX subsidy was reduced in August and abolished in November. The Central Bank also formally eliminated FENT that month. In March 1991, several measures were taken to eliminate other barriers to exports. Requirements that domestic demand be satisfied before any goods could be exported were abolished. Sanitary regulations for horticultural imports and exports were simplified and other controls on exports, with the exception of those designed to preserve the country's biodiversity and cultural heritage, were eliminated. Reforms also reduced the effects of direct and indirect taxes on exports. A new temporary admission - system was introduced in November 1990. While all inputs could now be imported the negative list was eliminated- the system was still highly restrictive, since only manufactured goods exporters were eligible. To qualify for duty exemptions on imports, exporters had to provide bank guarantees for the value of their imports until their products were exported. The system, which was administered by the  Ministry for Industry on the basis of technical coefficients, was slow and complicated and was used by only a small number of large exporters. In March 1991, more radical changes were introduced in the treatment of indirect taxes (Supreme Decree 034-91.EF). The temporary admission scheme was replaced again. Import taxes were suspended on all intermediate inputs (including replacement parts, packaging, and consumer goods) for goods to be exported within 12 months. Bureaucratic requirements were reduced, and any exporter, including traditional exporters and indirect exporters, can use the system. It is administered by the customs service based on the good faith acceptance - subsequently verified by customs of - exporters' declarations of imports and exports and of the input products and waste coefficients used. Banlr guarantees for the full value of the import taxes are still required. Also, in March the government introduced a system to ease the burden of indirect taxes for non- traditional exporters (Supreme Deme 052-91-EF). Exporters are granted a (transfernable) tax credit against import duties, sales taxes (IGV), and employer participation in FONAVI (a housing tax paid on inputs). The system, administered by the General Diectorate of the Public Treasury, offers an alternative to the temporary admission system. The reimbursement system, conceived of as a temporary measure, has not been extended to traditional exports because of budgetary restrictions. The new government also added a provisional direct tax of 10 percent on exports as an extraordinary contribution to the Social Emergency Program (Supreme Decree 23240-EF). Thgt tax and various lesser taxes on traditional exports were eliminated in November 1990, replaced by a new 5 percent tax on most traditional atports a d a 10 percmt tax on exports of large mining companies. The provisional tax b r mining exports was changed again in April 1991. Exports of large or medium-size copper, lead, zinc, and silver mining companies were taxed at 10 percent, 5 percent, or 0 percent, depending on where their products' prices fall within a range fixed by the governmeat (Supreme Decree 084.91.EF; Supreme Deme 101.91-EF). SmaIl mines and exports of other metals are exempt. In practice, the tax has functioned only as a tax on the two largest copper exporten. Anna Tabk A 41 F e w campositbn of aporb, 1970.90 . (fob, miUioa d U.S. dollars)  Chapter 5. Agriculture The agriculture sector merits special treatment, both because of its socioeconomic importance and because of the legacy of discriminatory treatment that endures to some extent even under trade liberalization. The policy of cheap food for the urban population - a policy that consistently accompanies import substitution policies in Latin America - has created strong distortions and inertia in agriculture. Characteristics and wolution of the agricultural sector Three features strongly influence the agricultural sector in Peru: the low ratio of arable land to population, wide agroecological diversity and large interzonal differences in production costs. Arable land Der ca~i@. Agricultural land resources are scarce in Peru, so their opportunity cost is - high. Of the 3.7 million hectares of arable land in Peru 1.3 million of them irrigated only- slightly more than 3 million hectares are actively cultivated. Moreover, important irrigated areas have been lost to agriculture as a result of salinization brought about by overirrigation because of poor water pricing policies. Thus Peru has one of the lowest ratios in the world of cultivated hectares to population - around 0.15. Another 5 million hectares could be brought under cultivation, but at great cost (IICA 1990a). Efforts in recent decades to increase the cultivated area through colonization and imgation demonstrate this: the cost has averaged more than US$5,000 per hectare under colonization schemes, and US$10,000 to US$12,000 per hectare under large irrigation projects (not counting production consolidation cost), and US$2,700 to US$3,050 per hectare (IICA 1990a) under medium and small irrigation projects. By comparison, excellent agricultural land in Argentina, Uruguay, and Southern Brazil costs US$2,000 to US$3,000 per hectare. Agroecolcgical diversity. Peru has a wide diversity of climate, altitudes, soil, and rainfall and consequently an enormous variety in its agroecological land. That diversity implies a potential to produce a great variety of agricultural products year-round, including the potential to produce fresh fruits and vegetables during the nonproductive seasons in the Northern Hemisphere. This potential further increases the already higb opportunity cost of land. differ- This great agroecological variety also means that in uroduction ~QS~S. production and transportation costs may vary substantially h m one region to another.  Principal products Peru's principal agricultural products are shown in table 5.1. Coffee and cotton are the major exportables. Part of sugar production is exported in order to take advantage of importing country quotas; some sugar is also imported - the international price in the nonquota segment of the market is low. Nontraditional agricultural products (asparagus, mangos, garlic) are also exported; their value in .I989 was US$llS million (IICA 1990). Of exportable products over the last decade, production of cotton and sugar cane has declined, while that of coffee and nontraditional agricultural production has grown. Wheat, rice, yellow corn, beef, and milk products are the major importables. Over the past two decades production of wheat has declined, while that of rice and yellow corn increased significantly (table 5.1). Cheap food policies contributed to the decline of wheat production, whiie rice and yellow corn production benefitted from new high-yielding varieties and from production subsidies designed to promote production in the jungle region. In the last decade, milk production has stagnated, and imports have risen. Tabk 5.1 Production, aports, imports, 8nd growth rata for the prineiprl 8 g r i c M producb, 1989 and 196647 A'-#=''-=' e- Pmdvct Wst d. mJdR) - 1 pot.Oar 1,690 0.6 R~W 1~ a 4.2 Ca~n 322 66 -1.1 Yellow corn 7a5 23 2.5 SUP=- 6,335 20 49 -1.4 Coffee 107 1s 2.8 Cornrtucb 225 0.6 Banr 56 0.7 wheat 160 140 -1.7 Beef 112 10 0.l Mi 802 35 . I& Potatoes, cornstarch, and beans are the major domestic products. (Many other products fall within t h s group as well, such as quinua, yuca, broad beans, and peas, but production is small.) Whether i - - these products ought to be classified as truly domestic in effect, nonaadables is problematic. If they are closely related to an important tradable product, their price behavior could be g o v d by trade and exchange rate policies (Escobal 1989). For example, then seems to be a strong amumer- substitution correlation between wheat, rice, and barley (tradables) and potatoes, quinua, and cornstarch (domestic). Such correlations must be considered in analyzing the effects of changes in foreign trade policy. Sectoral development Agriculture has not been a dynamic sector in Peru, growing at an average annual rate of 1.2 percent between 1970 and 1988, well behind industry (1.6 percent) and GDP (2.3 percent). And population growth has outstripped all of these measures of economic growth. Thus agriculture's share in GDP has declined, From 14 percent at the beginning of the 1970s to 12 percent in 1988. Agriculture has also lost ground as a generator of employment: it accounted for 45 percent of total employment in 1970, but only 32 percent in 1987. It does, however, continue to generate more employment than any other productive sector. As agricultural production per capita fell, exportable production also declined. Agriculture's share in overall exports declined by even more because of the overall growth of exports. It fell from 30 percent in the 1960s to 20 percent in the 1970s and 10 percent in the 1980s (IICA 1990a). These figures do not include the illegal production of coca leaf. Including that production causes production and export statistics to jump dramatically. Illegal coca leaf production is estimated at 190,000 metric tons (IICA 1990b), with exports in 1989 equivatent to 50 percent of legal exports. The gross value of coca derivatives is estimated at US$2 billion, nearly equivalent to the total value of legal exports. Evolution of agricultural pricing policies Agricultural pricing policies have had vastly different effects on agricultural exportables, importables, and domestic products (nontradables). l3ponabIes Exportables include traditional and n o n t d i o n a l products. The major traditional w p s are coffee, cotton, and sugar. Peru, like most developing countries, has taxed traditional exports basically for fiscal reasons (Kmeger, Schiff, and Valdb 1990). while protecting many importables through tariff and nontariff restrictions. This has left many exportables with negative rates of indirect protection and has implied a lower real equilibrium exchange me, overall.  Coffee and cotton were excluded from export promotion efforts designed to boost production by guaranteeing fixed minimum or support prices around harvest time. The price received for much of cotton production never even reached the level of world market prices, since cotton could be exported only after the needs of the domestic cotton fiber processing industry had been satisfied. These domestic needs were, in effect, defined by the processing industry itself, thus guaranteeing its minimum supply and giving the industry oligopolistic power over prices. Under those conditions, cotton could never command a price higher than the international price - domestic processors could always buy the fiber to be exported at that price - but it could drop below that price because of the oligarchic power of the processing industry within the domestic market. Subsidies on agricultural inputs, especially irrigation, fertilizer, and credit, which were especially important during the Garcia government, did not significantly shrink the overall negative rates of protection. This was the situation facing traditional exportables in February 1991: a direct tax of 5 percent, an indirect tax of 15-50 percent arising from the protection of importables, and an implicit tax on production arising from the 15 percent tariff on most agricultural inputs (fertilizers, pesticides, tractors). Moreover, there were almost no credit subsidies available at the time. In recent years, nontraditional agricultural exportables have fared better than traditional exportables because they were not taxed directly and received a tax certificate (CERTEX) subsidy worth about 10 percent. In net tenns, however, even ontr traditional exportables were probably unprotected because of the much higher rates of protection on importables. With CERTEX gone, however, nontraditional exportables are not much berter off than traditional exportables - both groups clearly have negative protection. Importables The conflicts inherent in Peru's agricultural price policies and the powerful political interests behind them become starkly clear in the importables sector. The stakes are high for several reasons. One is that the farm sector hires more than 30 percent of the economically active population. Another is that variation in the prices of importables has a strong effect on the prices of most agricultural products because of the high degree of substitutability between importables and many domestic products (potatoes, quinua, cornstarch). On one side of the price issue are producers, who seek high prices. On the other side are urban consumers, who seek low food prices (Amat and Paz 1989). and agroindustrial processors, who seek low prices on agricultural inputs. The pow= base for the low food price faction is in the urban population (70 percent of total population), especially in Mewpolitan Lima, where 30 Percent of the population and 36 percent of the electorate are concentrated. Public expenditure and the instruments of economic policy have traditionally favored those interests. To avoid increasing the cost of living, governments have avoided price hikes for mass consumption food items which constitute a large part of the daily diet of the working class. Among the most important of these items in Peru (sometimes called strategic food items) are wheat and wheat products, rice, potatoes, and cornstarch. The other powerful force in the low price camp - although much less often mentioned - is the agroindustrial processors of primary agricultural products. These industries have a high import component: milling, 75 percent; brewing, 54 percent; milk products, 42 percent; balanced food products, 36 percent; fruit and vegetables, 33 percent. These five industries together represent approximately 70 percent of the gross value of agroindustry production (IICA 1990b). With rare and brief exceptions, the cheap food factions have prevailed. Thus during 1970-84, the subsidy for wheat was the equivalent of 40 percent of the fob value; for soybean oil, 32 percent; for rice, 25 percent; and for hard corn, 13 percent (IICA 1990a). While relative protection of urban consumers does not translate directly to an equivalent lack of protection for domestic producers, the subsidies did imply a lack of protection for competitive domesfic products during that period (Ministry of Agriculture 1991). A brief period of policy change occurred in the mid-19805, in the first years of the Garcfa administration. AgriculturaI importables were substantially protected and, consequently, so were domestic substitutes. htemational prices for primary food products were low in those years and domestic prices, which were highly regulated, were higher than internatonal prices. Despite low international prices, producer prices were higher than consumer prices because of price guarantee mechanisms. Thus low international prices enabled the Garcfa government in the early years of its administration to achieve two difficult-to-reconcile objectives: reasonably low prices for basic foods and improved agricultural prices and profits. Agricultural profits rose during the first years of the Garcfa government, and agricultural production responded favorably: it rose 4 percent in 1986,3 percent in 1987, despite a fall in agricultural versus urban income, and 4 percent in 1988. Between July 1985 and September 1986, the rural-urban terms of trade rose 75 percent, and agriculture's share in national income rose from 8 percent to 9 percent between 1985 and 1986. However, a reversal began in 1987, when relative agricultural sector fell again. Agricultural production fell 4.2 percent in 1989 and 9 percent in 1990.  In the early years the Garcfa government financed the differences between the higher producer prices and (lower) consumer prices through the spread between the (low) international price and the (somewhat higher) consumer price. That difference, collected by government trade organizations (ENCI,ECASA), acted as an implicit tariff. However, as the exchange rate became strongly overvalued, the implicit tariff on primary food imports became a subsidy. In January 1986, in the absence of exchange rate overvaluation, the implicit tariff represented 65 percent of the import price of wheat; two years later, exchange rate overvaluation reached 274 percent and the implicit tariff was 89 percent (Escobal 1989). As the exchange rate became more overvalued, the subsidy on agricultural imports and several agroindustrial derivatives became so large (at the end of 1987 the price of the dollar was five times higher than the official rate used for those imports), that those products were re- exported to neighboring countries, a c h as Bolivia, Chile, and Ecuador. To compensate producers for the low prices of agricultural importables, the government boosted subsidies on agricultural inputs, especially fertilizer and credit. Fertilizer was sold at only 5 percent of the international price.ll Although imports doubled, supply did not satisfy the demand at that price. At such a distorted price, much of the fertilizer was purchased for reexport - or even for its container, which, with prices so low, could be resold at a profit. In summary, despite brief periods of protective tariffs, there seems to have been a net lack of protection for agricultwal importables over time, with especially high negative rates of indirect protection (overvalued exchange rate, protection of other importables). Input subsidies probably did i not compensate fully for the dre ct and i n d i i lack of protection, so value added must have experienced a net lack of protaxion. Case studies are needed to determine this precisely, however. For example, rice production has been heavily subsidized through the very low price for the irrigation water. Regional variations in prices would also need to be examined. The picture changed after the main tariff reforms of 1991. First, without exchange rate overvaluation foodstuff importables received positive nominal di r ect protection. Tariffs of 15 or 20 percent were applied to agricultural imports (and inputs), while specific duties were applied to food (flour and pastes, whole powdered milk, skimmed powdered milk and anhydride milk fat) and agricultural products (wheat, rice, yellow corn, sugar) (table 5.2). Neither measure suggests a strong interest in maintaining a policy of cheap food. Income from the specific duties is earmarked for development of the agricultural sector. However, it is difficult to d e b m h e whether that represents a 11. lnformrtionprovided by ENCI. greater contribution to the agricultural sector than that received previously. Second, negative rates of indirect protection are lower because tariff protection on nonagricultural importables is lower and most nontariff restrictions have been eliminated. Some changes have made the current situation somewhat less favorable than the past. Most agricultural inputs now experience nominal protection of 15 percent - which is also the effective rate .of protection since domestic prices have been liberalized. The government marketing organizations for wheat (ENCI) and rice (ECASA) no longer have a monopoly. Furthennore, whatever gains ENCI achieves from wheat sales are used to finance the high subsidies on rice and yellow corn production in the jungle region. Transportation costs between the jungle and the coast are also subsidized and may amount to more than US$80 per metric ton. Table 5.2 shows the basic structure of protection for agricultural importables. While the structure presented there has been affected by new levies of specific duties on various importable products, those levies tend to reinforce protection for agricultural consumer goods without changing the relative structure of protection presented in table 5.2. Nominal direct protection for final products is i somewhat higher than for inputs, which means that effective gross drc et protection (i.e., without correcting for changes in the real exchange rate) is higher than 25 percent. For nonagricultural importables that protection is somewhat higher, while for exportables it is negative, which means that agricultural importables probably enjoy modetately positive net or total effective pmte&on. Agriculturalco~podr 25.6 Inputn for agriahm 15.9 Agricultural inputn for i n d q 173 C l l p i podr roragrhla 21.o Domestic products Many domestic products are close substitutes in consumption with the main group of agricultural importables. Consequently, their prices tend to be positively wrrelated. While markets for those products have been relatively free of regulation, important problems exist (Geng 1990). Principal among them are the lack of an adequate network of wholesale markm, deficient ~ r m a t i o n mechanisms for producers, excessive and complex regulations, small scale and widely d i d  production, and low educational levels among producers. Many of these factors have hampered the formation of producer associations to achieve better information about prices and greater stability in supply and have contributed to great price volatility. Evaluation of agricultural policies . Agricultural policies have been flawed both conceptually and in their implementation. Institutional weaknesses and disincentives are in large measure responsible for these problems (IICA 1990a). In the last twenty years, six organic laws have been promulgated for the agrarian sector. M i t e r s of agriculture have been relieved of office on numerous occasions and with each turnover have come new policies and priorities. The IICA 1990 commented that '[alt the global level, mechanisms for coordination among sectors exist ....At a more operational level however, there is a nearly total lack of mechanisms for coordination between the agricultural sector and others, which affect its development. In part, this is apparently due to a lack of overall policy conception, allowing sectoral interests to prevail ....Important institutional faults exist as well' (IICA 1990b). Too narrow a sectoral ficus A partial equilibrium approach seems to hold sway in the analysis of agriculture, to the neglect of the eatnomywide consequences of such policies. A conspicuous example is the protective response (generally through implicit tarif%) to low international prices. In a country that i s a net importer of those products, protecting producers instead of consumers constitutes a net loss on ptoducts such as wheat and milk products. Tbe same tunnel vision affects policies within the sector. Policies seem to be based on the presumption that each product has a right to protection, a view that ignores the fact that strong general equilibrium restrictions for land affect the whole sector. How strong those restrictions are can be seen in the efforts to relax the restrictions on large irrigation projects undertaken to bring new land under cultivation at the high cost of US$10,000 to US$12,000 a hectare. The Miitry of Agriculture's anuouncement that self-sufficiency in rice production is a goal of agricultural policy is another egregious example of the failure to consider general equilibrium effects. Too many incentiws uvrkhg at cross-purposes Agricultural policies have combined in a haphazard way to form a complex and volatile mix of measures and countermeasures, incentives and disincentives. Thus, for example, to compensate fot the disincentives arising from cheap food imports, agricultural inputs (fertilizer, credit) were heavily subsidized; to make up for inadequate road maintenance, transportation subsidies were granted. This situation of incentives and disincentives at cross purposes has a high resource allocation cost. It also makes it impossible to determine the final effects on producers and on the economy of this intricate network of subsidies and duties superimposed on the same productive activity. ' , Not only did these incentives and disincentives create a complex pattern; but they were extremely variable as well. Consider the uncertainty that must arise when fertilizer prices can shift from their full international level to only 5 percent of that level. The strong uncertainty generated by such incentive volatility leads to inefficient resource allocation, both within the sector and in the economy as a whole. Excessive reliance on price policy to redistribute income Because of large differences in natural soil fertility and transportation costs, production costs vary widely across regions. In a free and competitive market for products and inputs, producers with different levels of soil fertility, different production costs, and different transportation costs would receive the same price and, therefore, different levels of profit per hectare. The price of inputs would also vary, with prices rising as transportation costs rise. Such differences would also lead to differences in profitability per hectare. To achieve a more even distribution of producer income, producer prices often varied by region so that prices were higher in less fertile regions with higher costs. A notable example of this differential pricing affected producers in the region called the "Andean Trapeze." A similar objective motivated transportation subsidies for producers in regions with higher transportation costs. Uniform pricing of agricultural inputs, which levelled out differences in transportation costs, also had a rediibutive intent. Using product or input prices to redistribute income entails economic costs. In the case of products, zonal pricing artificially boosts production in favored regions over levels that would prevail without subsidies, and by favoring existing products, constitutes a disincentive for change. Considering the potential gains from nontraditional agricultural products, the costs of such distortions are likely to be high - both to the economy as a whole and to the favored region itself, by creating disincentives to search for new productive opportunities in which the region might have a comparative advantage over other regions. Uniform price policies for inputs stimulate inefficient r&urce use.  Income redistribution objectives ought to be pursued through policies with lower economic efficiency costs. Fiscal policies that do not distort the efficient selection of products and inputs are the appropriate policy instruments - the provision of economic infrastructure, and directly redistributive policy are several possibilities. Protection of the status quo Agricultural price policy has also protected existing production patterns. Conspicuous examples are the fixing of prices at levels that guarantee "adequate" profits for current production or the fuing of explicit or implicit tariff protection levels for the same purpose. These policies have pernicious effects because they block incentives to change, supporting a pattern of production that was itself established in response to distortions created by previous policy measures.  Chapter 6. Anticipated Effects of Trade Liberalization This chapter presents a rough estimate of the impact of the trade reform - tariff rationalization and the elimination of import prohibitions - on production and employment in industry. The impact of the sequence in which nontariff restrictions were eliminated is also examined for three subsectors - motor vehicles, fertilizers, and pesticides. The probable effects on agricultural production and income distribution are also considered. Impact on industrial production and employment The trade reform has affected protection levels for different industrial sectors, with some sectors gaining and others losing. Table 6.1 shows estimated effective protection rates for different sectors prior to the refonn. Table 6.2 arranges these seaors according to their level of protection to show losers and gainers: those whose protection was higher than 15 percent (the approximate effective protection for all sectors after tariffs are unified at this level) and those whose protection level was lower. This analysis is supported by a recent survey of industrial sector sales, which showed paper and metal products among the industries negatively affected by economic openness and beverages, tobacco, and footwear among those that expect to increase production. The reform of August 1990 lowered the average effective p~tection of industry from 82 percent (June 1990) to 40 percent in December 1990, and to 24 percent in March 1991 (table 6.3). The reform of March 1991 introduced greater homogeneity in levels of effective protection, although high levels of protection still remain for such activities as clothing and footwear manufacture. Liberalization will benefit sectors that together represent 60 percent of total industrial production and more than two-thirds of industrial employment (table 6.2). Adversely affected industries, which will have to adapt production to the new competitive conditions or, in some case, close their pTants, absorb only 3.3 percent of total Peruvian employment. Furthermore, that adaptation implies resource reallocation whose impact on production and employment in other sectors will be positive. Table 6.1. Industrial scctor protection, cmpbyment, md production before the 1990-91 trade r e f o m pominal Effective EmproYnvnr kgd Collecd f GDP %o 6) ImponabIcs Milk ploductr iab.cc0 Foawaar Paper dpaper p m d w Basic c h c m k h d f e h n M&i Olbcrchcmiulploductr Steel proccrriag -utitccr  Table 6.2. Sectors likely to gain or bse from trade Ebernlimtion 5% of GDP 6) Basic chemicals and fertilizers Otba chemical products Machincry and equipment Rubber and leather products Pap= and pclpa prod- Nonelectrical machiuy canned f Lb s Divcrrc mud producta Nonmudlic minerd products Stscl proocsring O t h a manufacturedproducts O t h a foods Totrl Te3aila d 0.79 0.48 F m h 1.39 1.7 6 Tn~po~mtionrmtairl 0.94 0.29 h d - 0.18 0.a Tobreco 0.40 0.07 Nonfaroun t m q m m h a rmtairl 2.65 0.20 1.91 031 Rintia0 0.91 0.34 F ~ ~ d o i l 0.40 0.05 Millmg 0.61 0.34 BJFa). prod- 0.50 0.28 ao t b i n l 1.m 1.a6 M d i b 0 .47 0.15 Sugar d rmaproduca 0 .20 0.14 Mi produar 0 .28 0.16 -goodr 0 .09 0.10 ~ootwcu 0 .22 o.n Total 13.03 6.79 Noos: Qainaa ace tboro with effective ntcr of proteaion boforo tbs reform of lcn thn 1 5 loren ace tbom with rpsr lbovo 5 paccnt. 1 dN~iti0llrlarrpoa~. Source: M i i o n sla;mrtra.  Table 63. Effective protection of industry before and alter the reforms of August 1991 and hlarch 1991 July December March %of EdOymmf I990 I990 I99l W P d a r e (%) M i products Milling m d bakery goods Other food products Beverages m d tobacco aorhig - 8 d Footwear Furniture Piper PV prod- w i g Chcmicrlr, hit, and fatilizcn Mdiciie Otha ebcmicd productl Rubba M d plantic products -- Nobmecrllic mined prod. p - g - D i m m i c products Non-decrriemrchinoy M.chinay & quip. for bdurey Tmqmtmrrairl Ekxal impact - The fiscal impact of the reform depends on many factors elasticities of supply and demand, customs taris, tax rates, the effects of production and consumption substitutions, and success in improving resource allocation and stimulating ccommic growth. Only a m g h indication w ill be presented here of what the probable fiscal impact might be. Total tariff inwme for 1991 is estimated at about US$489 million, on the basis of projected imports for 1991 of USS3.260 million, a 15 percent effective tariff after the ref- of March 1991. These projections, in turn, are based on ad valorem duties of USS117.6 million for the first quarter of 1991 on imports of USS663.7 million cif. Export tax income is estimated at USS19.2 million on the basis of an cstimated USS.475 million in traditional exports for 1991 and an effective export tax rate of 3.2 percent. Thus, fiscal income f rom -58 - taxes on foreign trade are estimated at USS526.3 million for the year. If the general sales tax (IGV) and the selective consumer tax (ISC)are included, at an effective combined rate of 9 percent, total revenue from foreign trade will be US$796.4million. These figures reflect a recovery in fiscal revenue from foreign trade after the drop in 1990 (table 6.4). .Table 6.4 Tax revenue from foreign trade: 1989-91 ( *us of US dollars) Ad valorem tax 255.6 174.6 447.1 G e n d ulu tax and dectivc conrumtr turJ 150.0 109.2 270.1 Otherimporttua 113.0 88.6 The fiscal impact of the deepening of the trade reform after Mar& 1991 was also estimated through a simple comparison of tariff revenue in September 1990 and projected tariff revenue for 1991 (table 6.5). Although tariff payments drop from 1.8 percent to 1.6 percent of GDP, imports rise by USS474 million. This rise will boost tax revenues from the general sales tax and selective consumer tax and compensate for the decline in tariff income. Moreover, these figures overestimate collections because they do not account for domestic demand substitution, which, in the new situation, will be satisfied by imports, a phenomenon that will become more pronounced over time. Impact on specific industrial sectors By February 1991, the Peruvian government had reduced and consolidated tariffs into three rates (50 percent, 25 percent, and 15 percent). Prior licensing (which covered 50-60 percent of imports), the no-competition decision (allowing tax exemptions for imports that did not compete with domestic production), and most import prohibitions were suspended. Minimum domestic content requirements 'were eliminated, except for motor vehicles. Despite the rapid dismantling of quantitative restrictions, significant nontariff trade barriers remained in effect after the reforms of August and November 1990. Some simply continued to protect particular sectors. Some effects may have been unintended. The removal of some restrictions while others remained in effect led to anomalous results (or could have done so, if later reforms were not implemented as planned). And trade liberalization measures may have interacted with other reforms in unexpected ways that created the potential for resource allocation distortions. Some of these effects are examined in the three case studies that follow, which show how nontariff trade barriers continued to distort incentives and resource allocation, at least until February 1991. Rapid progress in trade reform since February 1991 has made some of the descriptions and recommendations presented here obsolete. Nevertheless, the studies provide important insights into policy alternatives. 3hr motor vehicle industry: a case of nontmjbamiem Before August 1990, imports of assembled automob'iles were prohibited (except those impoited by diplomats), and vehicle assembly packages were subject to a 1 percent tariff. The reform of August 1990, allowed the free importation of new automobiles at a 50 percent tariff and of new buses and trucks at .a 25 percent rate. The tariff for vehicle assembly packages was fixed at 15 percent. While trade in new vehicles was liberalized, the vehicle assembly in Peru remained heavily regulated, subject to domestic content rules and obligatory exports. Requirements for domestic content varied from 40 percent for buses and trucks and 38 percent fot automobiles, to 21 percent for diesel motors and 18 percent for motorcycles. Each year, motor vehicle assembly firms had to present to the Ministry for Industry a list of the officially registered parts they wished to incorporate into a given vehicle in order to comply with the content requirement. Alternatively, content requirements could be satisfied by exports of motor vehicle products. Beginning in 1988, automobile assembly firms were also required to export automotive industry products equivalent to 20 percent of the value of the Completely Kaocked Down (CKD) packages  they imported (for example, Nissan fulfilled the requirement by exporting parts to Mexico, Volvo by exporting trucks to Bolivia and Ecuador). Because of the depressed state of the industry, the export requirements were temporarily lifted by special decrees. As the model presented in Appendix V shows, the effective rate of protection for a domestic industry depends on the interaction of diverse elements of the protection and regulatory regimes and ' h e resultant pattern of incentives and disincentives. For domestic vehicle assembly, import prohibitions or tariffs that restrict imports stimulate domestic assembly by raising prices for assembled vehicles. Domestic content and obligatory export requirements and tariffs on imported packages discourage domestic assembly because they raise input costs. The overall impact of these factors depends on their relative magnitudes. Appendix V presents a model for estimating the effective protection rate for assembly operations, based on data for three truck models provided by Nisaa Motors of Peru for several real and hypothetical trade regime. (table AS.l). The calculations are illustrative only; they are intended to provide a general sense of the effects on the industry of different stages of the reform process and of various reform alternatives. The model results show effective rates of protection of 79 percent to 141 percent before the reforms of August 1990. These protection rates were the r a t of nominal protection rates for vehicles of 47 percent to 62 percent and nominal increases in parts costs of 19 percent to 31 percent. Effective rates would have been wen higher-1 12 percent to 177 percent-without the domestic content requirement.. In August 1990 the vehicle import prohibition was replaced by a 25 percent tariff and the tariff on packages was increased to 15 per- domestic content requirements remained. These changes reduced the effective rate of protection to between-0.03 percent and 17 petcent? If the requirements for obligatory exports had been in force, effective protection rates would have been wen lower. The March 1991 reforms lowered tarif& for motor vehicles to 15 percept and removed domestic content and obligatory export requirements, bringing the effective protection rate for assembly operations to between 17 percent and 18 percentU. Had a flat 15 percent tariff been adopted instead  and domestic content requirements remained the same, as some had recommended, effective protection rates would have fallen to between -13 percent and -50 percent. The protection regimen in place before the trade reforms was costly to consumers and to the economy as a whole. Import prohibitions on assembled vehicles increased vehicle prices, while domestic content requirements increased production costs and encouraged the production of domestic parts that were four or five times more expensive than their equivalents at international prices. Trade liberalization will allow consumers to benefit from lower automobile prices and the economy, to benefit from lower domestic transport costs. The long-term impact of the trade reform is more difficult to predict. With the lower effective rate of protection will come rationalization and reduction in the size of the motor vehicle industry. Some sense of the industry's future can be derived from data on the historical development of the industry, information on parts costs, and the effective protection rates calculated above. Historical data show that domestic assembly constituted 45 percent of the market in 1981, just prior to the earlier liberalization period, and in 1990 (table 6.6). With the reversal of liberalization measures and the passage of the new Law of the Automotive Industry in 1984, the domestic share rose to 88 percent. While it is true in Peru, as elsewhere, that'economic recovery stimulates sales of both domestic and imported vehicles, three factors make it improbable that the industry will maintain the same market share in an expanding market as it did during the earlier episode of reform. The first @ is that the effective protection rate is probably lower now than during the previous episode, when the maximum tariff was 60 percent (see NoguQ 1991 for a discussion of the liberalization effort of 1979- 1980). The second is that motor vehicle imports are now permitted, which creates additional competition from outside. The t h i i is that technical progress in the world automotive industry has probably increased the competitive edge of foreign producers over Peruvian assembly operations (see Karmokolias 1990 for an overview of industry trends).  Table 6.6. The Pauvian automotive industry, 196549 to 1990 (1.000 vchiclu) Domestic To& D m s t i c mmbb mmbb ImpOnr a - , -I rhau Year (1) 0) 0) (I)/@) Automobiie assembly, which is more complicated than truck assembly, is less likely rn be competitive with imports. Truck assembly also enjoys natural protection from high transportation costs and is more labor intensive than automobii assembly. Transportation costs are about USS1,500 for an assembly package from Sweden and USS18,OOO for a finished cab and chassis. While truck assembly operations will undergo basic change under the new trade regime, they could remain viable. Domestic production of diesel motors for Volvo and as spaces will probably survive as well. Other firms that currently assemble some models in Peru and have the diibution rights for others will probably con- on imports. Parts production will undergo change as well. Some Peruvian parts for vehicle assembly are competitive with imported parts (approximately 14 percent of parts at the 15 percat tariff rate, according to data from Nissan). That smaller competitive segment of the parts industry will probably provide parts for the remaining assembly operations and spates for existing autos. C o m r d o n s witb parts suppliers indicate tbat many of them have become parts importem. Fertilizers: import monopolies and the &cts of indfrcct taxation Five of the six fertilizer producers are admWtered by the state: PETROPERU (urea), FERTISA (dphate and ammonium nitrates), PESCAPERU (guano), INCASA (ammonium nitrates), and s MINEROPERU (phosphorates). INDUS (superphosphates and mixed fertilizers), the only private concern, is relatively small. The National Enterprise for International Trade (ENCI), a state agency, had a virtual monopoly on imports. Between August 1974 and December 1982, ENCI was the only finn authorized to buy fertilizer from domestic producers and to import and distribute fertilizer within the country. Even as 'its legal monopoly on imports and distribution ended in 1982-83, in practice ENCI continued to be the only importer because of its access to public funds to subsidize purchases, its capacity to handle large shipments, its established distribution network, and its exemption fiom import duties. In the mid-1980s, ENCI sold nearly 90 percent of all fertilizer sold in Peru; the remaining 10 percent was distributed through private firms and agricultural associations. ENCI recovered exclusive import rights for a time during the Garcfa administration, but the market was once again liberalized toward the end of that administration. Fertilizer was heavily subsidized throughout the country through ENCI. To compensate for transportation costs and to subsidize fertilizer sales, ENCI was partially financed by the Fondo de Reactivacidn y Seguridad Alimentaria (FRASA). According to some estimates (Castro Boissier 1990, p.7), fertilizer subsidies reached an astonishing 97 percent in 1988. '2he production of fertilizers in Peru is more expensive than abroad because of the small scale of operations, antiquated equipment, and high input costs. The refbrrn and stabilization programs of 1990 raised input costs even higher. 'Ihe stabilization program eliminated the input subsidies implicit in the low-price petroleum inputs bought from PEI'ROPERU and imposed a 134 percent sales tax on petroleum products. 'Ihis left domestic production, which was only marginally competitive with imports even in the best of times, in a clearly noncompetitive position. This drastic shift in circumstances for the fertilizer industry poses a difficult policy problem. Subsidies of fertilizer sales and petroleum inputs diirted resource allocation, so dimbating them was clearly justified. However, the high taxes on petroleum inputs caused another type of distortion. In principle, fertilizer will be produced in Peru if the current underlying cost of production is not greater than the import price plus the dimlution cost from the port of entry. If production costs are raised artificially by a tax on inputs, production that would be viable at international prices will not be competitive. Further complicating matters is a proposal for a credit system for indirect taxes on inputs for products to be exported. Implementing such a system (if fertilizer production was or could be competitive internationally) would create incentives to export Peruvian fertilizer, but not to sell it domestically. This could lead to a situation in which the domestic market would be supplied by  imports, while domestic production would be exported. An alternative (if fiscal restrictions permit) would be to exempt the petroleum products used as inputs in manufacturing industries from the tax in order to avoid those distortions. Another alternative would be to eliminate the tax on petroleum products sales in the long term and simultaneously apply the general tariff to imports of those products. Pesticides: import licenses Pesticides must be registered with the Ministry of Agriculture before they can be imported and sold. Samples must be supplied for evaluation by the Instituto Nacional de Investigacidn Agroindustrial y Agrfcola (INIAA), which verifies its contents and makes recommendations for its use, and by the Control Laboratory in Callao, which analyzes active ingredients. This process takes from six months to two years, and the process must be repeated every three years. Once the product is registered, each shipment must receive an entry license from the Agricultural Department of the Ministry of Agriculture, which takes three or four days. The delay, uncertainty, and cost involved in product registry and shipment licensing discourage potential importers and so constitute a nontariff barrier that benefits domestic producers. Legitimate standards to protect health when toxic substances are involved are not a t issue, but the administration of public health safeguards ought to minimize costs for both state and private firms.The need for repeated product registry and for licensing each shipment is questionable. Those measures ought to be simplified if they have not been removed by the general elimination of licenses and registries of the March 1991 reforms. Impad on the agricultural sector 'Ihe analysis of the possible effec$ of liberalization on agriculture looks at botb the effects on gmups of products, and in the aggregate, effects on agricultural exports and imports are examined briefly. The analysis primarily compares effects that could take place over the next few years if the liberalization proposed in this report were implemmted, w ith the situation prior to the e s t a b l i i e n t of the regime of specific tariffs (withvarying levels, depending on irrternatonal prices) on some agricultural products and of fixed d valorem tariff rates of 15 percent and 25 percent on the rest of imports.14 However, it also discusses briefly the effect of the variable tariff regime. Production Traditional ex~ortables. Exportable agricultural products should, in principle, benefit from the elimination of negative nominal direct protection, cost reductions arising from lower direct nominal protection of some tradable agricultural inputs, and the higher real exchange rate associated with - lower nominal protection for the importables of the economy as a whole the most important positive effect on the aggregate value of this group of goods. To the extent that the new export regime is applied to agricultural products, the effects of trade liberalization will be broader. The specific tariffs and high taxes on some important agricultural importables will have the opposite effect. They may even once again severely reduce incentives for the production of the exportables, counteracting the expansive effect mentioned above, although the negative effect will be weaker if the specific taxes compensate only for a decline in international prices. Another negative effect on the sector's aggregate value could come from the lower subsidy on the use of irrigation water. On balance, however, the positive effects appear to outweigh the negative effects, and it is expected that exportables will be somewhat favored by the new trade policy. m. Agricultural importables, with some exceptions, are likely to benefit from liberalization. Surcharges maintain the levels of direct nominal protection for agricultural importables, which will also benefit from lower negative indirect protection as protection is reduced for the rest of importables. The exceptions are products whose level of protection will be reduced, but these products have little weight among importables. In net terms, agricultural importable. are likely to be more favored than exportables. In particular, rice, wheat, sugar (also an exportable), sorghum, corn, and m ilk could benefit greatly if international p r i m rise while the specific tariffs are in effect. However, the elimination of the irrigation water subsidy will reduce incentives for rice production, perhaps even enough to outweigh the impact of that tariff.  Chicken, which has a mid-range of nominal protection of 25 percent, would have been included in the previous group except that yellow corn, its main input, is taxed by a tariff higher than that on the product, which will have a negative impact on its production value added. The value of beef, pork, and mutton will also fall because of the dominant effect of the reduction in nominal protection to 15 percent. Milk still receives special protection. Domestic ~ r o d u c ~ . prices of domestic agricultural products (potatoes, corn starch, quinua, The etc.) are closely tied to those of importables, especially wheat and rice. Thus domestic goods production, ought to benefit from the new trade policies, while liberalization of domestic prices, by boosting transportation costs, may lower incentives to produce some of those goods (e.g., potatoes). Recent changes in international prices may have the same effect. The net effect on the production of domestic agricultural goods is therefore uncertain. Summary. In the aggregate, agricultural production will probably grow as sectoral protection increases at the expense of the rest of the economy. The production of exportables will grow, as will that of some importables Qellow corn, wheat, sorghum, milk, among others), as long as hmational prices do not continue to fall. Meat, most imporrant domestic products, and rice production will tend to remain stable or contract. Agricultural exports will increase, and agricultural imports will probably remain stable, as long as international prices continue to fall. Income distribution Trade liberalization will probably have a regressive impact on lower-income urban groups. Wheat, rice, potatoes, and cornstarch constitute a large share of consumption for these groups, and as the relative price of these products rises, the relative p r i a of their food basket may also The redistributive effects are more complex within the farm sector. The small farmers of the coastal region, who work 25 percent of the arable land, with high relative incomes within the agricultural sector, will benefit as the aggregate value of their primitpal products (cotton, agarcane, and yellow corn, and their nontraditional exportables) rises. Other faaon will dampen the benefits somewhat: the increase in the relative cost of the food basket, the elimination of the irrigation water subsidy, and  the negative impact of liberalization on other products of the region, such as potatoes and milk c0ws.l6 The small farmers of the jungle region, who also work 25 percent of the arable land and have a high relative income within the sector, will also benefit from the boost to coffee, cacao, yellow corn, and banana production that trade liberalization should bring. The higher relative cost of the food basket will reduce the gains somewhat. It must be noted, however, that production in the region is heavily influenced by subsidies to promote the cultivation of alternatives to coca leaf. The small farmers of the Andean region, who work the remaining 50 percent of the arable land, are the poorest of the farming groups and even among the poorest in the country. World Bank estimates indicate that the Andean farmers constitute half of the poorest 30 percent of the population in Peru. The impact of trade liberalization on their income is unclear. Among their main products are potatoes, wheat, cornstarch, quinua and other nontradeable products. The relative price of this 'package" of products will probably rise as a result of the trade reform, which will affect farmers both positively (production) and negatively (consumption). Domestic price liberalization may have less favorable effects, as transportation and input costs rise in some areas and fall in others. The net effect of these factors is unclear. Several factors may attenuate the regressive impact of trade liberalization. One is the high level of on-farm consumption in the Andean region. (To the extent a farm family grows its own food, it is not affected negatively by price increases.) Another is the great potential for diversification in that zone, with its agroecological diversity. At any rate, any negative d i ' b u t i v e impact on that region ought to be handled through technological assistance, improvements in infrastruchve and direct redistributive measures, rather than through market interventions that distort incentives. More recently, temporary compensatory surcharges were placed on the value of food products and imported agricultural inputs, and the proceeds are intended to support agricultural sector development. Wheat and wheat products, hulled rice, sugar, yellow corn, and sorghum are the principal products affected. The surcharges seem to formalize implicit tariffs on those products existing up to February 1991. These surcharges increase protection of agridhue at the expense of the rest of the economy. - The surcharges would benefit agriculture twice once through higher prices and again as the funds are invested in the sector. 16.lbe &miartion of irrigation water atb8idii will roduce tho danger of rrliniration.od flooding in tbe fmure, witb a nu positive hnp.ctonproductiwiucaui~inthelong~~~. -69 -  The attempt to provide greater protection for agriculture, especially for agricultural importables, and with them domestic production appears to be an effort to modify the regressive distributive effects of foreign trade liberalization on the small farmers of the Andean region. While it is true this will improve the position of Andean fanners, it will also subside production in the coastal and jungle regions. Andean farmers would still be disadvantaged by higher costs for inputs and transportation. At m y rate, a less distorting way to achieve the objectives of the surcharges would be through the use of price bands, which are less discretionary and more transparent and systematic.  Chapter 7. What Remains to be Done? The Fujiimori government has taken significant and fortright measures to open the economy to foreign trade, Additional steps are now required to complete the reform and solidify the new trade structure. This chapter summarizes the main recommendations on tariffs, antidumping, and reference prices and nontariff barriers. Because of its socioeconomic importance, the agricultural sector receives special attention. Tariffs The tariff refonns instituted so far are clearly on the right track. The next move would be to bring all tariffs to 15 percent. All positive tariffi constitute an indirect tax on exports. Even drawback schemes do not eliminate discrimination in favor of the import substitution sector. To the extent that domestic goods are substitutes (in consumption and production) foi imported goods, tariffs raise the price of the import competing domestic goods relative to exportables, constituting, in effect, a tax on exports. That tax has been estimated for various Latin Amesican countries as at least half the average tariff rate1'. In Pem, with a current average tariff of 17 percent, that tax would be equivalent to at least an 8.5 percent tax on exportables and an 8.5 percent subsidy for importables. That means that fewer resources would move into exports than would be the case without airif%. The simplificationof tariffs to only two rates is also important, because it reduces the dispersion of effective protection among productive sectors, and lowers allocational inefficiency which is proportional to protection dispezsion. Nonetheless, considerable dispersion remains with the two tariff rates of 15 and 25 percent. Supposing that the 15 pe~rcmtrate is applied only to inputs and the 25 p e r m rate only to finished products, effective protection could be as low as 26 percent in cases where value added reaches 90 percent (only a few cases) or as high as 115 percent in cases where value added does not surpass 15 percent. If the rates on inputs and finished products were reversed, negative effective protection rates could result. This dispersion ought to be reduced or eliminated by dropping the 25 percent rate down to 20 percent or 15 percent. A eingle tariff rate has advantages both for resource allocation and for administrative simplicity. 17. (World Develmmmt Ranm 1987).  Tariff reduction and simplification also allow the real exchange rate to rise, which has the desirable effect of stimulating the export sector. The effect would be intensified if some of the remaining nontariff barriers were also reduced and if reference prices for food products (see chapters 3 and 5) were also eliminated or replaced with a price range on two or three of them. Ultimately, how effective customs tariffs are in raising revenue and influencing resource dlocation depends on the levels of smuggling and informal sector economic activity. In that regard, the optimum tariff is difficult to determine. For Peru, the high cost of border control and of transportation from important markets (United States) suggests that 15 percent is a reasonable tariff. Many of the products that are currently subject to the 25 percent duty have the high vdue-volume correlation that has traditionally been associated with contraband. While short-term fiscal considerations may block the move to a IS percent tariff for these goods, there may be a potential positive revenue effect even in the short-term if lower tariffs brought more trade out of contraband and into the tax net. In addition, the ongoing modernization of legal structures by encouraging informal sector activies to formaIize their status as long as the changes include less discriminatory and more efficient controls than in the past. Antidumping legislation and reference prices Once the first stage of the trade reform was in place, industries that were hurt by their loss of tariff protection began to clamor for other measures to protect domestic production from the threat of "unfair" foreign competition. . Peruvian customs legislation authorizes the establishment of official minimum prices when "abnormal trade practices cause or threaten to cause serious damage to the national economy, the state or domestic producers of certain goods." The customs authority can also establish reference prices for cases in which the importation is not a commercial operation and, therefore, the concept of normal price or freely competitive price does m t apply. A list of reference prices has also been drawn up for used goods imports. M c i a l minimum prices and refixem prices, as d i d o n a r y protection measures which encourage resource allocation to importables, can become a direct obstacle to imports and an indirect obstacle to exports. They also can incite Peru's trading partners to apply antidumping measures against Peru's exports-say, if reference prices werenot applied on inputs for products to be exported, thus enabling the good to be sold mon cheaply in the rest of the world than in Penr. The most sensitive areas in this respect are textiles and clothing, areas in which Peru has some comparative advantage. High-quality versions of a given type of product almost always have international prices above the minimum reference prices; thus, their domestic prices are not affected, while the domestic prices of low-quality versions of the products are increased by the application of reference prices to imports. 'Thus,reference prices encourage low-quality domestic production. That effect may seem to be of little importance in the absence of a high percentage of unskilled labor, but the possibility of comparative advantage in the production of some highquality products (clothing and some textiles) . . cannot be dismissed out of hand, even for the domestic market. Reference prices have the further undesirable effect of encouraging domestic consumption of highquality relative to lowquality products artificially and inhibiting domestic competition, favoring the formation of cartels among the protected industries. To stave off efforts to push through such economically dangerous and ill-advised measures as reference prices, which would result in the dismantling of the tariff reform, it is useful to have formal antidumping legislation in place based on the GA'IT antidumping code. Decision 230, approved in March 1991, contains nonns for preventing or conecting distortions in competition generated by dumping and subsidies within the Andean Gmup. That Decision, which is based on GA'IT codes, is also appropriate for foreign trade from outside the Group. At the national level, a Supreme Decree consistent with GA'IT nonns and Decision 230 has been proposed for debate. The project has been well received by politicians and the business community. National legislation governing unfair trade practices also has the advantage over reference prices of transparency of aaministrativeprocedures. Care must be taken to ensure that unfair trade practices legislation is not used for protectionist purposes. To that end, adequate verification k t be required not only of the existence of unfair practices but also of their damage to domestic industry. Any antidumping or compensatory duties should be set at the lowest amount that will eliminate the damage or threat of damage. Clear definitions of dumping and clear limitations on the period during which duties will in effect remain must also be specified. Nontariff barriers Before the reforms of August 1990,5040 percent of imports were subject to prior licensing or (in the case of imports h m Andean Pact countries) reghation by the Institute de Comercio Exmior (ICE).The license requirement was suspended in August 1990, until the end of December 1990. The  suspension was extended once, and the requirement was finally abolished in March 1991, along with other registers, consular approvals, permits, and administrative requirements (Supreme Decree 060- 91-EF). The period between August 1990 and March 1991, when license requirements were temporarily suspended but not formally abolished, was one of significant liberalization but considerable 'uncertainty. Thus the decree of March 1991, by abolishing all licenses, registers, and other administrative barriers, ended the uncertainty about the government's intentions to eliminate licenses permanently. All efforts to bring back import registers in whatever form, should be strenuously resisted. Even registers kept for 'statistical purposes' only or to 'supervise' imports to prevent dumping would cause delays adding to costs, and introduce an undesirable discretionary element. The appropriate way to deal with dumping is through transparent investigative procedures that are consistent with the GATI' antidumping wde. Remaining obstacles for exports Much of the antiexport bias of the trade regime has been eliminated: export subsidies are gone, average taxes on exports are low, and new regulations have been adopted to lighten the burden of indirect taxes on exporters. Assuming nontariff barriers for imports have actually been eliminated, any remaining bias now comes from a relatively low tariff, which will be lowered even more as Peru advances toward its target uniform tariff of 15 percent, with the consequent depreciation of the real exchange rate. Complementary actions to elimbte other bias against exports are also needed. Since Peru is not a dominant world supplier in its traditional exports, tbere is no general argument for disdmhatiig against them through taxes. Once the fiscal situation pcfmits, taxes on some exports can be eliminated1'. Also, the requirement that fub exports be certified by CERPER, a mte company, is inefficient and should be abolished.- While the recent reforms create the necessary conditions for export growth, especially among nontraditional products, export growth will continue to be affected by infrastructural insufficiencies and other market distortions. Traditional exports, which still constitute 70 percent of exports, are hampered by continuing problems with port and transportation infrastructure and security. Recent l reforms of labor law, ~ r a property rights, and banking institutions have substantially reduced domestic market distortions, but the reforms need to be consolidated and secured. Further action would be beneficial to increase flexibility in labor markets (along the lines of legislation such as D.S. 016-91-TR, R.M.032-91-EF, D.S. 0155-91-TR) in the areas of job stability, winpensation for years of service, union legislation, and social security. Such actions are necessary to improve resource reallocation among sectors and companies as a consequence of trade openness and to permit the intensive use of labor in new export projects, thus taking advantage of its relatively low cost. More specific recommendations are presented in World Bank 1991; actions taken by the government are detailed in appendix 1. The capacity of the export sector to respond rapidly to trade reform may be vital to the survival of that reform. What may inhibit that response is not the trade policy measures themselves but the speed at which the general process of market reform is real'ied. For example, export response in the fishing and mining sectors will be limited as long as they suffer the rigidities caused by extensive state omenhip of property. The following section offers suggestions for agricultural exports. Proposals for better use of agricultural potential Foreign trade policy Tradable products and inputs in the agricultural sector need to be treated like other tradables. In particular this means applying the same single tariff to imports of agricultural products and inputs as to other imports and treating agricultural exports like all other exports-no special tax, no discrimination between traditional and nontraditional exportables, and the same treatment of indirect taxes for other exportables. Equal treatment for all tradables is essential if trade liberaI'd011 programs are to be sustainable. Equal treatment is a powerful line of defense against pressures from those who lost protection as a result of Iibdization. And within the agricultural sector, equal treatment among products is the best guarantee that the sector will develop to the full potential implicit in the enormous diversity of agroecological niches-that the high opportunity cost of land use be effectively covered by each productive effort so that despite great differences in production each region can exploit its comparative advantages.  Equal treatment of agricultural and nonagricultural exponables might also be extended to the return of indirect taxes, if those administrative problems involved in calculating those taxes could be overcome at reasonable cost. All special restrictions on agricultural exportables should be eliminated, particularly any requirement. for satisfying domestic demand before exports are allowed (e.g., cotton). Quality wntrols for exports administered by the state should be applied for that purpose only and eliminated once foreign buyers of the products demand quality control by a private entity. Quantity restrictions on agricultural imports should be eliminated. Quality and sanitary controls should be no different from those that apply to same products domestically. Moreover, the controls should be eliminated on products that are permitted into other countries with adequate wntrols. Most restrictive regulations governing agricultural trade are ineffective and constitute a heavy tax on foreign W e . In many cases, the regulations simply duplicate quality control already imposed by the foreign-private sector or by importing countries. As a short-term exception to equal treatment for all tradable products, the use of price ranges, or bands, ought to be considered. The mechanism would be applied through specific and variable tariffs used to moderate the domestic impact of strong fluctuations in international prices. The following general criteria are suggested: - Apply the measure restrictively, that is, only when it can be clearly demonstrated that a strong variation in domestic prices would have very high economic or social cost for the country as a whole, and only when international prices can be easily and clearly determined. - Ensure that the price range is broad enough that variations in international prices ace usually reflected in domestic prices, except in times of truly severe fluctuations. - Apply the range only through a flexible customs tariff, avoiding the intervention of buyers (ENCI or ECASA). - Use the system only to moderate price changes arising from variations in Wmational prices, not .from changes in the real exchange rate or other factors. The argument for price bands is grounded in the potential positive shift in the supply curve for products that may occur when farmer 1111cednty about prices is reduced. This effect may in theory more than compensate for any resource misdlocation arisihg from the blunting of the opportunity cost signals of international prices. Chide's seven years of experience with price ranges on wheat, oilseeds, and sugar support this positive view of the eiTects on production (Espejo and Fontaine 1990). Each situation needs to be examined on its own merits, however. In the case of wheat, for example, distribution costs and sociopolitical factors must be considered as well. Other price and nude policies Liberalization of the domestic market for agricultural products and inputs, both tradables and domestic, should accompany trade liberalization. Subsidies on agricultural inputs provided by the public sector, such as water for irrigation, should be eliminated to stimulate efficient use of productive resources. Because liberalization of agricultural product and input markets is likely to reduce the income of poor farmers, especially those in the Andean region, special assistance-technological assistance, improvements in infrastructure and fiscal measures-may be needed to compensate for the abandonment of redistributive price policies. Without judging the merits of special subsidies on rice and yellow corn production in the jungle region, which are designed to encourage alternative crops to coca lead, several comments can be made about them. Fit,to reduce distortions, the subsidies should be extended to d l crops in the region: there may be other crops that would require lower subsidies. For the same reasons, the transportation subsidy should also be extended to other crops. Because this region is so far fiom major markets, high transportation subsidies are needed for unprocessed products. Subsidizing the establishment of agmindustrial processing plants in the zone could be more cost-effective than subsidizing the trausport of the raw products to distant plants. If, as seems likely, the subsidy on the processing plant would be less than the difference between the transport costs for processed and unprocessed products, the subsidy might be economically justified. State trade organizations should continue the process of general disengagement from direct economic activity. Instead, they should devote their regulatory attention to indirect actions to promote competition, provide technical assistance to producers, support producer organizations seeking greater ' transparency in trade processes, create adequate price i n f o d o n mechanisms, and provide adequate commercial infrastructure (transportation, central markets). Any remaining state trade functions should be privatized to ensure unambiguous and efficient management, or disbanded. Otherpolicies The implementation of trade liberalization measures should lead to important changes in the structure of production and land use. To encourage efficient implementation and to minimize low  adjustment costs in resource reallocation within agriculture and between agriculture and other sectors, certain complementary policy changes are proposed. There is broad agreement between officials and sector analysts Winisterio de Agricultura 1991) that public investment in the agricultural sector should move away from large and expensive investment projects designed to expand the agricultural frontier-large irrigation projects on the coast and colonization projects in the jungle-toward short-term, less expensive projects designed to increase the productivity of land already under cultivation. Technology transfer, farm promotion, agroindustrial development, and trade infrastructure seem to be the appropriate types of investment projects. Agricultural credit subsidies should cease because they violate the principle of neutrality between industry and agriculture. Directed credit programs should be dropped as well. Technical assistance for Peruvian agriculture must be more intensively provided than in the past. The great variability in agroecological conditions, the wide dispersion of producers, the small scale of most agricultural protection and the low level of farmer education call for more intensive research and technology trausfer assistance by the state than would be desirable in some other countries. These conditions create substantial externalities and difficulties for the private sector in adopting new technologies, requiring more active state involvement in technical assistance for the agricultural sector. Policies to increase flexibility in land use are also important. Land tenure policies of the past have restricted the free trausfer of land resources. So that agricultural production can respond appropriately to the trade liberalization process, flexibility in the trausfer of land is needed to ensure that agricultural adjustment occurs as rapidly and efficiently as possible. Already, 'there is a sufficiently developed land market in Peru, covering the whole country, which operates informally, without legal support for property titles and their wrrespondiig registry' (Chaquilla 1990). The next step would be to recognize that market and to legdize it (Qlaquilla 1990, and the World Bank 1990). The authorities have taken an important step in that direction with the approval of D.S.009-91 AG in March, 1991, which establishes freedom for land transfers and allows for land ownership and control by any type of individual or organization.. Making liberalization irreversible Experience in Peru and elsewhere (Choki, Michaely, and Papageorgiou 1990) has shown the impomce of strong initial steps toward trade liberalization to fortify the new regime and ensure its sustainability. The current trade liberalization program is the second in the last 25 years. The first program was implemented between 1979 and 1982. Trade reform was not, however, accompanied by adequate fiscal policy, and the fiscal deficit rose from 1.1 percent GDP in 1979 to 11.6 percent GDP in 1983. The expansive fiscal policy encouraged imports and discouraged exports, speeding the loss of international reserves. The fiscal crunch led to the erosion of the tariff reforms and in 1982 a 15 percent surcharge was added to the tariff. Similarly, exchange rate policy did not properly complement trade liberalization. The inflationary pressures generated by the fiscal deficit made the exchange rate difficult to manage and favored overvaluation of the currency (27 percent in 1982, Nogut%, 1985). As poor macroeconomic management led to increasing losses of international reserves, the government applied more restrictive policies. Both fiscal pressures and special interest lobbying led to the reversal of trade reforms that began in April 1981, when special treatment was granted to the state company, SIDERPERU. By the end of 1984 a new import prohibition list was established, and in 1985, the list of products requiring prior licensing was expanded and tariffs were raised. F i and multiple exchange rates were introduced as well. Expansive economic and exchange rate policies led to increasing demand for imports and stagnation in exports, and a growing loss of net intemtbnal reserves. As demand for imports grew, restrictive mechanisms multiplied; by September 1987, prior import licensing was required for all tariff items. The restrictions on imports did not end the loss of international reserves or prevent the subsequent need for devaluations and stabilization effbrts, but led instead to inflation and recession, which sharply reduced demand for imports (see appendix 6). The lessons of this experience are clear. To endure, trade liberalization requires coherence among exchange, tariff, and other policies and confidence on the part of economic agents (consumers and producers) that the rules of the game will remain the same. This can be achieved only through sound macroeconomic management, particularly by maintaining low fiscal deficits that do not stimulate inflation and that encourage stability and ideally, some depreciation of the real exchange rate. The Central Bank has intervened in the money market on several occasions-and may be pressured to do so again-to maintain or raise the 'real' exchange rate. While the resulting devaluations could make tariff reform easier, they would be counterproductive for stabilization objectives. The price of  tradable goods would rise and the purchase of dollars at a higher price would inject intis into the money supply, necessitating a Central Bank squeeze on domestic credit in order to achieve its monetary growth targets. Moreover, the benefits of a rise in the real exchange rate would be. temporary, especially in a context of rapid price adjustments (and overshooting) and the authorities would run the risk of destabilizing inflationary expectations as well as inflation. . Consequently, the authorities should minimize their interventions in the money market and instead take action to influence the real variables that affect the real price of money-reducing the size of government, privatizing public companies through debt-for equity swaps, reducing the price of public services to the long-term sustainable price, and leveling tariffs at 15 percent. A system of directed flotation in the exchange rate (and avoiding the reappearance of multiple rates) is the way to preserve coherence between monetary policy and exchange rate policy, and preserve a relative uniformity of incentives. Several other issues are also important to the sustainability of trade liberalization. The 5 percent tariff rate must be eliminated to prevent its becoming the weak point through which the entire tariff structure is breached, as it was in the 1980s. when pressure groups pushed to add to the number of tariff items and gradually increased effective protection. The 5 per- seems to have been introduced to assist the restructuring of the state company, SIDERPERU; it would be preferable to compensate the company directly from public funds for the payment of the o r d i i rate of 15 percent. Specific duties, reference prices, and compensatory duties are instruments on which power centers have always focused pressureand will continue to do eo in the future. Their treatment in law must be clarified as soon as possible and implementation methods m&t be transparent for all economic agents and bear no trace of a protectionist spirit. To avoid any return to prior import licenses, exchange licenses, or the innumerable nontariff measures suspended or e l i i since August 1991, the many Supreme Decrees dealing with these instruments must be by a law of parliament consolidated or promulgated by the executive on the basis of legislative faculties granted by the legislature. That same measure ought to raise the costs involved in reestablishing such instruments and establish the means through which economic agents affected by such measures could petition for their removal. Legislation must also protect the tariff structure from the a r b i i pedorations caused by tariff exemptions. During the liberalization expesience of 1979-81, the systematic granting of exemptions left domestic production, which was supposed to receive full tariff coverage, unprotected and led to an erosion in confidence in tariff reform. Any desire to favor a specific sector should be made explicit and should be achieved through budget transfers which make the social cost of special protection clear. One way to solidify reforms would be to commit Peru at the international level, through the multilateral negotiations of the Uruguay Round. International commitment by Peru would help make the trade reform more stable and strengthen Peruvian authorities against pressures from special interest groups, since new protectionist measures would be a violation of an international agreement.  References Abusada, R., (1990) "Consideraciones en Torno a una Apertura del Comercio Exterior en el Penl" mimeo, May. Abusada, R,, (1990) "Una Propuesta de Apertura del Comercio Exterior para el "Penl", mimm, Econsult, Lima, July. APOYO, (1990) "Factores que Afectan la Eficiencia de la Industria Manufacturers en el Penl", October. AMAT International (1990), Production and Marketing Analysis of Non-traditional Agricultural Exports, Report to the United States Aid Mission in Ped, L ima. Amat, C. y Pat, L. (1991) 'La Alimentacidn y el Sisterna Econdmico' Lima,Banm Central de Reserva del Penl, Nota Semanal6, Lima. Asociacidn de Plantas de la Industria Automotriz del Penl (APIA), (1991), Boletin 28, Aiio XXIV, Torno XXI, January. Bolob, C., (1976) 'Tariff Policies in PenS: 1880-1980'. disertacidn doctoral, Universidad de Oxford. Castro, P., (1990) 'Situacidn Actual y Perspectivas de la Industria de 10s Fertilizantes en el P d ' en h p a m a de la Naciones Unidas para el Desmllo 'Programs de Reestrucavacion Industrial (1990-2000) - Subsector F e r t i l i e s ' Lima, DPPElU89N17, March. Chaquilla, 0.. (1990) M e d o I . & de 2lerras A@colas en el Medio Rural, IICA, Lima. Dombusch, R., (1976) 'Expectations and Exchange Rate Dynamics' JournaJ of PolfrialEconomy, December. Edwards, S., (1989) Real Ehchange Rates, Dduation and @ustment, (Fa Press, Cambridge, Massachusetts). Escobal, J., (1989) Politicas de Predios y Subsidies Agricolas; Irnpactos Mameconomico y Sectoral, PenZ 1985-1989'. GRADE, Documento de Trabajo 5, Lima. Espejo, A. y Fontaine, E.R, (1991). 'El mecanismo de una banda de precios para el trigo y su aplicacion en Chile 1974-1990.' C b d e m s de Economia, 84, August. Geng, R., (1990) Gcstion Empersarial'del Estado en la Cornerciulizadon de M u c t o s e Imtnos Agricolas en el Pent, M i o de Agricultura, Grupo de Analisis de Politica Agraria, Lima.  Institute Interamericano de Cooperacion para la Agricultura (IICA), (1990). El Sector Agropecuario Peruano. Situacion y Perspectivas para su Reactivation. Lieamientos para un Programa Sectorial Agropecuario en el Ped: Diagnostic0 y Propuestas, IICA, Lima. Karmokolias, Y., (1990) "AutomotiveIndustry Trends and Prospects for Investment in Developing Countriesw IFC Discussion Paper Number 7. Krueger, A.O., Schiff M. y Valdes, A. (Ed.), (1990) Economia Politica de las Intervenciones de Precios Agricolas en America Larim, (Centro Internacional para el Desarrollo Economico, Santiago, Chile). MACROCONSULT (1991) Boletin Economico, March. Misterio de Agricultura, Oficina de Planificacion Agraria, (1991) Politica Agraria: Estrafegia, Objetiws, y Pricipales Acciones de Politica, Lima. National Society of Exporters, (1990), Annual Report. Nogues, J., (1991) Chapter 3 in Papagiorgiou, D., M. Michaely and Chobi, edc. LfkdWng Foreign IFode, Vol 4 (Oxford: Basil Blackwell). Nogues, J., (1990) "Trade Policies, Liieral Institutions and Growth: A Proposal to Sh2t PenS to an Open Economy', mimeo, November. Programa de las Naciones Unidas para el Desarrollo (1990) 'Programa de Reestructuracion - Industrial (1990-2000) Subsector Fertilizantes' Lima ,DP/PER/89/017, March. Quispe M. F., F., Zenon, y. C. Barrera (1990) 'Politica Comercial y Tipo de Cambio Real de Equilibrio: el Caso Peruano (1980-1990)'. Banco C entral de Reserva, L i i P d , September. Revilla, C. V., (10990) 'El Proceso de Liberalizacion Comercial 1978-1983: Lecciones de Una Apertura Fmtrada', Fundacion Friedrich Ebert, August. . Rodriguez, A., y Sjaastad, L. (1980) 'Atreso Cambiario en Argentim Mito o Reatidad?' Ensayos Economicos, 13, March. Rossini, G., (1991) 'Las Medidas Recientes de Apertura Comercial', La Moneda, 33, March. Rossini, G., (1990) 'Apextun Comercial, Estabilizacion Economica y Desarrollo Economico', mimeo, Gerencia de Investigation Economica, Banco Central de Reserva del PenS, Lii. World Bank, (1990) P d AgricuWd Sector Review,'mimeo, Washi11gton. Valdes, A. y Leon J. (1987) 'Politica Comercial, Industrialization y su Sesgo Antiexportador: PenS 1940-1983'. Ckdemos de EconbmtO, 71, April. APPENDIXES  Appendix Table of Contents Appendix 1. Calendar of the trade policy changes between August 1990 and June 1991 Appendix 2. The impact of reforms on the real exchange rate Appendix 3. Non-tariff barriers Appendix 4. International agreements: the Andean Pact Appendix 5. Protection in the motor vehicle sector Appendix 6. The restructuring of the customs system Appendix 7. The political economy of l i b e r a l i n  Appendix 1. Calendar of the Trade Policy Changes Between August 1990 and June 1991 Imports 'August 195W Tarifi Maximum tariff reduced from 84% to 50% and minimum tariff from 15% to 10% @.S. No. . 228-90-EF, Arts. 5, 6, 7, 8). Exemptions Minimum tariff applied to 39 exemption regimes @.S. No ,228-90.EF, Art. 5). Nontari'f'bammers Number of import prohibition tariff ranges reduced from 539 to 13 @.S. No. 228-90-EF, Art. 10). Prior import license, nocompetition decision, and other nontariff import measures, except those necessary for sanitary reasons, suspended until December 31, 1990 @.S. No. 228-90-EF, Art. 11). Tar@ Temporary surcharge of 10%on imports subject to the 22% and 50% rates and all other surcharges eliminated @.S. No. 257-90-Ef, Art. 4). Exemptions Tariff exemptions limited to only five regimens (educational centers, treaties, donations, duty free zones and BCR), plus exemptions for Tax Stability, Non-traditional Export Contracts, and the Stock Replenishment Regimen @.S. No. 257-90-W, Art. 2 and Circular No. 4691-90- INO/SUNAD). Nontari!barriers All import prohibitions eliminacd (D.S. No. 257-90-EF,Art. 5). n any condition (D. Introducing waste matexkd into the country prohibited, from any origin and i Lg. No. 611, Art. 17). November 193W Tarii The 10% ad valorem cif surcharge that affected imports subject to the 25% and 50% rates is eliminated (D.S. No. 294-90-EF, Art. 2). Exemptions ., The exemption for goods brought in by Peruvians returning from diplomatic and official posts or positions as functionaries of international organizations is restored (D.S. No. 294-90-EF, Art. 5) The suspension of duties on imports of capital goods from companies with Non-traditional Export Contracts is extended (D. Lg. No. 617, Art. 3). Nonta@fbaniers . Oleaginous inputs excluded from the list of products to be imported exclusively by ENCI @S. No. 294-90-EF, Art. 8). Invoices for automobile imports are to be authorized by the Consul @.S. No. 29440-EF, Art. 9). = Usedgoads Used capital goods may be imported without prior authorization of decision, and garbage compactors may be imported by local gwemments. 'Ibis leaves intact the prohibition on . importing used passenget transport and cargo vehicles, except as donations @S. No. 294-90- EF, Art. 7). = Tdfi . Calendar for the progressive reduction and unification of tariff rates within four years @S. No. 322-90-El?, Art. 1). = Nonta@Ymdem . Wheat flour is included in the list of products to be imported exclusively by ENCI @S. No. 149-90-EF, Art. 15). . State companies authorized to import food products through third parties @S. No. 31940-EF, Art. 1). State company wheat flour imports through t h i i parties are subject to the s i m u l t a ~ ~ ~ u s importation of 20% of bran @. S. No. 319-90-EF,A rt. 1). A margin over the cif value of state company imports through third parties, which will fonn part of the AgriculW Reactidon and Nutritional Security Fund, is eaabliied (FRASA) @ S .. No. 31940-EF, Art. 2). .  Used goods Imports of used vehicles for surface passenger transport exceptionally authorized; likewise for CKD packages (D.S. No. 165-90-EF). Agricultural products State companies responsible for food products importation are authorized to operate through third .. parties @.S. No. 319-90-EF, An. 1). January 1991 Tari,frs An ad valorem cif rate of 5% is approved for 25 tariff items of steel processing inputs @.S. No. A Supreme Decree project is published to modify the ad valorem rate for 141 tariff items of industrial inputs. Nontmy barriers The suspension of the Prior Import License, the No-Competition Decision, and other prior conditioning factors is extended until July 31,1991 @.S. No. 353-90-EF). A Supreme Decree project is published to control the manufacture, importation, and sale of 22 chemical inputs; the decree proposes creating a Special Registry in SUNAD and establiiing import quotas. The prohibition against imports of paper and carton residues for industrial r e c y c l i is suspended until June 30, 1991 @.S. 036-91-ICIWWD). 'Ihe lists of food products to be imported exclusively by ENCI (wheat, wheat flour, white sugar and raw sugar) and by ECASA (rice) is made more precise @.S. 00141-AG). The Technological Meat Processing Regulations are changed and the technical specifications for presentation and quality for meat and subproduct imports are established; a National Registry of Meat and Subproduct Importea is created (R.M. No. 00241-AG). TwiB The rate is reduced from 25% to 15%, and the items subject to the 50% tate are reduced to the 15% or 25% rates; the 5% rate for steel processing inputs comes into effect @.S. No. 01191- EF, Art. 1).  Exemptions Ad valorem cif exemptions for goods imports, except for capital goods considered in Non- traditional Export Contracts, are eliminated, according to Art. 16 of D.L. 22342 (D.S. No. 033- 91-EF, Arts. 5 and 6). Tax exemptions in favor of state companies are eliminated @.S. No. 067-9 1-EF, Art. 40). .?I Nonranf barriers The Prior Import License, the No-Competition Decision, the National Registry for Import Companies, the Unique Import Registry, the Import Registry of the Andean Group, and other conditioning elements prior to importation are eliminated, except those established for sanitary reasons, defense of the cultural heritage and the environment, protection and conservation of the flora and fauna, and defense of internal order and s d t y @.S.N0066-91-EF, Art. 1 and Circular NO. 46-45-91-SUNAD). The exclusive control by state companies of food products and inputs imports is eliminated @.S. No. 066.91-EF, Art. 1). The Sanitary Regulation for the importation of vegetable products and subproducts is modified (D.S. NO. 00991-EF). The Technological Meat Processing Regulition, which establiies requirements for meat and sub-product imports and for live beef for slaughter is modified (D.S. No. 003-91-SA). The National Obligatory Technical Norms are eliminAtnrl and the new version is limited to products that affect the physical security of pasons (D.S. No. 006-91-ICI'I). It is esabliied that the importation or exportation of bulk or semielaborated gold, together with gold obtaixied as a diiect result of a mining or metallurgic process, will be.totally free (D.S. No. 005-9 1- E m . Eumpttons The 70% reduction in import duties for the personal effects of Pennrians returning to stay in the country, as, for example, functionaries a d o r profbionals with contracts, is diminntnri (D.S. NO. 028-91-EF). The preferential regimen of the Pemvian-Colombii Customs Agreement is restored, suspeuding restrictions on the consumption of goods within the region designated by the agreement @.S. NO. 108-91-EF). The ad valorem cif exemptions for imported capital goods considered in Non-traditional Export Contracts made prior to DL.No. 033-91-EF and in contracts made after but whose petifions were presented prior to that law are maintained (D.S. No. 087-91-EF).  Nontanr bam'ers Sanitary requirements for wheat flour, soy cake, raw cotton, wheat, oats, barley, grain sorghum, dried vegetables, spices, fresh and ornamental fruit are approved, requiring a Sworn Statement, an official Sanitary Certifiqte from the country of origin and an inspection by the Vegetable Quarantine Inspector (R.M.2 10-91IAG). Sanitary requirements for imports of live mammals, semen, bovine embryos, red meats, subproducts, liquid and powered milk, cheese, sausage and gelatine, chicken, ducks and BB turkeys, eggs, and frozen chicken are approved, requiring a Sworn Statement, an official Sanitary Certificate from the country of origin and an inspection by the Animal Sanitation Inspector (R.M. 0231-91-AG). Sanitary requirements for imports of bulbs, tubers, vegetables, botanic seed and vegetable material for propagation are approved, requiring a Sworn Statement, an official Sanitary Certificate from the country of origin and an inspection by the Vegetable Quarantine Inspector (R.M. 0259-9 1-AG). Sanitary requirements for imports of animals, fish products and subproducts are modified and the -sanitary Import Permit is substituted with a Sanitary Certificate from the country of origin (R.S. 0011-91-AG). Administrative restrictions of imports of mimeographs and the l ib,including photocopiers, are eliminated @.S. No. 090-91-EF). It is made clear that the h i g n rrade functions of the National Rice, Coffee, Wheat, and M i Products Boards and of the Foreign Trade of Hydrobiological Products Agency ate abolished @.S. NO. 088-91-PCM). Used g& The operating procedure to be followed by customs for shipments of used goods is approved (Res. of the Superintendency No. 002878 and Directive No. 7-D-8-91-SUNAD). Agriclrl~alptoduas It is made clear that the compensatory surcharge does not exclude imports from countries with which Peru has established trade agreemeats w t ih tariff advantages @.S. No. 093-91-EF). May 1991 ~ t i o n s Benefits in favor of the handicapped for importing automobiies w ith special equipment whose cif value does not exceed USSb,000 are extended until Dec. 31, 1991 @.S. No. 111-91-EF, Art. 3).  Nontariibarriers Pesticide imports are subject only to the presentation of a Sworn Statement in which the registry number of the product is specified or the Sales Certification fiom the country of origin for non- registered products is presented (R.M. No. 276-91-AGDGA). Veterinary products are subject only to the presentation of a Sworn Statement in which the registry number of the product is specified or the Sales Certification fiom'the country of origin .. for non-registered products is presented (R.M. No. 277-91-AGDGA). Introduction into the country of organochloride pesticides (ALDRIN, ENDRIN, DIELRIN, BHCMCH, Canflecorotoxafeno and heptacloro), derivatives, and compositions is prohibited 0 . S . NO. 0022-91-AG). Agriculturul products Preferential tariffs granted in trade agreements for food products and inputs imports (meat and subproducts, powered milk, butter, and cheese) are temporarily suspended 0.S. No. 017-91- AG). A specific duty expressed in US$ per Metric Ton on food products and inputs imports is established, in connection with price variations in the interdona1 market, for products classified in 18 NANDINA items and the compensatory surcharge is eliminated. This tax is part of the taxable IGV base 0 . S . No. 0016-91-AG). Procedures for the application of the specific duty are approved, and fob reference prices for the application of that duty are published (Circular No. 46-41-91-SUNAD). Nm~MerS Remaining import restrictions are made mom precise: 1. Internment Cdficate (phamnceutical products for human use, ammtics, medical equipment). 2. Certificate of Sanitary Aptitude @ydrobiilogicalproducts). 3. Ministry of Foreign Relations authorbarion for geographical texts and maps. 4. Mitry of Transport and Communications permits for radii communications equipment. 5. M i of the Inmior (DIOSCAMEC) authorization for arms and munitions for personal use. 6. IPEN certification for food for human consumption ftom Europe. 7. Permit ftom the sector (ships and aircraft). 8. M i of Health certification for interment of human remains. 9. National Institute of Culture approval (art object and replicas, wez 100 years old). 10. Sworn Statement for agricultural pesticides and the l i b. 11. Sanitation requirements for vegetable products and subproducts. 12. Zawmitary certification for animal products and subproducts  13. Certificate of Origin to apply for Trade Agreement benefits (Circular No. 46-52-91- SUNAD). Agricultural products Reference prices (fob) for the application of specific duties on food product and input imports are published (Circular No. 4-6-50-91-SUNAD). Complementary policies August 1990 Instinctional reform The Institute of Foreign Trade (ICE) is deactivated and its functions, material resources, personnel, and files are passed to the Office of the President of the Council of Miters. Activities linked to international treaties in process continue until ICE is fully deactivated. Trade offices overseas are taken over by embassies and consulates (D.S. No. 099-90-PCM, Art. 2). A chief of ICE is named, for the period of its deactivation (R.S.No. 25640-PCM). A commission is created to make recommendations to the Resident of the Council of Miters on the deactivation of ICE and the period of transition. The chief of ICE, during the institute's period of deactivation, is charged with the responsibility of maintaiaing ICE services until it is fully deactivated, accord'ig to the recommendations of the commission, and of supporting the Social Compensation Program (D.S. No. 100-PCM). Integration policy The Administered Trade Roles signed with Colombia and Venezuela are diminated and, during 1991, that with Ecuador is maintained, with 10 tariff items @.S. No. 00341-ICXWG). i s One hundred items are withdrawn from the Exemptions Lt applicable to Bolivia, Colombia, and Venezuela (R.M. No. 012-91-ICTI/IG). Mjustmentpoliq Fixed-term labor contract regulations are modified (R.M.No. 032-91-EF).  The Regimen of Compensation for Length of Service is modified so that the worker will receive monthly benefits payments in a financial instrument (savings account) @.S. No. 0155-91-TR). The Labor Stability Law is modified, increasing motives for dismissing workers @.S. No. 016- 91-TR). The regimen of farm land as collateral is modified, allowing its use in the granting of commercial ., and bank credit to farmers @.S. No. 039-91-EF). Free availability of farm land is established and ownership and conduction of land may be in any corporate form @.S. No. 009-91-AG). The Regulations for the Treatment of Foreign Investments and Contracts for Foreign Technologies, Patents, Brand Names, and Licenses is approved (R. del Dir. de CONITE No. 003- 9 1-EFnS). Insurers are allowed to freely contract second policies domestically or overseas, according to SBS @.S. No. 038-91-EF and R. SBS NO. 164-91). SUNAD Circular related to the elimination of Administered Trade Roles (Circular No. 46-23-91- SUNAD). InstiMonal~nn ' The new chief of ICE is named (R.S. No. 140-91-PCM). Overseas ICE offices, except in Washington, D.C., and Brussels, are closed @.S. No. 074-91- XM). Compensated trade operations are no longer subject to bureaucratic control or administrative approval @.S. No. 075-91-PCM, Art. 1,2,3). Decrees govexning the System of Supervision of Imports and Exports are annulled @.S. No. 075- 91-PCM, Art. 4). Nine commissions of foreign trade are dissolved @.S. No. 076-91-PCM). Sanctions for Mtactions of the delivery of boreign cummy regimen end coedve collecting procedures are suspended @.S. No. 077-91-PCM). ICE attriiutions related to debt payment agreements in kind are transferred to the Ministry of Economy and Finance @.S. No. 078-91-PCM).  The administration of the Tax on Non-traditional Exports Recovery Regimen is transferred to the Mi nistry of Economy and Finance @.S. No. 079-91-PCM). A 4 u s m n t policy Tbe regulations of D.S. No. 015-91-TR, related to compensation for length of service, are approved @.S . No. 022-9 1-TR). Tbe Regulation for the Treatment of Foreign Investments and Contracts for Foreign Technologies, Patents, Brand Names, and Licenses is adjusted to Decision No. 291 of the Andean Group (R. del Dir. de C O W No. 04-91-EF/35). Integration policy Sunad Circular related to the withdrawal of items from the Peruvian L i t of Exemptions, applicable to Bolivia, Colombia, and Venezuela (Circular No. 46-33191-SUNAD). May,I991 IrmimionaZ r @ m It is made clear that the functions of ICE related to CERTEX have been suspended since D.S.No. 079-91-PCM took effect @.S. No. 112-91-EF). Mjustment poliq The Regulation for Len@ of Senrice is modified @.S. No. 024-91-TR). Agrarian cooperatives and corporate firms are allowed to choose their business model freely @.S. NO. 018-91-AG). June, 1991 = I ~ o o n a l ~ n n Ministry of Economy and Finance representatives are named to the Transfer Commission, responsible for handling the documentation and records related to edminifitrative functioning of CERTM, which w as taken over by the Ministry of Economy and F i( R .M .No. 248-91- EF43. a ) . The Ministry of Industry, Foreign Trade, Tourism, and Integration substitutes for ICE as the state representative in the stock capital of the Export Credit Insurance company (SECREX) @.S. No. 130-91-EF). Zntegrasionpoliq Decisimns 295 and 2% of the Commission of the Cartagem Agtcemera are publied in the Official Joumal, El Pmuuu, (R.M. No. 199-91-ICIWIG). August, 1m w o n tax An extraordinary contribution from all exporters to the Emergency Social Program, equivalent to 10% of fob exports value, is established @.S. No. 232-91-EF). Inceni tw s for exports The basic CERTEX percentage is reduced to 10% fob value. For purposes of decentralization, CERTEX is limited to only agrarian and agroindustrial products. The basic CERTEX percentage is reduced to 20% fob value for handiwork products. Complementary and Additional CERTEX are suspended. The unique CERTEX for clothes is limited to 10%fob value. The CERTEX for aquiculture is limited to 10% @.S. No. 22891-EF, Art. 9). Rponiax A tax on foreign money for traditional exports is created, with a rate of 5%fob value, with the exception of Large M i g exports, for which the rate is 110%. All other taxes on exports are abolished @.L. No. 622, Arts. 1,s). Iitmtiwsfir~m For reasons of decatmlization, CERTEX is extended to cover aquidture production @.L. No. 622, Art. 8). I-*wm It is made clear that the eliminntion of CERTEX extends to exports after DL.No. 622 takes effect@.S. NO. 33391-EF). lapontax The tax on foreign money for traditional exports i s suspended, as well as fw directs exports of d v e r in the fonn of ore, concentrate or refined, as long as the Handy and Harman average quote for the last three months is not over US% the Troy ounce @ .S. No. 016-91-PCM). March, 1991 &port restrictions All export restrictions, including those which condition exports to prior supply of the domestic market, are eliminated (D.S. No. 060-91-EF). Listings of export restrictions are eliminated and only export prohibitions are maintained (R.M. . NO. 118-91-EFnO). Because of the cholera epidemic, farm and agroindustrial products listed in 17 NANDINA sub- items are subjected to sanitary control (D.S. No. 059-91-PCM). Exemption petitions for foreign money delivery are simplified @.S. No. 037-91-EF). Incentives for exports A transitory mechanism to allow nontraditional exporters to recover taxes on their export production is adopted (fiscal credit) @.S. No. 052-91-EF). The regimen of Temporary Admission for merchandise, to be used in processes of 'active improvement' for subsequent exportation, with duties and all taxes suspended, is modified and simplified @.S. No. 034-91-EF). ' Export tax 'Ihe tax on foreign money for traditional exports is replaced by a tax on the expbrts themselves, is to be paid within 15 days of transaction payment, with rates of 0.5% and 10%for mininp products (copper, silver, zinc, lead, and their concentrates) and 5%on other nomining traditional exports (in effect until Dec.31, 1991) @.S. No. 084-91-EF). Mid-sized mining producers pay the export tax until June 30, 1991 @.S. No. 101-91-EF, Art. 2). ExponremStrZ&ns Sanctions for inhctions of the foreign money delivery norms are suspended, along with coercive collection procedures @.S. No: 106-91-EF). Imentiwsfor qorts The tax recovery mechanism is extended to include silver mining producers (D.S. No. 08641-EF, Art. 3). Domestic firms that supply exporters are allowed the benefits of the Tempomy Admission Regimen @.S. No. 086-91-EF). May, 1991 Export restrictions Obligations concerning foreign money delivery assumed during the period when the relevant regimen was in effect are valid and must be fulfilled according to the norms in effect up to the publication of D.S. No. 077-91-PCM (D.S. No. 118-91-EF). . Ten NANDINA sub-items are included in the list of products subject to sanitary control because of the cholera epidemic (Res. N* 092-9 1/ICE-Jefatura). Remaining export restrictions are clarified (D.S. No. 128-91- EF): 1. Unique Text of Prohibited Bports 2. Certification for textile products with quotas 3. Products subject to sanitary control because of cholera 4. Protection of Cultural Heritage 5. M i of Health certification of expatriation of human remains.  Appendix 2. The Impact of Reforms on the Real Exchange Rate The effect of tariff reform on relative prices The analytic framework presented here involves four types of goods: exportables, importables, non-tradeable goods (services), and goods produced by State companies, the prices of which are fixed by the Government. It is supposed that exportable and importable goods obey the law of only one price, due to goods arbitrage. That is: where P , is the domestic price of imports in intis, E is the exchange rate (the price of foreign currency in intis), Po,, the price of imports in foreign currency, and t,, the ad valorem customs duty or the tariff equivalent to quantitative restrictions (QRs). Similarly, the domestic price of exportables is indicated by: where t, represents the tax (ii percentages) on exports (mainly taxes on minerals). A tariff reform involves a reduction of the customs duties applied to imports, 2,. Measuring the relation between prices for exportables and importables, the increase in the relative price of exports is indicated by: Therefore, the percentage change in the relative price of exports is indicated by:  A tariff refonn that reduces t,, or a change in taxes on exports that lowers 1, , will change the price of exports in terms of imports. Changes in the terms of trade (international price of M relative to X) also play a role in the relative domestic price of exportables in terms of importables. The signal given to the economy in terms of relative prices is supposed to indicate the required change in resource allocation. . . It should be noted that the relative price indicated in equation 4 does not depend on what occurs in the nominal exchange rate, given that the price of exports in terms of imports (P,/P,,,) is independent of the level of the exchange rate. The relative price signal is, thus, given by 4: that is, the relative signal for the economy for resource reallocation. Given that, in practice, the price of importables is a weighted average of imports prices, including each relevant tax rate, the relative price of exports rises in a weighted average of tariff reductions. This can be calculated in the following way: on moving to a lower and flat rate of 15 percent customs duty on imports, it is supposed that the reduction is from a level of 50 percent for some gcpds, from 25 percent for others, and from a tariff equivalent to 100 percent for quantitative restriCti0ns.l The weights supposed for tariff reductions are given in table A2.1. It should be noted that the analysis does not refer to the tariff reductions implemented by the new . . nwhenit entered office in August, 1990, but rather to those applied to tariff levels of M ar ch 1991. Moreover, and more important, this exercise must be understood as an illustration of the model under discussion and not as a precise description of reality. I n f o d o n on the tariff equivalence of all quantitative and nontariff restrictions is not available, and neither is the weight which ought to be given to that category in the following calculation. 1.Tboortimrtsof 100parsllt,umqpivrlautuifffor~roaierionr,wudaivaddrom implich~mrdutbr, . a l ~ ~ t b o p a e 4 a t y o 0 f t b o ~ ~ ~ P e r w k n d ~ ( U ~ ~ ) ~ i b r p r o a u o ~ l r p b ~ t o - -q . nrbuPltomobilsl,~~mpPten,taavtiwr,~,d~.Alnorotompkso~of current ntm of affectivepratccba m h an be hund m Aburdr (1990)d Rwrini (1990). Sse Bolob for a hizor)l of pmtucth i n POCP.  It h possible to create a model for the way in which QRs were made more flexible by considering this a reduction of the "equivalent" rate of protection. To estimate that rate, it is useful to compare domestic with world prices in order to evaluate the tariff which would be equivalent to a QR. The relaxation of QRs would, therefore, be equivalent to a reduction in the equivalent tax. In both cases- nominal tariff reduction and the elimination of the QRs-it will be supposed that economic agents perceive them to be permanent measures, not transitory. Tariff and QR reductions, by themselves, will permanently alter the exchange rate in the direction of real depreciation. The Effect of Tariff Reform and the Elimination of Nontariff Restrictions on the Price of Nontradable goods While it is possible to determine the domestic price of importables and exportables through price arbitration, the price of nontradable goods (represented by "S," for services) is determined by domestic supply and demand, or: , ,the price of gwernment services. A linear where EX represents expenditure, Y,income, and P logarithmic version of the functions of supply and demand can be assumed in order to derive the impact of tariff reductions on the relative price of services. Taking into account that the supply and demand functions are homogeneous of degree zero in real terms, nominal prices and values can be divided by the price of imports to convert them into relative prices. Therefore: It should be noted that the relation expenditure to income is used as a proxy in the measurement of the relation of total demand to total supply. Thus, the natural logarithm of equation 6 yields a linear model of the supply and demand of services, in which all quotients have signs associated w i t . the supposition that all goods are substitutes, both in dmand and supply. In order to calculate the change in the price of services arising from the tariff reform, the natural logarithm of equation 6 is taken and the resulting expression is d i i e r d a t e d over time. This yields, where a hat denotes a percentage change in a variable.This can be expressed in the "Omegaa notation as: in which It should be noted that 0,+0,+0, = 1, by the Hichian price effect, itself and crossed, in supply and demand, respectively. The percentage change in the price of importables due to reductions in tariffs and QRs is , in equation 1 and dBerenththg: discovered by taking the natural logarithm of P where & indicates the percentage increase in the price in Idir for foreign cmency and pl ,the percentage increase in the world price of importables. Therefbre, with no change in the world price of imports nor in the exchange rate, the domeaic price will fall by - At, . (1 +Q Similarly, taking the natural logarithm of P, in equation 2, and diffenmhhg, the percentage change in the price of exports is given by:  The percentage change in the price of exportables depends on changes in export taxes, the depreciation rate of the inti, and the rate of increase in the world price of exportables. , , Let p, = P,/E be the dollar price of goods offered publicly. When their price in intis, P rises more than the price of the dollar in intis, E, the dollar price of those goods, p,, rises. Therefore, the percentage change in the price of goods offered by the government (state companies, etc.) can be indicated by: Since fia represeas the percentage change in the dollar price of goods offered by the government, its increase is in relation to the exchange rate or to tradable goods. When the price of goods offered by state companies rises more than the rate of devaluation, fig is positive. This is a useful way of desmiing the rise in the price of gasoline, diesel fuel, electricity, etc.. given that the price increase is calculated in dollars, (i.e. in relation to tradable goods purchased). The key charaderistic of the stabilization plan of August 1990 was that the increase in the absolute price of goods offered by the government w as much greater t han the depreciation rate of the etxchauge rate, The short-tem impact on private expenditure on public goods will now be considered. The demand Qa = Q P , ~ for goods provided by the public sector as a function of their price in dollars, or maintaining income constant, can bedescribed. Therefore, private expenditure in dollars on public sector goods is indicated by Pa = QP:-U ,Therefore, the pacemtage chauge in private expenditure due to hikes in the price of public sector goods is equal to (1 tl - .It should & be noted that this variation in expenditure represents increased income for the public sector, if the demand elasticity is less than one (which is probable in tbe short term). Any increase in real expenditure (in dollars) represents, at the same time, less real income for private s e a r e q x d i . If the elasticity of demand for services with respect to gm&r real public sector income is indicated  by y , . and the elasticity of demand for services in relation to private sector income. by 7 , . the net effect of that income redistribution on demand is given by: in which O is positive, if , q is less than one and 7 ,>7 , . The transfer of income that occurs when demand is inelastic in the short term will lead to greater expenditure on services if the propensity of the public sector to spend on nontradable goods is greater than the propensity of the private sector to spend on domestic goods. That is, if y , > yp and 7, < 1. However, it is expected that that effect will be small, given that it is divided by the level of domestic income. The way has now been cleared for calculating the percentage change in the price of services by using equation (7). Simply, in equation 7, , P and , P of equations 9, 10, and 11 are substituted in equation 7, as is the increase in real e x p e n d i i due to the transfer of income dependent on demand elasticity. That is: (for this solution, recall that P, + P, + P, = 1 ). The difference in percatage changes in prices can now be calculated to obtain the variation in the relative prices of goods. For example, with no change in the terms of trade for goods nor in export taxes, there we have: which verifies equation 4. Similarly, subtracting equation 9 from equation 13 to obtain the increase in the relative price of services, we have: (A.15) $, -@ , = [Q, - 1 1- At, + [Pi+OI@~ . (1 tJ + Based on the supposed parameters listed below, with respect to the value of the omegas and the increase in the dollar price of goods offered by state companies, the increase in the price of services relative to imports is estimated to be 41 percent. The supposed increase of 150 percent in the price of goods produced by state companies, together with an omega for government goods equal to 0.2, explain that strong effect.l An omega of 0.4 is assumed for imports, and a tariff reduction of 18.5 percent with respect to the initial level of imports prices. The tenn iP was equal to zero. Finally, with the world price of imports constant, the percentage change in the relative price of goods offered by the government is given by subtracting equation 9 from equation 11: Supposing that the percentage increase in the price of goods offered by state companies is 150 percent, the increase in the relative price of goods produced by those companies in terms of imports is 168.5 percent. kf'fects of changes in the price of government goods and semias on the red exchange rate Consider the impact for tariff l i b e r a l i n , in a context of s t a b i o n , through an increase in the relative price of goods offered by the government. F irst, the real exchange rate is defined: where P.is the level of world prices, E is the inti price of foreign currency, and P is the level of Peruvian prices (consumer price index). With world prices constant, the percentage change in the real exchange rate is given by: The consumer price index, on the other hand, is given by a weighted average of the prices of export, imports, non-tradablegoods, and goods offered by state companies.  inwhich a , + a , + a , + a , = l = 1. Therefore, the Peruvian rate of inflation is equal to a weighted average of the individual rates of price increases: Substituting equations 9 through 12 into equation 20 and solving (supposing no change in the world price of exportables): Therefore, the p e r w e change in the real exchange rate is given substituting equation 21 into equation 18 and solving: Ba sed on the paramaer suppositions described below, the real m e appmcMon is equal to 48 percent. The e!f'fect of capital influx on the real exchange rate , Recalling that, since Y-EX= CA = the cumnt acwunt, can be replaced in the demand for services by 1-CAJY, in which CA/Y represents current account surpluses as pexcentage of GDP. A greater influx of capital, due to highex real interest rates in Peru, would imply a h i e r deficit in the current account (financed by the influx of capital). Wi floating = w e rates, htemational reserves would not be u sed to finance that greater current account deficit. Only the greaser influx of capital would b c e it. 'Ibis section analyzes the impact of a high current account deficit on the real =change rate. Recall that, with equation 7, it is possible to express the price of services as a function of the relation expenditure to income, EXtY, or, equivalently, as a function of 1-CAtY:  Therefore, the effect of a higher current accounts deficit on the price of services is given by: A higher current accounts deficit implies dCA < 0, so that greater influx of capital which lead to a .higher current accounts deficit will cause an increase in the price of services. This implies greater real appreciation. The effect on the price of services depends on the surrounding parameters and the = 1, the percentage , = tam worsening of the current account. For example, supposing that y = q change in the price of services can be derived as a function of the higher current accounts deficit as a percentage of income. The Central Reserve Bank of Peru estimates the 1990 current accounts deficit at $1.347 billion (Nora Semanal No. 6,7, February, 1991), which, as a percentage of income in 1989 dollars of $28.077 billion (World Bank Debt Tables, 1990-1991, Vol. 1 1). would yield CAN = -0.05. A increase in loans of 2 percent of income would, therefore, lead to an approximate increase of 1 percent in the price of services, according to equation (23). I .general, the following result is obtained: Percentage increase in Influx of capital as prices of services percentage of income 0.5 1 1 2 1.5 3 2 4 Naturally, dierent suppositions about the values of the parameters y , and em would lead to different results. With inaeased terrorist activity in Peru, such as the bombing of bauks, and the partial prior confiscation of dollar accounts, the entry of capital seeking greater ptofit in Peru due to the restricted liquidity caused by the stabilization plan will mainly be realized by residents who sell dollars to acquire Intis. It seems reasonable to guess that this will hardly be 3 percent or 4 percent GDP, so that the impact on the real exchange rate will not be great. To estimate the effect on the real exchange rate, that change could be weighted by the importance of services in the consumer price index. For example, with a, equal to 0.3, an increase in hreign loans equal to 3 percent GDP would lead to an increase of 1.5 percent in the price of services, which would add half a percentage point to domestic prices. That would mean a half of a percent greater real appreciation. Parameter Suppositions Weights Omegas Effective tariff Increase in the Effect of the CPI and RC reductions dollar price of propensity toward goods offered by expenditure and companies controlled redistribution by the State Note: The parameta suppositions should be replaced by others with more extensive empirical fo~otls.  Appendix 3. Non-Tarif'f Barriers 1. Import company registry in ICE. 2 . State monopoly of wheat, wheat flour and sugar imports @NCI), and rice (ECASA). 3. The Technical Regulations for Meats, which established norms, such as the obligation to pay technicians' travel overseas. Only those registered in the National Register of Meat Importers could import meats. 4. Phytosanitary and Zoo sanitary permits were solicited from the Ministry of Agriculture, before merchandise was shipped. 5. License for bringing in agricultural pesticides granted by the Ministry of Agriculture. 6. Certificate for veterinarian products granted by the M i t r y of Agriculture. 7. Certificate granted by the Mitry of Agriculture for animal foods and supplements. Importers wexe obliged to register with the Ministry of Agriculture. 8. Publication in the Official Journal and prior opinion of the M i t r y of Agriculture for wood products imports. 9. Sanitary Registration and Certification by CONAMAD for importing pharmaceutical products, cosmetics, perfume articles, medical equipment in general, and industrialized foods. 10. Prior evaluation by lTINTJ2C for a list of manufactured products. 11. Control of comts, weights, and measures of packaged products by m C . 12. Special authorization of the Ministry of the Interior for imports of mimeograph equipment and photocopiers. 13. Radiological certification by IPEN for foodstuff imports. 14. Ministry of Foreign Affairs authorization for importation of geograph texts and cartographic publications. 15. M i t r y of Transportation and Communications authorization for importation of communications equipment. 16. Specific decision for imports of the assembly industries which b e t t e d from lower tariffs. 17. Consular permits for the invoices for automobile imports. 18. Prior import license and no-competition (with domestic production) decision temporarily rmsp-Jed).  Appendix 4. International Agreements: The Andean Pact At the heart of the stagnation and failure that have characterized the Andean Group for many years is the desire to replicate the import substitution model at a broader regional level. The attempt was made in vain to include the entire tariff universe in the duty-exemption frameworks that were to create a free trade zone. It also proved impossible to arrive at a regional agreement on a common tariff barrier for the Group. For more than 15 years, unsuccessful attempts were made to reach agreement on a set of common barriers (which were excessively high and heterogeneous). Achieving a common foreign tariff (CFT)implies massive resource reallocation within the Group, with diverse costs and benefits, which naturally made negotiation long and arduous. A high tariff meant that the attempt was being made once again to implement the substitution strategy in a market which, despite substantial growth, still constituted an economy somewhat smaller than that of a small European country. The failure of the integration process can be seen in two revealing indicators: & t, in the mid-805, Group countries still maintained more than half of all tariff categories outside of the regional integration programs. Second, less than 4 percent of total Andean Group imports occur among region wuutfies. The beginning of the debt crisis toward the end of 1982 comibuted to further paralysis of the Andean integration process. Growing fiscal deficits, tmmatic devaluations, protectionist policies and rising interest rates spurred regional protectionism more than it did toward other countries. However, the debt crisis moved the economies of the region timidly at f irst and more boldly toward the end of the decade, to reorient policy toward greater openness. The end of easy access to foreign indebtedness led to more realistic exchange and fiscal policies. At the beginning of the 1990s, all these factors combined to create a propitious situation lbr renewing Andean impation effbrts-now with the common objective of achieving greater integration of member countries into the world economy. The broadening of the Andean integration framework makes sense for member countries only to the extent that it increases the level of openness of the individual countries to the rest of the world and generates a larger economic space, with imprwed conditions fix compe&ion within the world economy. The goal, therefore, is to increase the overall trade of member countries, which in turn requires a low and uniform tariff barrier. The economic opemess programs of all Andean Group countries have the same goal: 8ubstantial reductions in the number of tariff rates and a lower level of  protection. Thus, the shared national policies of Group countries now make the establishment of a free trade zone possible, along with the eventual adoption of a common tariff policy. The policy of trade openness adopted by Peru is compatible with the integration process in that the level of protection adopted seems to coincide with levels already prevalent or at least programmed for the majority of Andean countries. In this sense, greater Andean integration appears to be a beneficial side 'effect of the more general movement toward promoting increased international trade between each member country and the rest of the world. If such global openness prevails, a level of trade will be generated that will be greater than the diversion of trade that tariff unions tend to produce. The economic openness poiides of the Andean countries Trade reform in Peru has been similar to that in four other Andean countries (see tables A4.1 and A4.2). Along with Bolivia, Peru's reform has been the deepest and most rapid. Bolivia implemented its trade reform toward the end of 1985, Venezuela began in 1989, and Colombia and Ecuador, in early 1990. BdMrr The Bolivian trade reform consisted of the elimination of quantitative restrictions on imports and the adoption of a fixed, unitwm tariff of 20 percent, initially, with a gradual reduction to the present level of 10 percent. Lakz, a 5 pe rm tariff was established for capital goods. For exchange rase policy, a system of managed flatation k m g h ao auction mechanism was adopted. Venezuela The Venezuelan trade reform proposes t a M reductions over a five-year period, to ahinate in 1993 in the elimination of import restrictions ,anda maximum tariff of 20 percent, with minimum dispersion. Curmtly, the maximum tariff is 50 percent. With respect to the e x a g e rate, the differential system was eliminated and one freely detemined rate has beem adopted. Table A4.1 Protection in Andean group countria: distribution oftariff universe, Januafl- 11991 (percat) Rcur Bolina Ecuador Colombia Peru Vrnezuelo Avenge 9 .8 17.~ US@ 16.8 17.4 Much.forFbl. ' k Am. tbae i s a tu of 2 paceat (XAW 14of Jmuuy 24. 1989)and a tu of 1 pccaat (YAW 92ofApril19.1988). w h i i nirc tbemt.gentoto20.4pacent. - d Addodtotbc~uatuof8perccntto13pace~t(YAW%ofl~.~rhrairatbeavengeto36.5~ Tabk A42 NominJ prota&on for tbt hdlutrkl u&u, 1991 ~wrumc~ooo6 9.99 n.10 2131 21.27 33.52 Inpum and Fudr 9.99 15. 18.54 15.70 11.16 CrpW Ooodr 8.64 13.01 13.74 1529 18.13 ~ O D o d r Indumrid 10.00 16.97 U.42 18.13 8.07 Colombia Colombia plans to increase openness over five years by steadily elimimthg import restrictions and reducing tariffs toward 25 percent, with minimum dispersion. Currently, the tatiff univetse within the prior licensiog regimea is 24 percent. 'Ihere is only one exchange rate, with a system of mini- devaluations.  Ecuador The first stage of the Ecuadorian reform saw a restructuring of tariffs, with rates varying between 5 percent and 60 percent (with the exception of the rate for automobiles, which had been fixed at 80 percent). At the same time, tariff exemptions are being gradually eliminated. In the second stage, quantitative import restrictions will be eliminated and new maximum and minimum levels will be established. In a third stage, to begin in 1992, economic openness will be broadened. Currently, there are multiple exchange rates. Since the second half of 1988, a system of flexible exchange rates with weekly devaluations has been implemented for commercial transactions; for other cash transactions, a free market or parallel exchange rate is used. Progress in Andean integration The agreements reached in the Fourth Andean Presidential Council (held in La Paz in November 1990) were aimed at broadening the integration process by achieving the following: accelerating the aut omatic tariff reductions of the liberalizationprogram-the mechanism of a uto matic tariffreductions that will lead to the elimination of tarifi on intraregional trade--with the intent of ending the process on December 31, 1991; dimhating the Adminimred Trade Regimen, which establiied production quantities subject to biiateral negotiation; dimhating the lists of exception, comspondig to products covered by regional programs of industrial planning; and elhiwing the ex&tions lists corresponding to the tariff segments excluded from the integration process, by the end of 1991. The Common Foreign Tariff was reduced from an average of 28.5 percent to 17.9 percent and rates w ere reduced from 19 to 7, with a minimum of 0 and a maximum of 50 petcent. This means substantially reduced protection at the regional level. The schedule for defining of the Common Foreign Tariffhas been moved up to the end of 1991, so that by the end of 1995, the Andean Tariff Union will have been created. C o l d i Peru and Venezuela will implement it by the end of 1993 and Bolivia and Ecuador, by the end of 1995. Colombia, Peru, and Venezuela have already eliminated the admh&ed trade registers among themselves, mahtaMng 20 percent of the register with Ecuador during 1991. The lists of exception will be eliminated for memba countries by mid-1991. W ih respect to the exemption l t iss t,B o l i Colombii Peru and Venezuela ate accelerating in 1991 the process of d i i n g of the lists that apply among th&vcs. The lists are to be completely eliminated by the end of 1995, culminating the process of total I i b a a t i n of intraregional trade.  Although those decisions reflect considerable energy in the integration process, it is too early to evaluate their real effect on intraregional trade. Even though estimates for 1990 intraregional exports have risen 24 percent, to more than USS1.3 billion, they are only 4.4 percent of total Andean sales for that year. If fuel exports are excluded, intraregional export share rises to 8 percent, and if traditional products are also excluded, its share is 18 percent. Currently, the segment of the tariff a v e r s e still outside the free trade zone is very large, and trade volume accounts for two-thirds of Andean trade because of the continuing impact of exemption lists, lists of exception, and difficulties arising from violations of the agreements-especially those related to the common foreign tariff. Table A4.3 presents a list of agreement failures, highlighting the large number of violations by Peru and Venezuela. Table A43 Fail- in the trade and foreign tuitt libadzation pr0y.m t o April, 26,1991 (1Yllmber of tariff L i i i z a t i o n program ( E l i m i i a of duties) N o t produced E l i m i i of - 11 Fodituitf Common minimum foreign tariff 1 521 742 These agreement failures largely reflect the different rates at which Andean countries are disposed to l i b d i z e trade, not only with thud-party countries but with their Andean partners. They also imply that, as in the past, Andean-level decisions have not in the least hindered the application of national economic policy. Although it is undoubtedly true that Andean agreements have had some effect on individual policies, it is easy to verify that Andean decisions will only be adopted if they coincide with certain predefined national policies. If a low flat-rate common foreign tariff (with no exceptions), is agreed to, the tariff union could be achieved by 1995. Bolivia, Peru, and Venezuela are already on the way. However, it will be necessary to accelerate trade reform in each member country. The persistence of disparate tariffs, with the subsequent failures in the Andean Common Market, which is atready being surpassed by  national trade policies, could wreck the tariff union. Adoption of the CFT could help identify economies that are natural trading partners and thus advance their integration into the world economy. The proposal of the technical organ of the Cartagena Agreement remains to be determined. To date, the Agreement Board is tending towards maintaining the CFT technical proposal, using the same criteria as in the 1970s, which resulted in policies the countries refused to adopt. These criteria included conditioning the nominal tariff in terms of production levels or making concessions according to the technological externalities that certain products may generate. On the other hand, the tariff reform in member countries has led to consideration of a CFT with only three rates, at an average between 10 percent and 20 percent. In spite of the progress implied by such a CFT proposal, it is unlikely that agreement will be maAwd on the basis of a CET determined by criteria that are much different than those that inspired the import substitution model-industrial planning. Thus, within the Group, modern economic thought about the role of tariffs exists side by side with the tendency to preserve the old ideas of the substitution model. The ministries of industry support the latter course while the ministries of economy or treasury and the central banks strongly favor a more liberal model. To the degree that conservative tendencies maintain their influence, adoption of a common tariff barrier is unlikely. The main obstacle will probably be several countries*r h a l to raise tariffs for basic products and inputs above the current 1 percent to 5 percent levels, rather than problems with maximum rates. Once the conflict over the lower rates is resolved, it will be easier to adopt a CFT based on an average of the countries*tariffs, in much the same way as occurred in the Emopean Economic Community. However, it is already understood within the region that simplicity must chatacterize the CFI'. At the May 1991 meeting of mhisten and presidents of central banks, which convened to discuss the Common Foreign Tariff, recommendations were made to seek a CFI' that will consider 'the criteria d to seek a low tariff level compatible of uniformity or of the least dispersed stnrcture possible' a with the necessary insertion of the Andean economy into the world economy. 'he common tariff and the consolidation of a free trade zone One of the most dramatic decisions of the meeting of the Presidents of the Andean countries (La Paz, 1990), was to implement the Andean Fret Trade Zone by December 31, 1991. Tbat decision will add a relatively small number of tariff items to the categories that are already traded without tariff charges. However, the newly included items comprise mostly products traditionally excluded fmm integration prognms because they are considered 'sensitive.' Accordhg to the praidents' for only a small agreement, countries will be able to protect themselves from Andean com~dtion number of products: 50 items in the new NANDINA nomenclature for Colombia, Peru and Venezuela, and 100 for Ecuador and Bolivia. The decision to implement a free trade zone has highlighted the urgency of also reaching agreement on the Common Foreign Tariff, the harmonization of national export incentive regimens, and a higher level of coherence in exchange rate, monetary, and fiscal matters. The implementation of a free trade zone by the end of 1991 created high expectations in the region. However, in spite of convergence among national policies, it is precisely the current heterogeneity in tariff and export incentive policy that seems to jeopardize development in the region. Currently, disparities in tariff rates for the importation of inputs from third countries are such that problems are likely to arise in the free flow of merchandise. Regional trade will include a large number of components imported from countries where tariffs on inputs are low (from 1 percent to 5 percent); but in cases where trade is with countries with more homogeneous tariff structutes (such as Bolivia, Colombia, and Peru), inputs for the local manufacture of the product would be subject to higher duty payments, putting them at a disadvantage. For that reason, the current tariff instrument, that is, the common minimum foreign tariff, ought to be replaced by the common foreign tariff very soon. Under prevailing circumstances in the Andean region, optimum implementation of a free trade zone requires the prompt creation of a tariff union. Another difficulty in establishing free trade in the Andean region arises fiwm the diverse export promotion schemes and the heterogeneity in existing systems of indirect taxes. It will be necessary to harmonize these systems to avoid such spurious types of trade as moving the same product in both directions to take advantage of export subsidies. Likewise, it is clear that neither curreat drawback systems nor those being implemented in the five Andean Group countries will be applicable for intraregionai trade, because they would cause distortions to the extent that the destination of products would be determined by where they can benefit fram reimbursement of those taxes. It must also be noted that indirect taxes normally reimbursed for exports to t h i i countries will impose a further restriction on the level that the tariff union common barriers will create, given that a high l e d would create a bias against trade within the Andean zone.  Relations with other economic blocks The rise of other economic blocks in Latin America suggests that the Andean Group should define its relationship both with those groups and with the General Secretariat of the Latin American Integration Association (ALADI). In fact, both MERCOSUR and the Andean Group are attempting to overcome the difficulties inherent in the integration process through the creation of tariff unions responsible for negotiations within each free trade zone. Since they exclude a common tariff for third countries, these negotiations tend to exclude large segments of the commerce of member countries, as is the case of ALADI. There is no formal hindrance to the survival and development of the Andean Group, MERCOSUR, and the possible free trade treaty with Canada, United States, and Mexico. Those relationships will develop through the granting of preferential tariffs, for example, between MERCOSUR and an Andean country or the entire Andean Group. In the case of the Andean Group, such tariff agreements require the approval of the Cartagem Agreement Commission. With an eye to the overriding goal of broadening Andean integration, such agreements must take into account the basic principle of increasing the volume of overall commerce with a tendency toward trade liberalization. Conclusions of the me eting of the sememies of the economy, planniag and finance and presidents of central banks, Caracas, May 2 and 3,1991 The participmts of the m e w of the secremies of the economy, planning and finance and the presidents of central banks, (Caracas, May 2 and 3, 1991), arrived at the conclusions presented below and agreed on a work program for 1991 to harmonize economic instruments and policies. Government experts will participate in a p-ent regional information system of macroeconomic policies. The information system will provide the data necessary to sustain the discussion and exchange of experiences that will allow progress in the harmonization of macroeconomic policy. Common methodologies will be identilied and g e n d guidelines will be established to assist Andean Group members in adopting policies that further the goal of harmonization. Given the importance of macroeconomic policy in the integration process, the ministers of economy, central bank presidents and planning authorities agreed to active, direct and joint participa- tion in the process of regional harmonization.  To consolidate the Andean integration process more efficiently and promote growth of intrwional trade, the need to deepen economic policy for stabilization of their economies, and especially, to achieve growth with price stability, was emphasized. Meeting participants highlighted the need for an integrated study of diverse macro-economic policies, especially of the tariff, exchange rate, fiscal, monetary, and payments policies. They stressed that the harmonization process must be flexible, with progressive intensity. The design of the harmonization instruments is to be completed by December 31, 1992. The common foreign tariff will define the preferential margin for regional production that will allow for competitive insertion into the world market. Participant. highlighted the importance of maki ng progress in forming an Andean tariff union, noting that common trade policy, together with the harmonization of macroeconomic policy, will avoid generalized distortions in competition, which would seriously affect regional free trade. For the definition of the common foreign tariff, participants recommended uniformity and the least possible dispersal, to allow for commercial openness to the rest of the world. They agreed to participate in activities to harmonize the economic instruments and policies as set forth in the May-December 1991 work program, presented below. Among these, the definition of the common foreign tariff and the harmonization program of export incentives are notable, together with activities to establish the bases for gradual macroeconomic policy harmonization. All Andean Gtoup secretaries of economy, planning, and h a and central bank presidents were exhorted to attend the meetings that will be convoked so they may personally comiute to the rapid o advance Andean integration. decisionmaking required t The Andean Group sege€aries of the economy and central bank presidents agreed to meet as the Cartagena Agreement Commission in order to take decisions within their competence.  Appendix 5. Protection in the Motor Vehicle Sector This appendix explains the impact of various aspects of the protection regimen for automotive sector and motor vehicle assembly operations, including the import prohibition on assembled vehicles, tariffs on finished products and import packages, domestic content requirements (minimum domestic contribution), and obligatory exportation requirements. Background The Peruvian motor vehicle industry has evolved behind high protective trade barriers. In 1984 as the trade liberalization of 1979-81 began to be reversed, imports of assembled vehicles were virtually prohibited, and domestic vehicle assembly and parts industries were highly regulated, mainly t h u g h domestic content requirements and official registers of assembly firms and parts producers. Eight companies assemble vehicles in Peru, importing kits of completely knocked down (CKD) components and combining the components with locally produced parts (table A5.1). Tabk M industry fa Pap 1 Idtutionrl shcture of the rutomoii~e . Mobor Pau Pawirrr,l i d by V o l k m q p NiMororofPoru PutRrwi,purJlp.w# Ck Inv. Corn. cum^ P0nrvi.n. l i i b y M i i Toyou of Pea Jlp.wa Sania Vabii of Pau SIuadirh volvo of pcru Swadi Motor lndureir Rmrvi. licsarsd by Cbevmlet kdurrrkAu&moeizBoa Huverta - - - - - - - - Source: Paw'i M i of Motor Vcbicle F-ria. 'Ihere is also an assembly operation for diesel motors from imported kits, Motors Diesel A n d m which is a joint venture of the Peruvian government (50 pscent),: Volw, (25 percent), and Perkins (25 percent). 'Ihat company sells diesel motors to Volvo and as replacement motors. The largest companies are Toyota-with about 48 p e r m of locally assembled vehicle sales in 1990- and Nissan-with about 20 percent. Trucks (iiuding pick-ups) and buses dominate the market (over 80 percent of domestic assembly in 1990) (table A5.2).  Table M2 Produdon structure of the automotive industry in Peru, 1990 . S h a r e of Motor Peru Nirran Motor .Bet. CunaKi Toyou Scaoi Vabi Volvo Motorindurtri. Source: P a w i k r o c i of Motm Vobiclo Frtoricr. 'Ihe industry is charackrized by the mudl scale of its opemtions. Only around 3,600 vehicles were assembled by the Peruvian motor vehicle idustry during 1990. Production during the first two months of 1991 w as at an even lower pace, at 68 vehicles. Peak production was reached in 1975, with approximately 36,000 vehicles. 'Ihe annual production record belongs to Volbwagen, with around 10,000 units. During the last decade the major fhctors influencing changes in mator vehicle production were foreign trade policy and fluctuations in domestic economic activity. Production was relatively high during the 1970s, as the industry expanded behirur high trade barriers. Protection was reduced in 1979-80, but protectionist measures were reintroduced in 1984. Production was low between 1983 and 1985because of depressed economic conditions, but it expanded in 1986 and 1987 as a result of increased protection and economic recovery. 'Ihe decline in economic activity in 1988 again decreased production. Trade libdization, togezher with the depressed economic d i n s in mid-1990, exerted more pressure on assembly operations. Rate of effective protection The rate of effective protection of motor vehicle assembly in Peru is determined by the net hpapaet of all policies that raise vehicle sales prices or affect imported or domestic p m costs. Prior to the reforms of August 1990, high sale prices prevailed because of the prohibition on vehicle imports except by diplomats. Ignoring impom by diplomats, the prohibition implied that motor vehicle prices 'would be determined by the interaction of the demand for vehicles by domestic consumers and the supply of vehicles by assembly operations. Between August 1990 and March 1991, assembly operations were protected by a tariff of 50 percent, which dropped to 15 percent at the end of March. After the switch to a tariff regime, the domestic price was determined by the price of identical or similar vehicles on the world market and by the tariff. The cost of domestic assembly would, in turn, depend on the tariff on imported parts (CKD packages), the domestic content requirements (htegracion nacional minima), and the obligatory export requirements for imported kits. Requirements of 39 percent domestic content for automobiles were in effect until March, 1991, when they were abolished. The tariff on imported kits was 1 percent until August 1990, and 15 percent thereafter. The obligatory export requirement of 20 percent of import packages was also in effect, but it is ignored in this analysis because it had been temporarily abandoned and because data are not available for calculating the cost of those restrictions for assemblers. The rates of effective protection of the entire regulatory and protective regime, if it were complddy enforced, would be lower than those calculated here because the obligatory exports requirements will also raise the cost of parts for assembly operations. A model for assembly operations The impact of the different aspects of the protection regime on the motor vehicle assembly industry can be illustrated by a simple model of assembly operations, the market for assembled vehicles, and the domestic components market. The analysis supposes a competitive market structure, which probably does not reflect the real structure of the Pexuvian market,given the protection regime and the small number of firms, although it serves to itlustrate how the different r&ctions affect the profitability of domestic production. Suppose that domestic assembly combines a quantity of components to create a lkished vehicle, that all those components are identical, and that a components are required to produce a vehicle. Domestic content restrictions require the local purchase of a proportion 6 of the components, 80 that arb parts are acquired from the domestic parts industry and a kit contains a ( l 4 ) parts. The domestic  market has an implicit value added supply curve that shows the number of vehicles that would be assembled for different amounts of value added per vehicle. The cost of components per vehicle would be added to the value added in order to determine the price at which the different quantities of vehicles would be supplied. Using this model for the assembly industry, figure A5.1 shows the impact on the motor vehicle 'industry of the prohibition of assembled vehicle imports, of the domestic content requirements and of the tariffs on imported kits (the regime in effect prior to the August reforms). The upper half of figure A5.1 corresponds to the assembled-vehicle market. DA is the demand curve for assembled vehicles, SVAis the value added supply curve, showing the value added per unit required by the industry to assemble that quantity of vehicles. The assembled vehicles supply curve (S,J is obtained by vertically adding the cost of domestic and imported components per vehicle to The cost of domestic components will be a6PcD, which, when added to the value-added per SVA. unit, yields the supply curve S'. The cost of the imported kit will be a(lb)Pcw(l +ti), which, when added to S', yields the vehicle supply curve Sk (If the obligatory export requirements are also in effect, the additional cost imposed on firms will also raise costs, and another tam should be added.) Any other restrictive policy which raises component costs for firms will displace the supply curve upwards (or to the left) and will discourage domestic production. The lower diagram in figure AS.1 corresponds to the components market. $, is the domestic components industry's supply curve. Given the domestic contea requirements, the demand for domestic components from assemblers will be -a. (Tf the obligatory export requirements are also in effect, the demand for components for exportation will have to be added to demand for assembly purpose.) Suppose that components are mailable at a given price in the world market Pew. The cost of components for domestic assembly will depend on the tariff on those components (t)(and any additional cost involved in package importation, such as obligatory export requirements). Equiiibrium in both markets is determined simultaneously. The position of the supply avve in the assembled vehicle market depends on the equilibrium prices in the component mar& but, at the same time, the equilibrium quantity in the assembled vehicle market determines the demand for domestic and imported components and, therefon, equilibrium in the components market. Given the prohibition to import assembled vehicles and ignoring imports by diplomats, equilibrium in the finished vehicle market will be at price PADand q u d t y &. The price of domestic components will be determined by the demand and supply of the d o d c components at Pa and , .Q Equilibrium in the imported components market (kits) will be at P, .,Q (1+Q and Each of those equilibrium prices determines the position of the curve in the vehicle market. When the prohibition on the importation of assembled vehicles was replaced with a 50 percent tariff in August, 1990, the tariff became the binding restriction. At the same time, the tariff on imported kits was raised from 1 percent to 15 percent. The domestic content requirement remained in effect. In 'terms of Figure As. 1, those changes would lower the price of assembled vehicles to PAW(1+t J and , shift the domestic industry supply curve upwards to S,*, given the higher cost of the imported kits. Both changes would lower the rate of effective protection of assembly operations and decrease production to QAT, although it would raise sales to DAT. The reforms of March, 1991, lowered the tariff on assembled vehicles to 15 percent, and eliminated the domestic content requirements. After that reform, tariffs on vehicles and components would determine the rate of effective protection and, since import packages and assembled vehicles enter at a nominal tariff of 15 percent, the rate of effective protection would be 15 percent. Some conclusions derived from this model should be noted: Under the protection regime prior to August, 1990, the equilibrium price of vehicles (determined by domestic supply and demand), and not the tariff on assembled vehicles, would be the important variable for determining the rate of effective protection. A binding domestic content requirement lowers the rate of effective protection of assembly operations by raising component costs, and may lead to negative protection rates if they are sufficiently restrictive and domestic p m are s i g n i f i d y mo& expensive than imported parts. That possibiiity is illustrated in Figure A5.2, in which the dotted line SAP (which lies above SVAby an amount equal t o the component costs under free trade, a P& is the industry supply curve under free trade. If the domestic content requirements are biding, increases in the tariff on imported parts would raise the costs of the assembly industry, cause assembly operations to contract, and reduce demand for domestic parts.  figure A5.1 Economics of the Motor Vehicle Industry  Egure AS3 Negative Effective Proteetion to Motor Vehicle Assembly  Calculation of the rates ol effective protection of assembly operations The rate of effective protection on assembly operations is the percentage by which the protective structure increases the value added in those activities (the difference between the final price of assembled vehicles and the cost of all the parts incorporated into it). Data on the cost of Peruvian parts, the cost of those same parts elsewhere, the cost of the imported kits, the price of assembled vehicles on the world market, their sales price in Peru, and the transportation costs for import kits and finished vehicles (all in June, 1990) were obtained from a domestic assembler, Nissan, for three models of vehicles assembled in Peru. Those data were used to estimate the rates of effective protection for automotive assembly under various hypothetical combinations of policy alternatives, including the pre-refonn regime, the regime which prevailed after the reforms of August, 1990, and after the policy changes in March. The rates calculated here must be considered illustrative rather than definitive, given that they deal with the experience of only one firm and are based on data provided by only one instead of on information gathered by independent sources. The first step in quantifying parts costs was to compare the cost of each Peruvian part with the FAS cost of that part overseas. Betwexm 8 percent and 13 percent (by value, depending on the model) of parts were cheaper in Peru and presumably would have bexm used in the absence of any restrictions. The cost of parts in a free trade regime was calculated by adding (a) the cost of parts which were less expensive in, and thus, would be acquired in, Peru, (b) the FAS costs of parts which would be more expensive in Peru; and therefore would be sourced abroad, and (c) the cost of the CKD package, under the content requirement. 'Ihis m d o d supposes that transportation costs for a larger number of imported parts would no higher than those for the kit actually imported under the domestic content regime. This method may therefore bias downward the estbam of rates of effective protection by underestimating the cost of parts in a situation of free trade and over-g, therefore, the value added. Then, the same procedure was followed, assuming a 15 percent tariff on parts. In this second exercise, it was supposed that assemblers would acquire in Peru any part with a value less than 1.15 times the equivalent FAS part. The basic data and the costs calculated for domestic and imported parts (CKDkits) in a situation of h e trade and under a tariff of 15 percent on parts are presented in table AS.3. In June, 1990, the time period to which the data apply, importing assembled vehicles w a s prohibited, domestic content restrictions of 3 8 4 percent were in effect, and imported kits were  subject to 1 percent tariff. The rate of effective protection under that regime (Ernj) was calculated using equation (1) below. The calculation of what the rate of effective protection would have been with the import prohibition but without the domestic content requirements illustrates the degree to which the latter factor discourages domestic assembly. Equation (1) was used, the index k in the first summation in the mumerator was changed to 1 ..kF to reflect the (lower) number of parts that would have been acquired domestically. The index of the second sum in the numerator was Qanged to kf+ l ..n to reflect the (higher) number of parts that would have been imported (included in kits). The rate of effective protection for assembly operations after the August reforms (ERPJ when imports were permitted with a 50 percent tariff, packages paid a 15 percent tariff, and the domestic content requirements were still in effect, was calculated using equation 2, with tA=0.5 and ti=O. 1 5. Equation 2 was also used to calculate the effective rate under alternative tariff regimes, with the content restriction. Table A-53 Midata for the dcdati01~ of effective rrta (USC) Cortofdomsaicputr (39%d o & controt) ~0fdomaricp.rtr (no dodconteottoquired) coaofdomarticputr (no domcuic marrat roquirod. 15%w o n puts) FAS cort of foreign pam (no d o d conkat required) FAS cost of foreign pmrm (no dOmcItic c o n k required, 15%w o n puts) CortofCIEDpdiage (39% domatic controt) Rcbyo--- CortofcKDp.dpge (no domsrtic contmt required) CortofcKDpdmgo (no domQltic mntalt rsquircd, 15%M o n puts) Vohidc nler pries Pnda import p r o h i b i i FAS cou of qpivdcot vchic)e T ~ c o u f o r ~lod~errpertsdmhem  The rate of effective protection, after the abolition of the domestic content requirements and the reduction of tariffs on vehicles to 15 percent in March 1991 (ERPM) was calculated using equation 3. Equation 3 was also used to calculate effective rates under different tariff systems, without the content restriction. The results of the calculations of the effective rate of protection under the actual regimes and hypothetical regimes (with and without domestic content requirements) are shown in Table 5.4. The range presented is the range across the different models. A5.4 bhateof &&e rrta of protcdion for vehicle mernbly h 1990, 1990, rrd Much 1991 (pacrbtr) -R=q"i--f Equations for calculating effective rates of protection to assembly where: P is the d o d c pria of the msemblod vehicle with the content restrictions md prohibition on imports P, is the FAS price of the died vehicle, plus Mght chges from qmthg wuntry ~isthedomssticpriceofcompmcati istheFASpri~ofcampanmtiind~w~ofwrpply n isthermmberofcompma~tsussdtoproducethev~cle kP is the number of parts ~#)urced domstically under the domstic conkat rsquinment kP is the number of puts ddommtically in the alweace of my trade or damestic anntrnt mtrictitm kTisthenumbgofpartseourcsd~dy&a15%~ffon~ is the 25% tariff on assembled vehicle8 rfta August r e h (-0.25) is the 15%tatiff on assembled vebiclc8 after the hduch reforms (-0.15) isthe 15%tarifftmkitsrftaAugustrefonns(-0.15) # is the 1%tariff on kits before the August reforms ( ~ 0 . 0 1 )  Appendix 6. The Restructurhg of the Customs System At the request of the Peruvian economic authorities, the issue of the reform of customs administration was addressed, on the basis of a working paper prepared by the World Bank Mission of November 1990. Meetings were held with different officials of the ministry of economy and finance, the national customs superintendency, and other ministries related to foreign trade. At the same time, meetings were held with the private sector to gather impressions about the reform. To complement those sources, it was decided to include the following materials: System of customs price controls. Administrative outline of the reform plan. System of custom price controls The norms for customs merchandii valuation are contained in article 4 of D.L.22619, which includes 45 rules based on the customs merchandise valuation system of Brussels. lhe current system The structure of the customs on - includes valuation units responsible for reviewing the prices declared in import and export invoices. Customs value i s determined on the basis of invoice prices, plus transportation costs and insurance, and adjustments when appropriate. Whe n invoice prices are reviewed, customs uses reference prices for identical or similar merchandise found in catalogues, booklets, or brochures and data ftom previous importations. There are no orderly and systematic files of prices with all the necessary information, nor are stadstics available ftom foreign trade, nor is there a microfilm system of invoice prices. The national customs superintendency has not established procedural nonns to be applied in the valuation process, nor has it prepared operating manuals.  Critical analysis of the system The lack of instructions codified in manuals gives great autonomy to customs agents, which, together with the lack of orderly price files, renders control ineffective; and there have been complaints that the system delays and halts the processing of paper work. 'Recommend~~ons During the period of reform of the customs system in Peru, estimated to last one year, a special unit of qualified technicians could be created to review the prices declared in import and export invoices. The customs administrations of Callao and Jorge Chavez airport would send copies of the invoices processed every day, with their respective receipts, to the national customs superintendency. Such a mechanism would create a control over the day-to-day business of the customs agencies. The new system would have magnetic files of prices, organized according to tariff position, which would serve as a support device for price control. Moreover, a special chapter in the compendium of customs norms is recommended which would provide for establishing valuation and procedural norms in operating manuals, including the duties, obligations and responsibilities of customs agents. For certain cases, customs should require price certification from foreign consulting firms. But it is not advisable to base price controls on reports from foreign firms as that function should not be totally delegated to outside concerns. Administrative outline of the d o r m plan General w~siderations Customs reform is part of the foreign trade liberalization policy. Curredy, Peruvian foreign trade is overregulated, hampered by the intervention of numerous departmenss from diierent ministries, which are legally obliged to exert control over imports and exports often causing delays and paralyzing import and export processes. A commission of three persons should be created to investigate the elimination of unnecewry controls and to propose ways to implement controls that should be maimbed so as to avoid creating obstacles to customs procedures. The commission should produce a report 45 days after it is convened. Infrastructure The following steps should be taken to ensure adequate infrasuuctural support for the customs reform: A well-equipped office should be made available to a reform committee of six to eight persons. This will allow the committee to operate in isolation from the sector it has been commissioned to reform. One person could be designated to perform the necessary investigations, issuing a report within 30 days. A building should be bought or rented in which to install the National Customs , Superintendency and the customs administrations of Callao and Jorge Chavez airport. The building should be located near the port and airport. For this purpose, a commission of three persons could be created, issuing a report within 30 days. Customs legislation and operating norms should be analyzed and evaluated in order to propose reform of the following laws and regulations: General Customs Law Regulations of the General Customs Law Organic Law of the National Customs Superintendency Contraband Law General Tariff Law Compendium of Customs norms Operating manuals Six qualified technicians should prepare proposals to modifythe laws to be reformed, preparing the technical reports necessary t o justify their recommendations, within six months. At the same time, the compendium and operating manuals should be prepared. IaE,maatbn on processing Onaite evaluation should be undertaken of the equipment alternatives that could fulfill the specifications. For this purpose, it is recommended that an outside expert in customs data processing be hied. This technician should evaluate the IBM equipment in the national customs school in the light of the proposed specifications and estimate the cost of its implemeambn, should it be approved. Operations Operational units for the reformed customs administration should be structured according to the procedures stipulated in the new regulations. This will require a working group of six persons for three months. of the new system 'Implententananon Implementation will be in successive stages, beginning with the national customs superintendency and the customs administration of Callao, in its different offices, continuing with the customs administration of Jorge Chavez airport, and concluding with other customs services. To that end, it will be necessary to hire professionals and qualified technical personnel who, together with selected current personnel, will structure the new operational units. This stage should last six months. Hu au m tesourccs Personne l from the customs services and the Central Reserve Bank of Peru, qualified Pemvh technicians, and foreign experts could be h i to constitute the corhmittee responsible for the reform of the customs system.  Appendix 7. The Political Economy of Liberalization At the end of the last century, Peru's industrial activity was especially linked to the exploitation, transformation and exportation of a wide range of natural resources. From the beginning of this century, Peru has developed an efficient industry for the production of capita1 goods. Peruvian foundries and machine shops competed with foreign concern to produce machinery for mining, sugar and cotton processing, and railroads. By mid-century, the food processing industry accounted for nearly half of the industrial sector, and by the 1950s, the industrial sector diversified mainly into chemical products, basic metals and paper products, was achieving dynamic growth, witb growth rates higher than those of the overall GDP (table A7.1). By 1959, a definite industrial policy was being followed, influenced by the import-substitution industrialization strategy implemented in other countries of the region. Tbat policy is embodied in the Law of Industrial Promotion, No. 13270, promulgated in 1959, whichcreated cariff and tax inceatives that lowered the cost of machinery and raw material imports for substitution i n d d e s and established o establish new industries. New tariff structures created in 1964 (Law 14816) generous tax incentives t and in 1967, increased-the eff&tive proteetion for imjmt substitution industries by raising tariffs on consumer goods and lowering those on capital goods. Tabk A 7.1 A m d rrte d grow& of orarll GDP and of the minuf&turiq GDP fn rdstcd periods Sheltered by Law 13270, industry grew an average of 7 percent anrmally between 1960 and 1%8, while overall GDP grew at a rate of 6.2 percent (table 36). By the eed of the 1960s, industry produced 25 percent of GDP, although it was h d a m W y dependent on ovesseas supplias for  capital goods, raw materials and semi-finished products and investment, technology and managerial expertise. The industrial process was not integrated into the rest of the economy but, rather, was highly dependent on the availability of funds produced by the primary sectors. Moreover, it introduced a significant degree of vulnerability into the economy because any balance of payments adjustment had an amplified effect on industry, since it was the country's principal importer. High levels of protection and over-valued exchange rates inflated industry's apparent share of the economy-protectionist policies made industry's value added appear greater than it really was, since it was valued in domestic market prices (rather than border prices) in national accounts. The military government that took power in 1968 renewed the thrust to promote import substitution at a time of general disenchantment with that industrialization strategy in the more industrialized Latin American countries. However, that renewed push occurred in the midst of profound changes in property relations within Peru's productive enterprises, as part of a plan to change the economic and power structures of the country, so that the State would assume an ever greater role, displacing national and foreign private sectors (Reform Law of July 1970). Ever more extensive controls were placed on access to foreign currency, inputs and credit channels, eventually covering state participation in investmeat programming, control of employment--~~en in private companies-and creation of t a r E and other types of import restrictions (quotas, licmses and prohibitions). Substantial resources were channeled to the industrial sector. Between 1970 and 1975, the real installed capacity of that sector grew 25 percent. While private sector dissatisfaciton grew, there was also real and sub- growth of earnings as a remlt of genemu government incaives. The balance of payments crisis of 1976 halted that growth abruptly. By 1977 and 1978, the growth was negative, while the fiscal deficit, inflation and the current account deficit account were rising. Finally, in 1979, the high costs of protection were recognized, and economic policy emphasized stabilization and trade libdization. A new tariff was approved and most non-tariff restrictions were replaced by ad valorem duties (between 0 and 155 percent) and by a l ist of temporarily prohi'bited products. Responsibility for trade policy was transferred from the M if or Industry to the M i of Economy, while the new Coxwitmion of 1979 gave control of tariff policy to the executive. That transfer made it possible for the new government to continue the tariff reform begun by the military govesament, reducing the maximum tariff from 155 percent to 60 pacent (table A7.2). Although the reduction affected only about 14 percent of tariff items, it provoked a violent reaction that was intensified by the government's amrollnad plan to continue to reduce tariff rates until they reached the goal of a flat 25 percent rate.  Table A 7 3 Evolution of the nominal taritf structure (Ad vabrem and surcharge) (Arithmetic average by tariff ssctioas) 1. Livestock .. II. Vcgerable productr m. FUJ & oils IV. Indukial productr Foods lad bevaaga V. Mineral productr VI. Chuniul producta W. Plluticr wr.Hi,lkinr LX woodladchucorl X. Muatlsvrodin the produdion pq# XI. Tdermterirlr M d mMufuurod pmd- xu. Footwou, ham dtbelika m.Stooc. p k n a aad m e a t prod- XTV. Jewdry Xv. command .naprod- - l E XVI. - .Idr p p l W m. Tvltarion mrtairl m. Opticrlpbatognphic CiImlaOgmphic W L Armr.Idmudbm XX.Othen ma. Altobjem Notes: T h c d u a u c f r o m D c c 4 a r b c r o f ~ y s u . I d r c i ~ 1 m ~ ~ . (1)Includa15~~mtbeUfiff. (2) Includol l o ~ s u r c l u r p a n cmvrluo. (3) locludor 15 pcrccat wchuge an QF vrluo, vrlid u of Aum (4) Includor 17 pawot auchrgo on QF vrluo, vrlid u of M.y.  Reaction to the new tariff measures was exacerbated by the gradual but steady appreciation of the exchange rate which the military government had begun six months before leaving ofice in an attempt to slow the monthly rate of inflation. The inflation had been stimulated by the surplus in the current account, resulting from the stabilization program begun in 1978 and the sharp increase in the terms of trade during 1979-1980. The new Government continued to use the exchange rate as a tool 'to combat inflation until the end of 1982, when the effects of expansionist fiscal policies had made the international reserve situation critical. International financing of Latin American countries ended in 1983, as commercial creditors reacted to Mexico's inability to make debt service payments. In that year Peru also suffered two severe natural disasters-extensive flooding in the north and severe drougnt in the Southern Andes. Peru's expansionist fiscal policy continued as tax income fell, obliging the Central Bank to apply contractive monetary policies and to devalue continually in order to defend its reserve positions. It was in that environment that reversal of the trade reforms began to gain momentum: first, in response to the need for fiscal revenue and then in reaction to pressure from the industrial sector. The crisis, although the result of the international situation and the n d disasters and the lack of a coherent fiscal response to those shocks, was blamed on trade liberalization. Demand rose for a return to the system of prohibitions, licenses and tariff increases in order to "promote savings'. New tariff exemptions were granted to agriculture and industry, and coverage of the positive list of i imports, and the Lst of imports subject to prior importation licensing was expanded (table A7.3). The net result was a level of effective protection high- than at the end of the military government. That protectionist momentum gathered speed during the Garcia Government, which emred office in July, 1985. In August the list of products subject to prior licensing was increased to cover 60 percent of total imports. A new Institute for Foreign Trade 0 was created to 'fomndate, execute and supervise foreign trade policy, and direct and coordinate all activities that promote development'. .. The Institute's principal mission during the Garcia government was to admmsm and control nontariff barriers to imports. In September, 1987, all items were subject to the import pmhitbiion lists. The program administered by the IFT and subject to Central Reserve Bank approval, also came to incorporate foreign exchange restrictions. That paraexchange rate system was complemented in 1988 by a payment authorization requirement (kr imports), within the framework of daily currency proP-@* The c o m b i o n of a myriad of non-tariff barriers and exchange rate controls created a chaotic system of protection in which political power, sometimes exercised by a minister, some$hes by an official of lower rank, granted ad hoc protection and allocated foreign currency for impom acwrdinp to seventeen different exchange rates. The system was totally fluid, with wnstant transfers of items from one category to another for exchange rate, tariff and nontariff effects. In its last months the Garcia government returned to a policy of lower nontariff protection and reductions in the range of tariff variations. Table A73 Nontariff nstrictions on imports, 1979-86 1979 1979 1ABO 1P81 198.3 1P84 1986 Mturh Dec. July Dec. Dec. Dec. Dec. ltemr subject to prior licauing 1038 1258 343 111 118 126 U75 hn. - d Y prohibited .ad #hen Source: N d o d Inrtituto of St.tirticr of Pau. The new process of openness 'Ibe new government that entered office in July 1990 viewed the process of openness withim the context of a stabilization effort designed to end the hyperinflation that had raged fir more than 24 months. Trade reform and stabilization measures were to be implemented together, accompanied by a drastic correction of relative prices and a floating exchange rate. Trade liberalization was to reorient the productive incentive structure toward exports, assist in the correction of strongly distorted relative prices and help stop the inflation process. The principal m easures approved in August 1990 included the fixing of a maximum tariff rate of 50 percent and a minimum of 15 percent. Moreover, 39 exemption program wese elimiad, the import prohibitions S i were cut back, and the prior licensing requirement was suspended. Toward the end of 1990, the tariff structure was reduced to only three rates: 15 percent, 25 percent, and 50 percent. However, in the last quartex of 1990, the government began to show signs of  indecision regarding trade policy. In September, a temporary surcharge of 10 percent was added to products subject to the 25 percent and 50 percent rates, but was eliminated after one month. The convergence of the tariff structure toward a single rate of 25 percent by the end of 1994 was also announced. Erratic signs continued to be given in January, when the planned reduction of tariffs on 141 industrial inputs was not put into effect. A new 5 percent tariff was applied to steel processing 'inputs, to provide additional protection to the state steel company. The lack of coherence in ttrade policy undermined its credibility among producers and gave rise to profiteering maneuvers as producers sought exceptions one sort or another. Pressure increased for changes in the lists of items subject to the highest tariff rate, and demands for exemptions in the tariffs on inputs and other products continued. The published calendar of gradual miff reductions lost all relevance in the face of delays that drastically undermined the confidence of investors and industrialists. The time preceeding the target date of 1994, for unification of the tariff structure at 15 percent, came to be viewed as a lengthy opportunity for undermining the entire reform program. In that atmosphere of uncertainty, important business sectors began to pressure the government to define its policy. Those sectors consisted of broad groups of businesses, including not only traditional exporters, but also a large group of industrialists who had decided, during the last three years of recession, to di their productive strategy toward exports. The catalysts for that business support were the appreciation of the real exchange rate, which had sharpened since the beginning of the stabilization plan, and the Camd Bank's decision to c o n t h e with d c t i v e monetary policies. The effects of a floating exchange rate and r d c t i v e monetary policy (which generated high interest rates) combined with the business sector's perception that an increase in imports was vital, to generate active private sector support for increasing openness. With a floating exchange rate, business Wests d d not presnue the goverment, as in the past, to devalue the exchange rate. At the same time, ~11- about the success of the stabilization process undermined credibility in the capacity of the Central Reserve Bank to raise the real exchange rate through purchases in the free market. In that context, the National Exporters Society commissioned a study in support of immediate unification of hrEs at a rate of 15 percent (Proposal for Tariff Unification, NES, February 12, 1991). The main arguments in the study concerned the need to generate aedibility for tariff policy, to increase imports and to raise the real exchange rate. The report argued that the three rate structure maintained levels of effective protection which were too high for certain sectors, that tariff unification would not affect the bulk of the industrial sector, and that such a tariff reduction would be neutral in fiscal terms. That business proposal coincided with the changes in the government's economic team of February, 1991. The new minister received the proposal favorably and used it to argue for tariff unification. In the end, however, it was finally decided to retain the 15 percent and 25 percent rates, in the 'understanding that the 5 percent rate, which still applied to a few items (steel processing inputs), would be e l h i w e d . Moreover, the great majority of nontariff measures were eliminated. However, together with that new liberalization thrust, import prohibitions were established for used goods in the areas of clothing and footwear, along with specific per ton duties for 18 food items. That measure was promptly rescinded, in part when the fixed surcharge was replaced by a specific variable duty. The level of the variable duty depended on international prices of those goods. The deepening of the reform was well received by the business sector (table A7.4). Table A7.4 S u n t y ol buiDar pawns with reapst to government .dMtJ S C A L E SmbibtiOopnDgnm --my T u policy T d m h PhrocLlmmh of^ Lbar - o f miamodm -I- l t e u o l u w - On the other hand, the impact of more clearly defined trade policy on the level of imports and the real depreciation of the inti have been significant. Bemeen March and May, the Wdollar exchange rate rose from 560,000 t o 850,000, or by 51 percent, a real depreciation of 24 percent. Imports for April-May rose 30 percent with respect t o the January-February level (table A7.5). Current support for trade openness, in contrast to the opposition of 1979-81, is, however, not as strong as it seems. There is pressure, both within the g ov- and the business community, to reverse certain aspects of the trade refonn. Three main sources of pressure to reverse the reforms or at least diminish their intensity can be identified. First, there are the industrialists who produce importcompeting goods and have suffered substantial losses in their levels of effective protection. Their concerns take the form of pressure to move products from the 15 percent to the 25 percent rate, and to expand the range of products covered by the 5 percent surcharge. There is constant pressure to establish minimum evaluation prices for customs and to apply antidumping duties. Second, there are elements withii the executive that try to defend their respective sectors or who feel that their power has been diminished by the reforms. Those interests are diverse and, in some cases, at odds with one another. Those in charge of agrarian policy try to expand the list of agricultural produds subject to specific duties. The ministry for industry, in coordination with industrial producas or institutions linked to that sector, oies to raise the level of effective protection by increasing the number of tariff rates and reinstating nontarifF barriers in the fonn of technical norms with clearly protectionist aims. State companies and Parliament exert pressure for expemptions from tariff payments. State companies argue disingenuously that the state should not have to pay taxes to itself, and Parliament tries to achieve redistributive goals by keeping transportation @uses, trucks and parts) or food and medicine prices down. Finally, there are always political interest that oppose the current liberalization process. These range fromthe accusation that liberalization lewes domestic labor unproteed, to opposition to establishing trade policy out of step with the Andean Group members, which except for Bolivia, have moved much slower in implementing Iibesalization measures. This has created an opportunity to pressure for a reversal of the reforms under the pretext of concan for adherence to the tariff harmonization process of the Andean group.