12 - Permanent orrice Copy Report No. 1671 -ME , Mexico - anuacturing Sector: Situatioon, Prospects and Policies f1L upX (In Two Volumes) Volume 1: The Main RepOrt May 1, 1977 Latin America and Caribbean Region FOR OFFICIAL USE ONLY Document of the World Bank Thisdocuent as a restricted distribution and may be used by recipients This document han a teir off cidal duties Its contents may not otherwise be disclosed without World Bank authorization Currency Unit: Peso (Mex$) Prior to September 1, 1976: US$1.00 - Mex$ 12.50 Mex$ 1.00 = us$0.08 Mex$ 1 million = US$80,000 On September 1, 1976 the fixed parity of the peso was abandoned. During 1977 the value of the peso has fluctuated in the range of 20-23 pesos per US dollar. As of August 31, 1977 the values were: US$1.00 = Mex$ 22.87 Mex$ 1.00 us$o.o4 Mex$ 1 million = US$43,733 Fiscal Year: January 1 through December 31. FM OflCAL UK ONLY GLOSSARY OF ABRREVIATIONS AND ACRONYMS CANACINTRA Camara Nacional de la Industria de Transformaci6n National Chamber of Manufacturing Industries CEDI Certificado de Devoluci6n de Impuestos Indirectos Certificate of Reimbursement of Indirect Taxes CFE Comisi6n Federal de Electricidad Federal Electricity Commission CONASUPO Compafiia Nacional de Subsistencia Popular Public Corporation for distribution of low-priced basic food products FIDEIN Fideicomiso de Conjuntos, Parques, Ciudades Industriales y Centros Comerciales Trust Fund for Industrial Parks and Cities and Commercial Centers FOGAIN Fondo de Garantfa y Fomento a la Industria Mediana y Pequefia Fund for Promotion of Small and Medium-sized Industries FOMEX Fondo para el Fomento de la Exportaciones de Productos Manufacturados Fund for the Promotion of Manufactured Exports FOMIN Fondo Nacional de Fomento Industrial Fund for Industrial Promotion FONEP Fondo Nacionel de Estudios de Preinversi6n Fund for Pre-investment Studies FTE Free Trade Equilibrium (refers to exchange rate) TEPES Inatituto de Estudios Politicos, Econ6micos y Sociales Institute of Economic, Political, and Social Studies (analytical arm of the PRI, Mexico's principal political party) IMCE Institute Mexicano de Comercio Exterior Mexican Foreign Trade Institute INFONAVIT Instituto del Fondo Nacional de la Vivienda para los Trabajadores Workers' Housing Fund SIC Secretaria de Industria y Comercio Ministry of Industry and Commerce SICARTSA Siderurgia Lazaro Cardenas, Las Truchas, S.A. Lazaro Cardenas - Las Truchas Steel Company SMI Small and Medium Scale Industry All monetary values are expressed in US dollars, unless otherwise noted. iThis document hs a retricted ditribuion ain may be usd by rci*sfa only in t petformuace ft *hai nflkr;^l .41tims 1 nxt " eninns v nt dtherwi e be disel d WitbK>t W-Vi lnk authofintin. This Report is based on the findings of a mission which visited Mexico in October/November 1976. The mission was composed of: A. Nowicki Chief J. Bergsman Deputy Chief D. Keesing Export analysis T. Hutcheson Finance; Import controls D. Cook Small and medium-scale industry J. Levitsky Small and medium-scale industry F. de la Balze Small and medium-scale industry J. Kendall Public sector enterprises H. Choi Capital goods sector S. Swayambu Capital goods sector W. Oettinger Capital goods sector D. Weigel Private banking A. Tejano Statistical assistance TABLE OF CONTENTS Page No. VOLUME I: INTRODUCTION ........................................... i-ii SU1MAIRY .. ................................................ 1 ANALYSIS ............................................... 9 I. FOREIGN TRADE POLICIES ... ............... . .... . ..... . 9 Exports ......................................... 10 Imports and protection . ..............* ......... 15 Net protection and export incentives ............ 18 Choices for the Future ............................ 21 Conclusions ....................................... 28 II. PUBLIC AND FOREIGN-OWNED ENTERPRISES .... .......... 29 Public Sector Enterprises ......................... 29 Conclusions . .... ..................................*** 34 Foreign-Owned Enterprises ............... .. ........ 36 Conclusions ............ ........................... 39 III. INDUSTRIAL FINANCING .............................. 41 The Financial Situation ............ .. ............. 41 Conclusions ............................. 43 IV. TECHNOLOGICAL AND SECTORAL ISSUES ........... ...... 45 Employment and Wage Policy ..... ................... 45 Small and Medium-Scale Industry (Summary) .... ..... 48 Capital Goods (Summary) ........... .. .............. 54 Regional Development .............................. 58 'NNEX I: Tables and Charts .ANEX II: Technical Note: Free Trade Equilibrium Exchange Rates ANNEX III: Mexico's Industrial Structure References VOLUME II: ANNEX IV: Small and Medium-Scale Industry ANNEX V: Capital Goods TABLE OF CONTENTS (Continued) LIST OF TABLES AND CHARTS IN VOLUME I Text Tables 1. Overview of Policies Suggested for Consideration ............ 7 2. Manufactured Exports, 1970-1975 ..... ........................ 11 3. Exchange Rates, Protection and Export Incentives 1970, 1975, and December 1976 ...................................... .......... 20 4. Alternative Exchange Rates, Protection and Export Incentives 25 5. Prices and Profits in Public Sector Manufacturing Enterprises ................................................ 32 Annex I Tables 1.1 Imports as Ratios to Domestic Demand, 1967-1974 1.2 Manufactured Exports not including Assembly Plants: 1965-1976 1.3 Exports as Ratios to Gross Value of Domestic Production, 1967-74 1.4 Mexico's Manufactured Exports (Including Assembly Plants) by Destination, 1973 1.5 Public Sector Enterprises in Manufacturing - Investment and Financial Performance 1.6 Foreign Investment Enterprises in Manufacturing - Investment and Financial Performance 1.7 Indicators of Financial Development (1973) 1.8 Nominal and Real Interest Rates 1.9 Monetary System Balances, 1970-76 1.10 Private Banking System Income Statements, 1966-75 1.11 Labor Costs, 1976 1.12 Manufacturing Sector Wages 1.13 Average Wages in Manufacturing: United States and Mexico: 1960-1976 1.14 Capital Goods Imports in Relation to Domestic Consumption in Argentina, Brazil, and Mexico: 1973 TABLE OF CONTENTS (Continued) LIST OF TABLES AND CHARTS IN VOLUME I (Continued) Annex I Charts 1.1 Real Effective Exchange Rates for Imports and Exports of Manufactures 1.2 Sectoral Distribution of Credits from Banking System Annex II Table 2.1 Export Incentives and Disincentives, and Estimated Overvaluation, 1970 and 1975 Annex III Tables 3.1 Value Added in Manufacturing 3.2 Industrial Structure in Brazil, Turkey and Spain, 1970 3.3 Value Added per Capita in Mexico, Brazil, Turkey and Spain, 1970 3.4 Value Added per Worker 3.5 Value Added per Worker in Brazil, Turkey and Spain, 1970 3.6 Actual and Predicted Value Added in Manufacturing - 1970 and 1975 MEXICO: MANUFACTURING SECTOR: SITUATION, PROSPECTS AND POLICIES INTRODUCTION i. This report is written during a period of unusually rapid economic change in Mexico. One month before the mission visited Mexico, the exchange rate of 12.50 pesos per dollar which had been in force since 1954 was abandoned. The system of rebates of excise taxes to exporters of manufactures (CEDIs) was dropped and a new tax imposed on manufactured exports; after a few weeks the tax was withdrawn. On December 1, 1976, Mexico's new President took office, instituted an administrative reform which created new ministries and transferred to them responsibilities for many of the policies analysed in this report. The new government has launched changes of many policies and reviews of others. An "Alliance for Production" charter between government and the private sector has been signed. ii. Analyses of economic policies can be complete only when they include a review of the efficiency with which the policies are implemented and an assessment of their results. In the case of this report, many of the new industrial policies were introduced only very recently, after the mission's departure from Mexico. The authors of the report were therefore unable to review the new policies adequately. In a broad sense, however, it is apparent that the Government is moving to improve policies on many fronts: the bias against exports of all kinds, in favor of import substitution of manufactures, is being reduced; public sector manufacturing firms' behavior is being brought more into line with overall development goals, and steps are being taken to reduce the credit squeeze previously faced by private industry. It is to be hoped that further progress will be made in these directions and that yet additional steps leading to increased employment will also be taken. iii. Some of the specific measures discussed in this report, that have already been enacted or are under active study by the Mexican authorities, are the following: The exchange rate is floating at a realistic level, and CEDIs were restored. Administration of both import tariffs and import licenses has now been united in one place, the Tariff and Foreign Trade Commission; the Commission has established a timetable for review of all 6,000 product classi- fications requiring import licenses, and 880 have already been dropped from the list. However, the government has declared its intention to continue to rely on the licensing system, albeit less heavily. Public investment programs are being reviewed, in particular those in industries where private capital may be willing to invest. The Federal Electricity Commission, Pemex, and the National Railways have declared their intention to identify their equipment purchasing plans in more detail and more in advance, to give domestic suppliers a better opportunity to prepare to meet their needs; Nacional Financiera is considering a new program to finance capital goods production, and the Tariff and Foreign Trade Commission has declared its intent to end preferential access to imported equipment by public sector firms. The Bank of Mexico has reduced marginal reserve requirements which may increase credit available to the private sector; it is also allowing interest rates paid to savers to rise so as to reinvigorate the flow of savings to Mexico's financial intermediaries, and the highly liquid "bonos financieros" are to be gradually replaced by a series of certificates of deposit with different fixed terms. Programs provid- ing credit and technical assistance to smaller firms are being strengthened, and the Bank is considering financial assistance to Mexico for those programs. The annual increase of minimum wages in January 1977 was held to the reason- able level of 10 percent. The geographic division of the country into zones for the application of investment incentives to less-developed regions is being revised, although it is not clear that even the revised zones will concentrate incentives sufficiently, or whether the nature of the incentives will also be improved. Finally, the administrative reform of January 1977 (not discussed in this report) should permit increased effectiveness in policy design and implementation. Other positive steps, not related to the analysis in this report, have also been taken, and still others are under active consideration. iv. These and other important changes have not altered the mission's view--which, as it understands, is shared by the Mexican authorities--of the importance of three basic objectives for manufacturing sector development, and of the importance of designing policies to meet those objectives. The present report is conceived precisely with this purpose in mind. The objectives, while economic in their nature, have important social ramifications. They are: rapid and efficient growth in production, balance-of-payments management, and the creation of productive jobs for Mexico's rapidly growing labor force. It is recognized, of course, that Mexico has other objectives as well and that in some cases policies that would promote the goals treated here perhaps cannot be put into effect immediately and in their "pure" form because of the need to accommodate other objectives or constraints. It is hoped, however, that any such compromises can be held to a minimum, so that the three objectives treated here can be met as fully as possible. SUMMARY 1. Over the last several decades, manufacturing has grown rapidly in Mexico and today plays important roles in employment, import substitution and exports, and through its large impact on the economy as a whole. In 1970, the last year for which complete census estimates are available, manufacturing accounted for approximately 23 percent of GDP and employed 17 percent of the labor force. In 1975 the share of GDP was still 23 percent. 1/ Growth in the volume of manufactured output was 9 percent per year during the 1960s, con- siderably more rapid than the 7 percent per year growth in GDP. During 1970-1975 manufacturing growth slowed to 6.2 percent per year, only slightly over the annual GDP growth rate of 5.7 percent. Development of the sector has followed a pattern common to several other large LDCs, with import substitu- tion almost complete in consumer goods (both durable and nondurable), less so in intermediate goods and least well advanced in capital goods. (See Table 1.1 and Annex III. For more details on Mexican industrialization, see Aspra, 1977; Nacional Financiera, 1971; Solis, to appear; and Villarreal, 1976.) 2. Efficiency, measured by international price comparisons, presents a mixed picture. Average price levels for manufacturing as a whole were 17.9 percent above international levels in 1960, 15.6 percent above in 1970, and 19.0 percent above in 1975. 2/ (These figures are net of estimated peso overvaluation of 6.0, 2.8, and 18.4 percent in the three respective years. Realized nominal protection of manufactured goods, based on prices converted at the official exchange rate -- i.e, including peso overvaluation -- is estimated at 25.0, 18.8 and 40.9 percent. See Annex II.) Thus the average efficiency of the sector compares well with other large import-substituting LDCs, where protection has been considerably higher, but not so well with the more open, specialized economies. Moreover, the averages in Mexico hide a fairly wide variation; many processes are highly inefficient, as shown both by price comparisons and by plant visit analysis. In a comparison of prices of 394 products in 1975, 52 products (13 percent) had prices more than 50 percent above international levels and 26 of these (7 percent of the total) were priced at more than double the international levels. (These price differen- tials also are net of overvaluation, estimated at 18.4 percent in 1975.) This scattered inefficiency is protected by an import licensing system which virtually ignores price differentials in its usual refusal to permit the private sector to import goods that are produced domestically. 1/ Data on the different variables are not collected on a consistent basis and hence the estimates of shares may not be accurate. 2/ Throughout this report, measures of protection are based on estimates of how much Mexican prices exceed international prices. The quantity measured therefore reflects how much Mexican producers actually take advantage of various protective devices. Estimated changes over time in protection may be biased downwards because the 1960 study made more adju.3tmerts for quality than did the 1970 one, while the 1975 study t.-ade none. Thus protection may have increased somewhat more over time tlian the estimates indicate. -2- 3. Export performance of manufacturing has also been mixed. In 1975, the value of manufactured exports (including total value of export sales of assembly industries -- maquiladoras -- which accounted for half of the total) amounted to $2,090 million. At 12.50 pesos per dollar this was equivalent to 2.6 percent of GDP and 6 percent of gross value of domestic production of manufactures. (Exports were depressed in 1975 because of the recession in the United States and the overvaluation of the peso. In 1974 exports had been somewhat higher.) Manufactured exports grew very rapidly in the 1970-74 period -- about 26 percent per year in real terms -- but the level they at- tained is still low for a country at Mexico's stage of development. Most of these exports came about in response to special situations such as the in- crease in assembly plants of US firms, or special programs such as the auto- motive parts exports that are required of the Mexican automotive industry. The rapid growth in manufactured exports occurred in spite of Mexico's over- all policies which, except for these special cases, implicitly discriminated against manufactured exports. In 1975, disincentives to manufactured exports included the overvaluation of the peso estimated at 18.4 percent, as well as the higher cost and lower quality of inputs (owing to protection and to sales taxes on inputs), as well as sales taxes on exports. These disincen- tives together are estimated to have been equivalent to an implicit tax on exports of some 26 percent. Positive incentives including sales tax reim- bursements under the CEDI scheme, low-interest export financing by FOMEX, and somewhat better-than-average access to import permits for inputs, are estimated to have been equivalent to only about 16 percent of what sales revenue woul;'- have been in the absence of any incentives or disincentives, even when the value of CEDIs are increased to reflect that they are not subject to income tax. Thus net incentives for manufactured exports in 1975 were negative, about minus 10 percent. 4. The situation as of mid-1977 is somewhat better, as a more realistic exchange rate plus restored CEDIs reduced net disincentives to close to zero. Mexico appears to have great potential as an exporter of manufactures, with her advantages of natural resources, abundant labor, growing industrial sophistication, and closeness to the large market of the United States. This potential is suggested by the growth in manufactured exports that has already occurred, and is confirmed by the mission's analysis. Better policies would help to develop this potential. 5. In recent years, the whole package of Mexico's foreign trade policies (protection, export incentives and disincentives, and the exchange rate) has not been fully appropriate either for promoting sustained and efficient growth in output and employment or even for managing the balance of payments. Policies that give greater promise of meeting these goals include (a) a higher real 1/ exchange rate (in terms of pesos per dollar) than in the past, maintained at a stable real level by reasonably stable prices if possible or otherwise 1/ Throughout this report, "real" means net of any effects of inflation, i.e., in terms of constant prices. Keeping the exchange rate stable in real terms therefore means adjusting it in proportion to any changes in the ratio of Mexican prices to international prices. -3- by frequent adjustments, and (b) low import protection and commensurately adequate net export incentives for manufactures, with low dispersion among products. Since September 1976 the exchange rate has moved in this direction. As to import policy, low and more uniform protection could only be achieved by substantial liberalization or complete removal of the import licensing system, since this system affords potentially very high protection and thus can be used to shelter very great inefficiencies where they exist. For exports of manufactures, the cancellation of sales tax reimbursements under the CEDI scheme, and price increases, have offset the positive effects of the higher exchange rate; as of December 1976 net export incentives for manufactures are estimated at minus 14.0 percent, which is even more negative than in 1975. Export profits for most Mexican manufacturers were still too low to motivate sustained growth in a wide range of exports. The restoration of CEDIs announced on April 1, 1977 will wipe out most but not all of the net negative incentives. 6. Placing greater reliance for both protection and export incentives on a higher exchange rate and keeping that rate constant in real terms could help increase both import substitution and exports. This, in turn, would increase growth in total production and employment. Mloreover, growth under this strategy would be more efficient, because incentives would be more equal among products. Manufacturing as a whole still would, and should, be promoted, but the strategy would induce somewhat more specialization within the sector in those products that can be produced most efficiently. The strategy would also permit fuller utilization of capacity, greater scale of production, and hence reduced costs. Gains in quality would also result from reduced protec- tion and greater exposure to export markets, and thus Mexican consumers would benefit in terms of both price and quality. 7. The possibility of greatly increased exports of petroleum and petrochemicals opens up a new option for Mexico's foreign trade policy. The increased revenues from these exports can be used to relax constraints on domestic demand and investment, and hence to facilitate faster growth of GDP and employment. To do this, the peso/dollar exchange rate must be maintained at or near its present level in real terms, and other export incentives must also be maintained (or strengthened). Then not only petroleum products but other exports will grow, the total increase in foreign exchange earnings will be larger, GDP and employment will increase and the additional foreign exchange can be used to pay for the additional imports that will be demanded. The danger to avoid is allowing incentives to increase other exports to weaken. This weakening would occur if the increased foreign exchange earnings from petroleum induced the authorities to allow the dollar value of the peso to appreciate in real terms, to reduce CEDIs, etc. Such actions would reduce the profitability of other exports, with the net effect of substituting petroleum for other exports with no gain in GDP or employment. 8. Employment and wage policy, in this report, are analyzed only from the viewpoint of the manufacturing sector. In this limited context, the most effective measure to increase employment would be the foreign trade policies just mentioned. These policies would promote both import substitution and - 4 - exports of products Mexico can produce efficiently, which should imply much greater use of Mexico's abundant supply of labor. Another complementary ap- proach would be to improve support of small-scale industry, through greater access to credit and technical assistance as well as the improved access to material inputs that would result from liberalization or abolition of import licensing. Also, fringe benefits that are financed from taxes or other charges that are proportional to wages add as much as 50 percent to basic wage costs. These burdens, combined with other labor laws that make firing or laying off workers very difficult, greatly increase incentives for employers to substitute equipment for labor. Shifting the financing of some fringe benefits to taxes on value added or on income, in place of taxes or other charges that are proportional to wages, would reduce the cost of labor to potential employers. This would increase employment, and through it, equity, as the relatively high-paying jobs typical of the manufacturing sector become available to a greater part of the Mexican labor force. Finally, the large wage increases of 1976, which were far above price increases and productivity growth, could rapidly erode Mexico's competitiveness if repeated in the future. 9. Fiscal policy is treated only partially in this report, but is important for the objectives dealt with here. (Fiscal policy will be dealt with in greater depth in the report being prepared by the economic mission that visited Mexico in April/May 1977.) Large public sector deficits have fueled inflation, which in turn has made Mexican exports less competitive, has led to greater use of import licensing to counteract the fall in the real peso price of imports, and has negatively affected saving and financial intermedia- tion in Mexico. Moreover, to help finance the growing public sector deficits of recent years, credit has been withdrawn from the private sector. Thus inadequate fiscal discipline in the past has reduced industrial investment and growth, efficiency of production, and exports. Reversing these trends is important for the future. Moreover, the adoption of foreign exchange policies to increase output, exports, and employment (as summarized in paragraphs 5-7) would increase the need for managing public finances in a non-inflationary manner. Analysis of the details of a sound fiscal policy is beyond the range of this report, but one important aspect could be improving the management of public sector manufacturing enterprises, especially requiring them to cover a fixed share of their current expenses and investment requirements, so as to increase their incentives to cut costs and to adjust prices to cover reason- able costs more fully. 10. A return to steady industrial growth requires controlling the infla- tion that has been accelerating for the last few years, contributing to the climate of uncertainty that had brought domestic private investment in Mexican manufacturing to a virtual standstill. The Mexican government is now taking steps to reduce inflationary pressures. The mission's analysis of the manufac- turing sector suggests that an approach to price stability involving restraint by government, management and labor should consider: (a) A sound fiscal policy, as mentioned in paragraph 9. (b) Reduction of the high protection for some products permitted by import licensing, combined with tariff revisions to -5- rationalize the system, would benefit consumers by doing away with the shelter- ing of the few highly inefficient producers that take advantage of the licens- ing system to pass on their high costs by raising prices in the domestic market. (c) Changes in taxation to tax away some of the increased profits that could result from the changes in foreign trade and wage policies, and to increase motivation to re-invest most of the rest of such profits. (d) Not allowing real wage increases to exceed productivity gains, especially for the next year or two. 11. Many public sector enterprises in Mexico function well. However, some others are a significant drain on public resources because of high pro- duction costs and low product prices. As a whole, public sector manufacturing enterprises have been incurring losses of 2 to 4 percent of net worth in each of the last few years. To increase the contribution of public sector enter- prises to Mexico's development, the management situation which permitted this should be reviewed, and the recent Mexican Government reorganization seems designed, at least in part, to do this. Another important aspect of the performance of public sector enterprises is their virtually unlimited access to duty-free imports, especially of capital equipment. This is one of the principal causes of lagging development in capital goods production in Mexico. These policies are also responsible, in part, for the large debts that some public sector enterprises have accumulated. Equity contributions by the government would be useful to restore financial health to enterprises that are basically productive, while increased fiscal discipline as required by the recent administrative reform and other laws and regulations should guard against repeating these problems. 12. Private foreign investment has played a growing and useful role in Mexican development, and can continue to do so in the future. The restriction on foreign ownership to 49 percent of equity (less in some branches of industry) is sometimes relaxed if the investment is in a sector or region judged particularly important to the country's development. Explicit public notice of such relaxation in specified conditions, rather than by ad-hoc implementation, might help to make the policy more effective. Also, the strategy now being pursued of promotion of tripartite joint ventures, by public, domestic private, and foreign capital, offers many advantages and should be strengthened. 13. Mexico's well-developed system of financial intermediation has been badly set back during the last few years. Credit available to private indus- try has been doubly squeezed, by a drop in the mobilization of savings and by a rise in the share of total credit going to the public sector. These trends have reduced credit outstanding to the private sector as a whole from 22 or 23 percent of GDP during 1970-73, to 15 percent of GDP in 1976. The private industrial sector has been squeezed even further, as agriculture and commerce have maintained their shares of the total. Credit outstanding to manufactur- ing enterprises dropped from 36 percent of value added in manufacturing in 1969 to 32 percent in 1975; in 1976 it appears to have continued to fall, at an even faster rate. This credit squeeze could be reversed by reducing the public sector deficit, restoring reasonable real interest rates paid to depositors, and managing foreign trade policy so as to increase confidence in -6- the peso. The Bank of Mexico seems to be aware of the need to provide ade- quate working capital to help firms affected by the adjustments to higher exchange rate, price increases in public services, etc., and that the situation requires careful monitoring and prompt action to avoid unnecessary and costly failures. 14. Taking a long-run view, Mexican industry has been growing and much of it is reasonably efficient. However, in the last few years the situation has been considerably less favorable. Compared to most of the past few decades, the 1974-76 period has seen increasing inflation, declining invest- ment, restricted access of the private sector to credit, and wage increases that were more rapid and less related to productivity growth. Since August 31, 1976 the abandonment of the peso-dollar parity of 12.50 to one, and subsequent instability in the exchange rate, prices and wages, had produced confusion and a crisis of confidence in the months that followed. Mexico needed a new set of policies that both in fact and in appearance confront both the short-run problems and the longer-run needs of the economy, providing a return to healthy growth. This confusion and air of crisis seem now to have passed, as the basic soundness of the economy, the tradition of good perform- ance in the past, and many of the actions the new Government has already taken or has declared its intention to take increase confidence in Mexico's future. 15. One of the problematical aspects of the situation when the mission visited Mexico in November 1976--perhaps difficult to avoid in that transi- tional period--was the lack of a coherent set of policies. Would wage and price increases be allowed to erode the hoped-for benefits of the devaluations of the peso? How can the competing claims of workers, public sector enter- prises, private business, and the government itself be resolved in a way that reduces inflation and allocates resources to where they are needed and where they would be most productive? How can the words and the actions of the government help to increase confidence, continue to attract both domestic and foreign saving, and invigorate investment? Any report must, necessarily, discuss policy measures in sequence. But the measures are dependent on each other for success, and must be so designed. The sequential nature of the exposition should not be mistaken for a lack of interrelations between the individual items. 16. Text Table 1 presents a brief overview of the policies discussed in this report. The three overall objectives that orient our analysis are listed across the top: efficient growth, balance of payments management, and employment. From top to bottom are the major policy areas studied: the exchange rate, import controls, export incentives, prices, wages and labor policy, public sector enterprises, private investment both foreign and domestic, and industrial credit, as well as the two more narrowly defined parts of the manufacturing sector that were studied in more depth: small and medium scale industry, and capital goods. Entries in the table summarize the measures suggested for consideration, and their likely effects. 17. Text Table 1 contains several examples of the interdependence of policies. To cite just one here: Removal or substantial liberalization of the import licensing system would probably increase imports to some extent. -7- Text Table 1 MEXICO: OVERVIEW OF POLICIES SUGGESTED 'FOR'CONSFUDFRATION O B J E C T I V E S Policy Areas Efficient Growth Balance of Payments Employment and Income Management Distribution Exchange Rate: Keep rate at or near equilibrium Creates export incentives Restores confidence and Promote exports and import in order to substitute for most and protection against brings both capital and substitution in all activi- protection and export incentives imports while avoiding current accounts more ties; thus increasing em- and to manage balance of payments price distortions and need toward balance even at ployment. Available income at a high GDP growth rate; keep for import controls that higher growth rates. increases and its distribu- real value more or less constant would otherwise lead to Eliminates implicit tax tion becomes more equitable, in future. A level in the range inefficiency and lagging on exports that stemmed hased on real contributions of 20 to 22 pesos per dollar, at productivity; increases from overvalued peso, and to grovth ratber tban on December 1976 prices, could be import capacity enabling thus increases exports. distorted incentives. appropriate with other measures growth to accelerate. suggested here. Import Controls: Dismantle or greatly liberalize Strengthen role of prices Probably increase imports. Limit price increases and licensing system in a phased in resource allocation; However, higher exchange thus help to maintain con- sequence starting with inputs spur to increased efficien- rate and tariffs will limit sumers' purchasing power. into producer goods. Rational- cy. Improve access of these, and greater exports ize tariff system to provide manufacturers to inputs will more than pay for protection averaging, perhaps, at reasonable cost and them. 10 to 15 percent for manufac- quality. turing. Eliminate waivers of public-sector tariff payments. Create an administratively ef- fective system of drawbacks on import duties for exporters. Export Incentives: Equilibrium exchange rate plus Exports would be products Reverses two negative as- Increased demand for labor somewhat stronger CEDIs; main- which are produced effi- pect of past policies, by in manufacturing and in all tain incentives indefinitely. ciently. making export profitabi- export activities; indirect Since import substitution lity (a) high and (b) nearly effects further increase opportunities are becoming equal to that of domestic employment throughout the limited, and Mexican manu- sales. Enables balance of economy. facturing is maturing, a payments equilibrium at shift toward export markets higher growth rate and with is both natural and neces- more capacity to import. sary for continued rapid growth. Prices: A sound macroeconomic policy, Relative prices allowed to Helps to keep exchange Limit price increases and beyond the scope of this report, move closer to relative rate roughly constant, thus help to maintain con- is essential. Within such a costs (through liberaliza- in real terms at least. sumer's purchasing power. policy replace spotty price con- tion of imports and reduc- Maintains Mexico's compe- trols and below-cost provision tion of implicit tax on titiveness in both domes- of goods and services from some exports that stemmed fromi tic and foreign markets. public sector manufacturing overvalued peso); will enterprises with economy-wide increase efficiency and real price restraint, with target growth. level consistent with a modera- tely restraining fiscal and monetary policy. Wages and Labor Policy: Limit increases to productivity Reduce incentive to substi- Maintain Mexico's advantage Encourage employment ex- growth during stabilization tute capital for labor; of lower labor costs vis- pansion both by use of labor period. Replace annual renego- reduce costs of negotia- a-vis major trading part- as a substitute for labor- tiations with 2 or 3 year agree- tions and uncertainty. ners; induce expansion of displacing capital equip- ments providing for automatic ad.- Greater use of available exports and of efficient ment and through greater justment. Replace financing of supply of labor will in- import substitution. growth; improve income some benefits from wage taxes crease output. distribution among workers by financing from general reve- by increasing the number of nues. jobs in industry (which are relatively high-paying in comparison to the rest of the economy). Text Table 1 (cont.) O B J E C T I V E S Policy Areas Efficient Growth Balance of Payments Employment and Income Management Distribution Public Sector Manufacturing Enterprises: Partly replace financing of defi- Motivates greater effi- Reduced inflation will make cits and expansions through ciency in operation and balance of payments manage- Treasury funds by revenues of keep prices from falling ment easier. the enterprises. Remove waiver below costs; reduce defi- on import duties for inputs. cit of public sector. Reduced demand for credit from public sector will increase supply to private sector. Private Investment: Replace subsidies through cheap- Restore incentives to in- Increases employment by er access to equipment and vest in efficient sectors reducing incentives to imported inputs, by temporary through (a) better access substitute capital for income tax relief (where to imported inputs, (b) high- labor and further by in- subsidies are warranted); chan- er export profits, (c) tem- - creasing investment. nel employer's gains from porary corporate income tax Promote equity by discoura- wage restraint and export prof- relief in areas where sub- ging luxury consumption. its into productive uses sidies are needed and (d) through interest rate, tax and wage restraint: these credit policies that favor measures compensating for savings and reinvestment while higher costs of some equip- discouraging luxury consumption. ment and other current in- puts. Alleviate credit squeeze. Foreign Investment: Apply 49% participation limit Continue to attract needed Maintain or increase with flexibility; make permis- capital and technology. foreign capital inflow; sible relaxation explicit in facilitate purchase of cases such as investment in technology where needed priority sectors or regions. (e.g. in capital goods). Continue to promote tripartite Motivate investment for joint venture with domestic export markets. Improve private and public investors. access to export markets. Industrial Credit: Hiuher real interest rates and Restore flow of savings to Maintain output; expand Maintain and expand em- stable real value of peso. Reduce financial system; restore investment for export ployment in efficient public sector deficit. Assure share of private sector in production and efficient enterprises that would adequate flow of credit for work- access to savings; maintain import substitution; otherwise fail for lack ing capital. output by avoiding squeeze reduce new foreign in- of credit; promote equity on working capital. debtedness by expanding the by allocation of credit the flow of finance from based mainly on efficien- within Mexico. cy and future potential, rather than existing wealth or foreign connec- tions. Small and Medium Scale Industry: Increase access to credit and Remove constraint to growth Provide increased employ- technical assistance, especially of small-scale sector. ment and increased rewards for smaller firms. Liberalization of import to small-scale entrepre- licensing will also move neurs. in this direction. Capital Goods: Eliminate privileged access of Move towards Mexico's compa- Substitute for imports Expand production in a public sector to imports; provide rative advantage in this where domestic production sector with a relatively high moderate tariff protection; sector. is not too costly; increa- labor/output ratio. analyze growth possibilities sed specialization should for intermediate inputs and end also increase exports. products; promote new projects and assist existing firms ac- cordingly. -9- However, avoiding overvaluation of the peso, eliminating large-scale exemp- tions from paying tariffs, and providing moderate tariff protection for products where tariffs are now low or zero because of reliance on refusal of import licenses could contain any increases in imports within safe levels. The same exchange rate policy plus restoration of CEDIs at an adequate level could increase exports by more than enough to pay for the increased imports. This policy package would induce growth in output in efficient enterprises (and in particular could provide part of an appropriate environment for promotion of efficient import substitution and exports of capital goods). At the same time, the import liberalization could help to restrain unreasonable price increases and help maintain consumers' purchasing power in the face of wage restraint, and is in itself needed to cut costs and provide access to inputs of good quality and reasonable price, in order to promote exports. ANALYSIS I. Foreign Trade Policies 18. This set of policies includes the exchange rate; import tariffs, permits, and other regulations; and export incentives and disincentives. It is assumed that the important goals to be met are (a) in the short run, helping to increase confidence in the peso and in the economy, and (b) in both the short and the longer run, managing the balance of payments and in- ducing efficient growth in output and in employment. 19. The key themes of the mission's suggestions on foreign trade policy are increasing generalized incentives for exports of manufactures, and general- izing and rationalizing the already strong but uneven incentives for import substitution. Mexico's manufactured exports have been growing rapidly, but are still only a very small share of Mexican manufacturing production. The country has great potential for continued growth of exports of manufactured goods, which could be realized by better exploiting its growing know-how, abundant labor, natural resources, and proximity to the US market. A stronger export orientation for all Mexican manufacturing could serve many important purposes. (a) More exports are needed for managing the balance of payments, including the severe debt service burden of the next few years. Even import substitution that is efficient in the long run will not help much in the next year or two because most remaining opportunities for import substitution in Mexico would themselves require considerable imports of equipment and some current inputs. (b) More exports are needed to provide employment; exports provide markets for additional production and many export products that would appear under generalized incentives would be more labor-intensive than manu- facturing on the average. (c) Finally, such exports almost by definition are goods which Mexico produces efficiently and thus they would contribute to a greater growth of output in real terms. 20. Mexico's present system of protection through licensing provides potentially very high protection to many manufactured products, and less to - 10 - others. The system insulates Mexican producers from motivation to cut pro- duction costs or to increase quality. This allows considerable inefficiency in domestic production, most of which could be eliminated if motivation were sufficient. Some of this inefficiency consists of producing goods in which, at least at present, Mexico cannot be competitive. But by far the largest part of the inefficiency consists of higher costs and/or lower quality that could be improved upon if the producers had to compete with imports or if export incentives made competing in export markets more attractive. To induce greater efficiency, Mexico could increase access to imported products where domestic costs are unreasonably high, and also increase the profitability of exports. This would permit more Mexican producers to increase volume and profits through successful competition in international markets. Increases in competition as well as more efficient specialization could induce faster growth. Past Policies 21. The exchange rate was kept constant at 12.50 pesos per dollar from 1954 until September 1976. Price increases in Mexico stayed about equal, grosso modo, with those of the United States and others of Mexico's trading partners until the early 1970's; protection against imports, as well as export incentives, also did not change very much. Around 1971-73 this behaviour changed. Prices rose faster in Mexico than in her principal trading partners. Mexican authorities chose to allow the peso to become overvalued, and to com- pensate for the overvaluation by increasing apparent protection and apparent export incentives. 22. In any process of increasing overvaluation of a currency, with com- pensating increases in protection and export incentives, the true change in protection (or in export incentives) is the net effect of the two opposite events: increasing overvaluation decreases protection (or export incentives) while the explicit measures such as more restrictive import licensing or tax rebates for exports provide compensating increases. For this reason, in this report we refer to the explicit measures as "apparent" protection or "apparent" export incentives, while the effects after estimated overvaluation has been netted out are referred to as "net." 23. Exports: Mexico's manufactured exports have grown rapidly from a low base. Over the five-year period 1971 through 1975 they grew by 30 percent per year, of which about 12 percent per year was attributable to price in- creases and 16 percent per year represented real growth. Text Table 2 sum- marizes this growth. - 11 - Text Table 2 Mexico: Manufactured Exports, 1970-1975 /a (million dollars) "Maquiladora" Year Assembly Plants Other Total 1970 218.8 353.5 572.3 1971 270.0 454.2 724.2 1972 426.2 574.9 1,001.1 1973 651.2 837.9 1,489.1 1974 1,032.9 1,242.9 2,275.8 1975 1,020.6 1,069.3 2,089.9 /a The figures exclude primary nonferrous metals, sugar and petroleum products. The full value of product of assembly plant exports are reported, making them com- parable with other exports. 24. In relative terms these industrial exports are still rather small -- only 3 to 8 percent of those of European countries with populations roughly comparable to Mexico's but whose income is three to six times as much (West Germany, United Kingdom, France, Italy), and also very much smaller in per capita terms than manufactured exports from successful LDC exporters such as Taiwan, Hong Kong, The Republic of Korea and Singapore. However, Mexico's industrial exports, including assembly products, are roughly double those of Brazil in per capita terms. 25. Nearly half of Mexico's manufactured exports come from plants that assemble products for the US market based at least in part on inputs imported in bond from the United States, taking advantage of special provisions in the US tariff that limit import duties for such products to value added outside the United States. About two-thirds of assembly-plant exports consist of items such as electronic parts, television and communications equipment; clothing is a distant second. 26. Exports other than those of assembly plants are more diversified. Exports of non-electrical machinery, electrical machinery, and transport equipment industries have grown considerably, to $270 million (25 percent of non-assembly-plant manufactured exports) in 1975. Most of these exports consist of parts or components, led by automobile engines and parts, and most go to the United States. However, automobile parts exports are required of the automotive industry, and may not all be produced at internationally competitive costs (see paragraph 28). Chemical exports ($204 million in 1975 -- 19 percent of non-assembly manufactured exports) are largely based on - 12 - natural resources such as sulphur, lead, zinc, and barbasco (used to produce hormones); these go mostly to the USA and LAFTA countries. Textile exports, based on locally produced henequen and cotton, and food product exports, led by molasses and simply preserved fruits and vegetables, are also important. In spite of the concentration of exports in a few sectors, however, the only sectors exporting 10 percent or more of their output in 1974 (the last year for which these comparative data are available) were basic chemicals, fertil- izers, and pharaceuticals--all principally based on natural resources. (Large food product exports were mostly sugar, not treated here as a manu- factured export.) (See Tables 1.2, 1.3, and 1.4 for more details.) 27. Manufacturing exports have been inhibited by a number of serious disincentives. The most important of these negative elements have been (a) the progressively more overvalued exchange rate that prevailed through the 1970s, until the floating of the peso in September 1976, and (b) problems of high cost, low quality and unreliability of manufactured inputs, mainly owing to indiscriminate protection by the import licensing system. The overvalued exchange rate squeezed export profit margins, as the peso value of exports was kept from rising while peso costs of wages and other domestic inputs rose. At the same time, some of the intermediate inputs produced in Mexico and protected by refusal to grant licenses to import competing products were too expensive or not good enough in quality to permit their use in competitive export pro- ducts. Other disincentives include (c) the inflexibility of some of the labor laws: high costs of discharging workers and difficulties in achieving tempo- rary layoffs discourage investors from expanding their labor forces; (d) high costs and delays of shipping through the main Gulf ports of Veracruz and Tampico; and (e) heavy use of discretionary measures in policies affecting exports: "deals" offered and rules applied vary greatly over time, and even among firms in the same industry. A potential exporter cannot be sure what incentives he will get, whether he will be able to import certain inputs, etc. The uncer- tainty that results is a great discouragement to investments to produce exports. Restrictions on imports, on the part of countries which would other- wise be buying more Mexican goods, are also an inhibiting factor. 28. Many exports that appeared in spite of these disincentives have been brought into existence by special policy measures. For example: (i) Automobile manufacturers are required to achieve either exports or increased domestic content in specified ratios to imported inputs in order to raise the numbers of vehicles they are allowed to produce each year. At least two-fifths of the exports are required to come from parts manufacturers. Accord- ing to one study, in 1974 Mexico's motor vehicle industry achieved $154 million worth of exports, consisting mainly of automobile engines and parts, compared to $498 million worth of imports. (Vazquez Tercero, 1976) To the extent that these exports are not produced at competitive prices, the costs of subsidizing both exports and domestic manufacture are borne by Mexican consumers. - 13 - (ii) Costs that processors have to pay for many natural resource inputs are reduced, by various means such as export taxes, export restrictions, or price controls that hold prices in Mexico below the world market price. The coverage of these measures has varied; at one time or another they have extended to most potential agricultural and mineral exports, including industrial inputs varying from zinc and sulphur to citrus and strawberries. In the case of babasco, used to manufacture hormones, exportation has been prohibited outright. Such measures, by lowering the costs of raw materials relative to the market value of exports, contribute importantly to Mexico's exports of manufactures based on natural resources. Costs are borne by primary producers whose sales prices are controlled or whose exports are taxed or restricted. (iii) Industrial exports under accords with LAFTA countries--to which Mexico exported over $125 million of manufactures, narrowly defined, by 1973 (see Table 1.4)--are based on mutual waivers of protection, giving these goods privileged access in markets such as Brazil or Argentina. Costs are borne indirectly by Mexican consumers, who must pay high prices for other products imported from Mexico's LAFTA partners in exchange. (iv) Low-interest financing by FOMEX has been a significant help to many exporters of manufactures. The advantage to exporters who receive these loans, which are available for virtually all manufactured exports and without excessive administrative diffi- culties, is on average equivalent to an additional 2 percent of export prices. (v) Many enterprises consider it good policy to export to increase their chances of being allowed imports, and many others export modest quantities at prices that may be well below average costs, in order to achieve fuller utilization of capacity. Some exports remain unprofitable in the long run but are continued because they cover variable costs; thus some of the exports in recent years are a lagged effect of investments undertaken earlier when conditions were more favorable. (vi) Finally, the most important of the incentives that were generally available to exporters of manufacturers were reimbursements for payment of indirect taxes--CEDIs. 29. CEDIs, in effect from 1971 to August 1976, were the single most important incentive to manufactured exports. These tax rebate certificates were temporarily eliminated after the fixed parity of the peso was abandoned in August 1976, were later reinstated on a selective basis for industries that could demonstrate a "need" for them (this is an example of the ad-hoc applica- tion of incentives, which is much less effective than automatic across-the- board procedures), and were reinstated generally just as this report was - 14 - completed in April 1977. While CEDIs were in effect before August 1976, the producer for export received certificates worth 11 percent of the sales price of exports if the domestic content (material inputs made in Mexico plus value added) of his product was 60 percent or more; these certificates could not be negotiated or transferred, but could only be used to pay taxes. This tax rebate was not subject to income tax (which is generally 42 percent for corporations) so that for many recipients the after-tax effect was equivalent to a 19 percent increase in their sales price. However, some exporters pay little or no income tax. In calculations for this report the mission has assumed that two-thirds of all CEDIs went to companies with sufficient income tax liability, so that on average the value of CEDIs was equivalent to a 16 percent increase in the sales prices of exports. Under the old system, CEDIs helped exports considerably but had some shortcomings. For exporters whose product consists mostly of purchased inputs and only very little of value added, CEDIs increased profit margins enormously, while for others with a different cost structure or with low tax liabilities, their value was small. The system did nothing to compensate producers of inputs that were incor- porated into exports by other producers, and the CEDIs given out bore very little relation to the net contribution of each export to the balance of payments, since the reward was the same whether domestic content was 60 percent (which could mean much less in international prices) or 100 percent, and regardless of the extent to which the export used exportable inputs or imported capital equipment. 30. Under the new CEDI system decreed in April 1977, the value of the reimbursement varies from 25 to 100 percent of indirect taxes paid, depending on (a) how much of the gross value of product represents value added (as opposed to purchased raw materials and intermediate goods); (b) the domestic content of the product, and (c) whether the exporter has increased his exports in the last year. For example, if the product has only 30 percent domestic content, and the value added by the exporter is relatively low, the CEDI is only 25 percent of the taxes paid. More domestic content means higher CEDIs; at 80 percent or above the exporter receives 50 percent of taxes back. Similarly, higher value added means higher CEDIs; high value added and 80 percent or more domestic content earns the exporter 80 percent of his indirect taxes. Finally, export increases can earn up to an additional 20 percent. To receive reimbursement for all of his indirect taxes, the exporter must have 80 percent or more domestic content, high value added, and a 15 percent annual increase in his exports. 31. Other export incentives are: (a) Administration of import and in- vestment licensing by SIC favored export activities. Requests for import licenses were more likely to be granted, and were sometimes granted more quickly, if the need for the import was justified by a need to compete in export markets. (b) In general there was increased official concern over manufactured exports, as evidenced by various promotional activities such as those of IMCE. (c) The Banco de Mexico appears determined to direct financing to export-oriented firms in the present credit squeeze. 32. With proper policies, Mexico appears able to become a major exporter of industrial products, on a scale many times as large as today, provided that its industries are encouraged to specialize in what they can produce best. - 15 - This achievement would speed the nation's advance into the ranks of the indus- trial countries, not only through the exports themselves but also because the types of policy environment and managerial and technical efficiency needed to meet this challenge are also what MIexico requires to improve its production for the domestic market, to raise income and to increase employment. 33. Mexico's comparative advantage in manufactured exports appears to lie especially in two broad areas. First, some manufactured exports can be and are based on local primary raw materials; progress in these industries depends heavily on the associated primary producing sectors. Second, there is an especially promising future in exporting products to the US that com- bine two features: substantial requirements for manual labor, and moderate transport costs and/or need for fast delivery. In products for which trans- port costs are too high, Mexico cannot compete with US producers; where transport costs are very low, Mexico cannot compete on even terms with Asian countries as a source of cheap labor. Mexico does enjoy special advantages in an enormous variety of exports subject to moderate transport costs, such as some building materials, furniture, other household items, a wide range of metal products and parts and assembly activities, perishable food products, and fashion-sensitive clothing requiring rapid delivery. While some of these exports may be specialty items produced in small batches, many other opportunities depend on mass production of goods in which Mexico has a poten- tial cost advantage vis a vis the United States. To realize such advantages requires specialization within each industry, so that high volumes of output can be achieved. Examples where early success can be expected include motor vehicle parts, components for machinery and transport and electronic equipment, assorted simpler metal products, and glass products. However, specification of the precise items is difficult to predict. The best approach may be to adopt generalized incentives and let the market indicate just where Mexico can best compete and, more, to make the market stimulate the potential exporters to come up with better and with new export items. 34. Mexico's potential can be realized rapidly by embracing a strategy that has been followed in periods since World War II in a number of coun- tries, with excellent results -- West Germany, Italy, Japan, Austria, Fin- land, Norway, Greece, Taiwan, Korea and Singapore would be leading examples. The essence of this strategy is to create a stable policy environment with a strong government commitment to expand exports and keep them profitable. This requires giving producers access to the inputs and other help they need to be competitive, and having an exchange rate, relative to domestic wages and othe: costs, at a level where local prices appear low and the country becomes a really attractive place to invest and expand output. In this strategy the exchange rate becomes the main source of protection against imports as well as of export competitiveness. By compressing the foreign currency value of incomes, non-essential imports can be kept in check and, since export earnings grow swiftly, the country can afford whatever imports of capital equipment and intermediate inputs are necessary. 35. Imports and protection: Mexican manufacturing has developed rap- idly and has diversified to a wide gamut of products during the last two or three decades. By 1969 imports contributed only 10 percent of total supply of - 16 - manufactures in Mexico. Of non-durable consumer goods, imports were less than 2 percent of total supply; for intermediate goods the share was 11 percent and for durable consumer and capital goods together the share was 29 percent. (Nafinsa 1971). The share of manufactured imports as a whole increased in the early 1970s, but this was due to the overvaluation of the peso and to inventory speculation in anticipation of possible devaluation. Looking at sectoral detail, by 1974 imports were significant only in basic chemicals, capital goods, synthetics and plastics, and the miscellaneous category. (See Table 1.1). 36. This development was promoted in large part by a conscious policy of import substitution during the last few decades. Control of imports has relied more on quantitative controls than on tariffs. The percentage of products for which import permits are required has risen steadily over the last 20 years; from about 33 percent in the late 1950s, the percentage rose to 50 percent in the early 1960s, reached 75 percent in the early 1970s, and in November 1976 was about 85 percent. The percentage by value of actual imports that required licensing also rose, although less rapidly, from 28 percent in 1956 to 74 percent in 1974. 37. The general principle of operation of the system is that imports that compete with locally produced goods are not allowed. In some extreme instances of high cost of the domestic product, the authorities have granted import permits. But as a rule they were prepared to allow substantial mar- gins of inefficiency -- as reflected in the prices -- to protect goods pro- duced in Mlexico. Of course domestic competition, the possibility of smug- gling and the increase in the size of the domestic market reduced inefficiency in many established industries. Nevertheless, the erratic levels of protec- tion generated by the licensing system are reflected in the wide dispersion of rates of protection, already noted in paragraph 2. Imports, moreover, have been subject to numerous exemptions from tariffs, most notably imports by the public sector and some imports of capital goods, so that tariff collections were only 6.5 percent of total merchandise imports. More than one-half of imports (33 percent of private sector imports and 89 percent of public sector imports) paid no tariff at all in 1975. 38. Actual protection as measured by international price comparisons, however, was higher. The apparent average level of protection of manufac- turing was 25 percent in 1960, 18.8 percent in 1970, and 40.9 percent in 1975. The pattern of protection is typical: primary production receives low rates, traditional industry higher rates and "modern" consumer durables still higher rates. Unfortunately, the two-digit industrial classifications that were used in studies of protection aggregate products with widely dif- ferent levels of protection in Mexico, and therefore impede analysis of appropriate groupings of products. For example, the two-digit sectors enjoy- ing the highest protection, on average, in 1975, were textiles, beverages, basic chemicals, synthetics and plastics, and cosmetics. However, the really high protection is only evident at higher levels of disaggregation. High protection is scattered throughout most manufacturing sectors, usually for a few products in each. This is the result of the licensing system which grants protection with little or no rationale, other than to pro- tect domestic production, and with little or no regard for relative costs. - 17 - 39. While protection. did help to induce industrialization in Mexico, and industrialization has been a key element in raising incomes, diversifying and increasing exports, and modernizing Mexican society, there have been signi- ficant costs associated with protection in Mexico. One way to estimate some of these costs, as presented in Bergsman (1970), is to view protection as leading to higher prices owing to three different factors: (a) higher costs that are unavoidable, in which case protection is sheltering industries which, at least at the time, cannot be efficient in Mexico; (b) higher costs that could be reduced, in which case the industries could be efficient in Mexico but the producers are using protection to relax their cost-reducing efforts; and (c) monopoly returns, in which case the industries are efficient in Mexico but the producers use protection to charge higher prices. Reducing protec- tion would reduce all three effects, in the first case by replacing domestic production by imports while in the second and third case domestic production would continue but prices would be reduced. Estimates for 1960 place the size of the first effect at 1.5 percent of value added in manufacturing, and the second and third combined at 11 percent of value added in manufacturing. This means that Mexican manufacturing could have produced much more output if distortions induced by protection had been absent, with only very small replacement of domestic production by imports. Changes since 1960 suggest that the allocative costs may have risen; the total of these effects may be esti- mated currently at about 15 percent of value added in manufacturing, or over 40 billion pesos. 40. Indiscriminate protection behind an import licensing system is also a major hindrance to manufactured exports. Such protection not only increases the costs of intermediate inputs but in many cases results in lowering their quality. Experience in other countries suggests that real success in export- ing manufactured goods virtually requires that exporters have access to inputs at world market prices. 41. Another kind of cost of protection is that incurred by the govern- ment in administering the licensing system and by the private sector in dealing with it. Administratively, the licensing system causes some prob- lems for many importers but not in all cases. In general, if the import is permitted its quantity is not limited. Thus, firms can import without restric- tion certain inputs or machinery which by no stretch of the imagination are produced in Mexico. In other cases, firms resign themselves to using the domestic input even if it is more expensive or of lower quality, adapt their production processes to the situation, and pass the price or quality differ- ence on to the consumer. Administrative problems arise when it is not clear whether the domestically produced good is "really" the same as the imported product. For these cases some firms and trade associations maintain staff of "expediters" and "public relations experts," and invest a substantial portion of the time of top management to "cultivate" the import licensing authorities. The Ministry of Industry and Commerce in turn employed about one thousand persons and substantial computer time in the process of granting permits and other matters regulating imports. Of course firms that are too small or too far away from the capital to become involved in the import permit process have to buy from distributors, often at a substantial premium, or do without - 18 - imports entirely. Controlling imports by licensing discriminates against small-scale entrepreneurs because they have less access to imported inputs and less ability to insure protection of their products by influencing the authorities to deny licenses for competing imports. Capital goods production is also very sensitive to the negative effects of licensing, because most capital goods are made of literally hundreds of parts which must meet certain tolerances or other quality standards. Uncertainty as to availability of imports and mistrust of protected domestic suppliers lead producers of end- product capital goods to produce more of such parts themselves, at smaller production volumes than is economical. 42. Net protection and export incentives: The combined forces of export incentives and disincentives, protection, and exchange rate policy can be summarized in numerical terms. Measures of many of the relevant variables are available only for 1970 and 1975, which is sufficient to show what happened: 1970, very roughly, is representative of the period 1960 through about 1971; changes from 1970 to 1975 show the trend from 1972-73 through August 1976. The analysis is summarized in Text Table 3 and Chart 1.1. The estimates for all incentives and disincentives are expressed in terms of equivalent effect on sales prices, so as to be comparable with each other and with changes in the exchange rate. For example, a 6 percent increase in input prices would be expressed as (6) (.62) = 3.7 percent, if inputs were 62 percent of product price. 43. From 1970 to 1975, apparent protection to Mexican manufacturing (as measured by price comparisons) rose from 18.8 to 40.9 percent. On the export side, the apparent effect of all incentives and disincentives taken together was changed from an implicit tax on manufactured exports of 12.0 percent in 1970 to an apparent subsidy of 7.1 percent in 1975. This apparent subsidy consisted of the positive effects of CEDIs (16 percent) and other incentives (3 percent), offset by the effects of sales taxes (8 percent) and higher input costs due to protection (5 percent). (See Annex II for details.) 44. As noted earlier the apparent increases in both protection and ex- port incentives were in large part compensation for an increasingly over- valued peso. Mexican prices were rising, relative to prices in her principal trading partners. Moreover, Mexico's balance of payments deficit was increas- ing. Both of these required devaluation or some substitute; the Mexican authorities chose to increase protection and export incentives as an alterna- tive to devaluation. As a result, the apparent increases in protection and export incentives were in large part illusory. 45. To make more meaningful estimates of levels of protection and of export incentives, the mission has estimated "free trade equilibrium" exchange rates for 1970 and 1975. These are the exchange rates at which Mexico's bal- ance of payments would have been at or near equilibrium if all protection, export incentives and disincentives were removed. Details of the methodology are explained in Annex II. Estimates of protection or export incen- tives relative to this exchange rate, referred to as "net protection" or "net export incentives," measure the net effects of all of Mexico's foreign trade policies, relative to a situation in which imports and exports are neither - 19 - promoted nor discouraged. These net estimates can also be compared over time -- because the effects of any changes in overvaluation of the peso have been removed, the estimates measure net or true changes in protection or export incentives. Changes in net protection or net export incentives can also be approximated by changes in real effective exchange rates, as shown in lines 4 and 5 of Text Table 4, on page 25. 46. The estimated free-trade equilibrium exchange rate depends on what other policies are in force. For example, it would be increased by greater governmental spending or by faster growth; it would be decreased by more conservative macroeconomic policy or by increased exports of petroleum. Thus it is not necessarily the "right" exchange rate (this latter depends on the objectives of policy) but simply the rate at which Mexico's balance of pay- ments would be in equilibrium given a particular set of other policies. 47. Overvaluation is estimated by this method at 2.8 percent in 1970 and 18.4 percent in 1975. 1/ When the apparent levels of protection against imports of manufactures (18.8 percent in 1970 and 40.9 percent in 1975) are reduced to net out overvaluation, they fall to 15.6 percent in 1970 and 19.0 percent in 1975. On the export side, net export incentives foir manufactures are negative in both years: -14.4 percent in 1970 and -9.5 perlcent in 1975. 48. These policies kept Mexican export profitability low, and certainly much lower than the profitability of sales within Mexico. In 1975 a typical manufactured product worth, say, $100 in the world market, could be produced and sold in Mexico for the peso equivalent of $118.40. 2/ The same Mexican producer would receive the peso equivalent of only $90.50, however, if he exported his product. If his profit margin on the domesticisale was 20 percent, exports would have implied a loss of $4.22 per unit.i 49. These conditions were not much better as of December 1976. For exports, the depreciation of the peso added about 60 percent to unit revenue, but the loss of CEDIs plus labor and materials cost increases since 1975 were equivalent to an even greater drop in unit revenue. Net export incentives were still negative, at about -14.0 percent. On the import iside, net realized protection was about the same as in 1975, 3/ as was the potential protection 1/ The estimated overvaluation is slightly less than other approaches have calculated. Reasons for this are explained in Annex II. Briefly, the two main reasons are that the estimates take into account the (im- portant) effect of protection on raising costs of inputs (an effect that would be removed in free trade, and thus increase exports) and also the effect of structural changes such as import substitution and export growth in sectors such as petroleum and tourism. 2/ Valuing the peso at its estimated equilibrium level. 3/ Approximated by changes in wholesale price indices. - 20 - Text Table 3 Mexico: Exchange Rates, Protection and Export Incentives: 1970, 1975, and December 19/6 (in current prices) 1970 10'75 December 1976 (pesos per dollar) 1. Exchange rate 12.50 12,50 20.21 2. Apparent protection for manufactures (in percent) 18,8% 4Q.9% 25,5% al 3. Apparent export incentives for manufactures (in percent) -12.0% 7.1% -9.3Z e/ Effective exchange rates for: 4. imports of manufactures b/ 14,85 17,61 25.36-al 5. exports of manufactures c/ 11.00 13.39 18.33 e/ 6. Ratio of effective rate for imports to effective rate for exports 1.35 1,32 1.38 7. -Overvaluation (in percent) 2.8% 18.4% 5,4% 8. FTE exchange rate d/ 12.85 14.80 21.31 9. Net protection for manufactures (in percent) 15.6 19.0% 19.0Z at 10. Net export incentives for manufactures (in percent) 14.41 -4.5Z -14.0% e/ a/ Apparent protection for December 1976 was approximated by applying changes in wholesale prices in Mexico, relative to wholesale prices in the United States, to protection in 1975. b/ Effective exchange rates for imports are defined as the average ratio of prices of imports to prices in the domestic market, and thus reflect average realized protection. c/ Effective exchange rates for exports include the estimated effects of various incentives and disincentives, all expressed in terms of how much sales prices would have to change to have the same effects on profits after taxes. See Annex II. d/ The estimates of the FTE exchange rates for 1970 and 1975 are explained in Annex II. For 1976, the estimate assumes that the rate changed exactly in proportion to price changes in Mexico relative to price changes in the United States, from a 1975 base. e/ CEDIs, not in effect in December 1976, are not included. Their restoration in April 1977, plus the movement of the exchange rate to 22.40 per dollar (approximate- lj the same changes as in Mexican prices during January-March 1977), reduce net export disincentiyes to close to zero. Relationships among the items in the table are as follows: line 4 = line 1, increased by percentage shown in line 2 line 5 = line 1, increased by percentage shown in line 3 line 6 = line 4 divided by line 5 line 8 = line 1, increased by percentage shown in line 7 line 9 = percentage by which line 4 exceeds line 8 line )0= percentage by which line 5 exceeds line 8 - 21 - afforded by the licensing system. Sales in the domestic market were still much more profitable than export sales, as the effective exchange rate for manufactured imports remained more than one-third higher than that for manu- factured exports. 50. The restoration of CEDIs, although at a lower level, announced in April 1977 will improve incentives for manufactured exports considerably. It has not been possible to take the new system of CEDIs into account because the detailed regulations were not available when this report was completed. Choices for the Future 51. Mexico's foreign trade policy is now in transition. The value of the peso is at a more realistic level, relative to 1974 and 1975, and perspec- tives are improved. To move toward the goals assumed here (increase confi- dence, manage the balance of payments, induce efficient growth of manufactur- ing output and employment) further steps might prove useful, but these would take some time to put into effect. Therefore the discussion focuses both on alternative policy packages that could be targets for two to three years from now, and also on choices for the transition. 52. Targets: Three aspects of foreign trade policy would be important in achieving the stated goals. First, confidence that the real value of the peso will be kept more or less stable is crucial for sustained export growth. 1/ Potential exporters know that no system of tax rebates or other promotional devices can compensate for continously increasing overvaluation of the currency. 2/ Should the Government choose to establish a new fixed peso value, Mexico would return to the fixed-parity situation that its people are used to. Obviously, a fixed parity can be maintained only when the Mexican economy has achieved price stability relative to its trading partners. If, however, a very rapid stabilization strategy is seen as technically not feasible or if its social or economic costs would make it not worthwhile, it may be better to recognize this at once and make a commitment toward a less rigid anti-inflation program plus an adjustable parity for the peso, at least for a period of a few years, so that an appropriate real value can be main- tained. The worst of both worlds would be to try to fix a new nominal value 1/ Although the exchange rate is obviously a macroeconomic policy instru- ment, it is so important to manufacturing sector development that it is included in this analysis. As with other measures, the purpose is to estimate the effects of alternative policies. 2/ The importance of tourism in Mexico's foreign trade is another reason for avoiding overvaluation of the peso. Tourism exports (i.e., money spent by foreign tourists in Mexico) and tourism imports as well (i.e., money spent by Mexicans when they travel abroad) are sensitive to exchange rate differences, and effective schemes to cheapen the peso for foreign tourists or to limit Mexican tourist expenditures abroad are difficult to implement. - 22 - for the peso but fail to stabilize prices enough to maintain it. Under a floating or crawling peg regime, the authorities may fear a loss of influence on monetary variables (especially interest rates). However, the close geographic and economic ties between Mexico and the United States and related free flows of capital have already limited Mexico's influence on these variables, even with a fixed nominal value of the peso; it is not clear that an adjustable parity would further reduce Mexico's flexibility in this respect to any significant degree. 53. Second, create a system of incentives that make exports profitable. To foster a strong and continuing growth of exports, profits on export sales must be, at least, about equal to those on domestic sales, and Mexican busi- nessmen must be convinced that the government is committed to maintaining the profitability of exports in the future. Sustained growth in manufactured exports, to increase them above the small percentages of output shown in Table 1.3, requires investment in plant and equipment to produce for export markets. Such investment in turn requires the prospects of continued reasonably high profits. As concluded in paragraphs 61-65 below, an exchange rate in the range of 20-22 pesos per dollar in terms of December 1976 prices, with CEDIs generally available at face values equal to the effects of protection on inputs and of sales taxes, which act as disincentives to exports, is probably the minimum necessary to motivate sustained growth in manufactured exports. Such an exchange rate would increase incentives both to import substitution and to exports; this would help manage the balance of payments even with the licensing system greatly liberalized or abolished, and promote more rapid and more efficient growth in industrial output and employment. 54. Third, rely more on the exchange rate itself, and less on additional measures for protection and export promotion. During the 1970s Mexico relied less and less on the exchange rate itself, and more and more on protection against imports and on incentives to export, to maintain realistic effective exchange rates. Two weaknesses of this approach have already been demonstrated: it is not good in keeping up with inflation, and it causes distortion because the protection and the incentives have stronger effects on some sectors than on others. These distortions induce growth in less efficient industries, and permit efficient industries either to gain very high profits or to relax and let costs rise. The possibility of retaliatory actions by trading partners is a third weakness of the approach. The United States, which buys three-fourths of Mexico's manufactured exports, has been increasingly imposing countervail- ing duties against measures it determines to be subsidies on exports. 55. The distorting effect of protection and export incentives is some- times intentional, as policy-makers seek to promote specific industries. Such measures can be desirable when based on well-thought-out national ob- jectives, externalities, learning-by-doing effects, etc. and applied in specific cases after careful analysis leading to a decision at an appropri- ate level in the government. But a system in which distortions are the norm rather than the exception, and are determined by thousands of small decisions taken every year by lower level civil servants, is bound to lead to substantial inefficiency. This is exactly the effect of a generalized licensing system. - 23 - 56. The first concrete steps to implement such a strategy have already been taken: the exchange rate is in the economically appropriate range, CEDIs are to be restored (albeit at a lower average level), and liberalization of import licensing has begun (albeit for only 884 items, many of which were either seldom imported or for which licenses were issued more or less on request; some 4,900 other items remain subject to licensing.) 57. Since the tariff system has long played a very small protective role in Mexico, it would have to be reviewed and adapted to function in the absence of the licensing system. An important part of the overall strategy under discussion is that protection against imports would come primarily from the exchange rate. Tariffs would be kept low, but certain tariffs that are now very low or zero could be raised slightly, especially for infant in- dustries such as capital goods. A structure of tariffs averaging, say, 10 to 15 percent for manufactures, with protection for sectors with much undev- eloped potential at, say, 20 to 25 percent, might be appropriate as a target to be achieved in several years. With the higher exchange rate, average protection as measured by the real peso cost of imports would fall only a little, but the very high protection afforded some products by the licensing system would be ended. 58. Licensing or other restrictions could be retained for imports of products deemed harmful or unnecessary. Even for luxury imports, however, the government might wish to consider whether the nation would not be better off if high excise taxes on luxuries (both imports and domestic products) replaced the present practice of refusal of licenses. This would permit those few people who are willing to pay exhorbitant prices to satisfy their import demands, with most of the price they pay going to the public treasury. 59. The elimination of exemptions from tariffs for public sector agen- cies and for private sector importers of capital goods (implying also that these agencies not be automatically reimbursed by the Treasury for the tariffs paid) would help domestic producers of capital goods compete on less unequal terms with foreign producers. The objectives sought by those tariff exemptions now in effect -- export promotion, decentralization, and increased investment -- can be achieved by other means (such as temporary income tax relief) which would be at least as effective and would not lower the price of equipment and thus tend both to reduce employment (see paragraphs 126-130), and to dis- courage purchases from local producers. 60. The elimination of licensing coupled with tariff reform, including elimination of tariff exemptions, would produce considerable revenue -- an important consideration during a period of budgetary restrictions. In 1975 the loss in tariff revenues due to exemptions alone was about Mex$ 6.5 billion. At estimated 1977 import levels, the changes recommended here would produce additional revenues from tariff collection on private sector imports of between Mex$ 9,000 million and Mex$ 14,000 million--roughly 10 percent of the expected budget deficit. 1/ 1/ This calculation uses tariff rates averaging 20 to 25 percent on manufactured imports; the assumption is that the lower tariffs averaging 10 to 15 percent would be approached over a period of several years. - 24 - 61. The only exemption from paying tariffs that would help to achieve the goals assumed here would be on imports of goods that are re-exported. Giving exporters access to imported inputs at world market prices is an impor- tant part of most of the successful export promotion packages in other coun- tries. There are several different ways in which such access could be achieved; administrative aspects are important in designing the system so that exporters can really get what is promised. One possibility would be to give exporters CEDIs for import duties actually paid. Alternatively, CEDIs given to all exporters of manufactures could include the average cost of import duties (averaged over the manufacturing sector as a whole). This would simplify administration even more, and the small bias it would create against using a lot of imports in exports would not be harmful. 62. In the longer run, revision of the system of indirect taxation to increase its efficiency and give systematic incentives for exports, as is done in most European countries and many others, could be considered. A value added or one stage indirect tax based on the destination principle, so that all exports receive tax rebates as a matter of course, could replace the cascaded sales tax which discourages division of labor through subcon- tracting and specialization. 63. Given the level and structure of protection, and the level of CEDIs, the level of the exchange rate needed to promote both exports and import substitution can be estimated, roughly, as a function of cost levels. Detailed, accurate and up-to-date information on cost structures were not available at the time the mission visited Mexico, but the method can be illustrated with approximate aggregate data. The latest month for which reasonably complete price data are available is December 1976. The exchange rate for that month was 20.21 pesos per dollar; assume it is kept at the same real level through some target year, say, 1980. Further assume that protection and export incentives are changed as suggested above, so that in 1980 protection against imports of manufactures averages 10 percent of the official exchange rate, and that for exports of manufactures CEDIs are available at a face value of 10 percent of exports, which is just about equal to the negative incentives of sales tax on inputs (3.7%), sales tax on output (4%), and net protection on inputs (2%, assuming that licensing is greatly liberalized and tariffs are adjusted as described earlier). 1/ Net export incentives would then be equal to the adjustment for the non-taxability of CEDIs (about 4%) and low-interest export financing (assumed at the 1975 value of 2%), or a total of 6 percent. 64. This set of measures is shown in Text Table 4 and in Chart 1.1, under "1980 Case A." 2/ The real effective rate of exchange for imports of 1/ The percentages in parentheses are again in terms of equivalent before- tax sales revenue, as in Annex II, and can therefore be compared directly with changes in the exchange rate. 2/ In Chart 1.1, the lines connecting the years are only to facilitate reading the chart; in fact the variables did not follow the straight lines shown. However, data to calculate all annual values are not available. - 25 - Text Table 4 Mexico: Alternative Exchange Rates, Protection and Export Incentives (in constant prices of December 1976) December 1980 1970 1975 1976 Case A Case B (December 1976 pesos per December 1976 dollar) Exchange rate 18.99 18.00 20.21 20.21 22.23 Apparent protection for manufactures 18.8% 40.9% 25.57%- 10% 10% (in percent) Apparent export incentives for manufac- tures (in percent) -12.0% 7.1% -9.3%.e/ 6% 6% Effective exchange rates for: a imports of manufactures b/ 22.56 25.36 25.362' 22.23 24.45 exportSof manufactures c/ 16.71 19.28 18.332f 21.42 23.56 Ratio of effective rate for imports to effective rate for exports 1.35 1.32 1.38 1.04 1.04 Overvaluation (in percent) 2.8% 18.4% 5.4% 6.7% 8.0% FTE exchange rate d/ 19.52 21.31 21.31 21.31 24.00 Net protection for manufactures (in 15.6% 19.0% 19.07%i/ 4.3% 1.9% percent) Net export incentives for manufactures -14.4% -9.5% -14.0% et/ 0.5% -1.8% (in percent) a/ Apparent protection for December 1976 was approximated by applying changes in wholesale prices in Mexico, relative to wholesale prices in the United States, to protection in 1975. b/ Effective exchange rates for imports are defined as the average ratio of prices of imports to prices in the domestic market, and thus reflect average realized protection. c/ Effective exchange rates for exports include the estimated effects of various incentives and disincentives, all expressed in terms of how much sales prices would have to change to have the same effects on profits after taxes. See Annex II. d/ The estimates of the FTE exchange rates for 1970 and 1975 are explained in Annex II. For 1976 and 1980 Case A the estimates assume that the nominal FTE ecchange rate would change exactly in proportion to price changes in Mexico relative to price changes in the United States from a 1975 base -- i.e. that the constant-price rate would stay the same. However, the estimate for 1980 Case B assumes faster growth than the 1970-76 period or than Case A. Thus the FTE exchange rate for 1980 Case B is estimated to be sane- what higher than fo_the earlier year.s or for 1980 Case A. The actual FTE exchange rate for 1280 tYould depend on the actual other policies in effect and on the structure of the economy at that time. e/ CEDIs, not in effect in December 1976, are nnt- inltlAaA. g'1eir rcsteration in in April 1977, plus the movement of the exchange rate to 22.40 per dollar ( anproximate- ly the same changes is in Mexican prices during January-March 1977), reduce net export disincentives to close to zero. - 26 - manufactures, including protection, would fall 12 percent from its 1975 level, while that for exports of manufactures would rise 17 percent. Since the projected exchange rate is estimated to be slightly overvalued, net protection for manufactures would be 4.3 percent and net export incentives for manufactur- ing would be 0.5 percent. 65. This set of policies would go a long way toward achieving the goals set. The effective exchange rates for imports and for exports would be close to each other, thus making production for export almost as profitable as pro- duction for domestic markets. Export profitability would be much higher than in 1975; by a rough estimate it would be close to what it was in 1971 or 1972 when CEDIs were first made available, before Mexican price rises offset much of their good effects. The effective exchange rates for imports and for exports are very close to the projected free trade equilibrium exchange rate, which indicates that they are consistent with past growth levels. To expand output and employment even more, however, the package for 1980 might include an exchange rate higher than is assumed in Case A. 66. As an alternative to Case A, assume an official exchange rate 10 percent higher in 1980 -- 22.23 pesos per dollar, in terms of December 1976 prices. (That is, the actual exchange rate in the target year would be 22.23 pesos per dollar, multiplied by a coefficient reflecting price increases in Mexico relative to her principal trading partners, since December 1976.) With apparent protection and export subsidies as they were in Case A, both import substitution and export promotion are more strongly supported than in Case A. (See "Case B" in Text Table 4 and Chart 1.1.) In Case B, the real effective exchange rate for imports of manufactures is only 4 percent below the 1975 level. Protection for domestic production for the-domestic market would thus be maintained on the average, but with fewer instances of very high protection. Manufactured exports, on the other hand, would be highly profit- able -- even more so than they were in the early 1970's. 67. Case B policies would induce faster growth than would those of Case A, but would be slightly more inflationary. Compared to Case A, Case B would imply a slightly higher one-shot cost push stemming from the higher cost of some imports (and those protected domestic substitutes sold under non- competitive conditions), and from demand-pull factors stemming from higher prices on export goods and higher employment generally. Of course, under either Case A or Case B, eliminating the import licensing system would force compensating price reductions for products now enjoying very high protection. In the long run, increased employment and growth in the Mexican economy are highly desirable and their negative effects in terms of inflation can be dealt with by fiscal and monetary policies as well as more specific programs to increase supplies of scarce inputs, whatever they may be. Thus Case B may be better than Case A for the long run. But right now, faced with an unusually great need for an anti-inflationary policy, Case A may be preferable until price increases are more firmly under control. The choice between the two policies may, therefore, be essentially a question of timing. - 27 - 68. Transitional Measures: The measures just outlined in paragraphs 48-63 would represent considerable changes from the environment to which most Mexican manufacturing firms are accustomed. The evidence strongly suggests that most firms could make the transition succesfully: the exports that Mexican manufacturers have already achieved, the moderate to low actual differentials between Mexican and international prices for most Mexican manufactured products, the analysis mentioned in paragraph 37 that concludes that in most cases where protection in Mexico is sheltering high costs, those costs could be lowered if necessary, and the fact that average peso costs of imported manufactures would not fall significantly under the measures described. Nevertheless, care would be required to minimize both economic and psychological problems of a transition. 69. The sequence of policy changes could be planned so as to start with measures that generally increase profits and export possibilities, and then to follow with reductions in protection. This is in fact what the Mexican government has done in the last few months, allowing the exchange rate to move to a more realistic level and restoring CEDIs. 70. Since Mexico has lived with the import licensing system for so long, it probably could not simply be abandoned all at once. The number of products subject to licensing could be reduced over a period of a few years, perhaps starting with intermediate goods and capital goods. Alternatively, or comple- mentarily, temporary higher tariffs could be applied to high-cost produc-s as they are decontrolled, to give producers of each product a reasonable time to reduce costs. The desired results of this program would not be achieved, however, unless any tariff increases are clearly specified as temporary only, and are reduced over a reasonable time specified in advance. 71. Pending complete elimination of import controls, the government might consider granting automatic permission to import any item regardless of its tariff, on which the importer was willing to pay a fixed high tariff (say, 50 percent). This would place a ceiling on the amount of total pro- tection from the combined effects of import permits and tariffs. Also, ex- porters could be automatically granted licenses to import inputs, with sub- sequent periodic checks to insure re-export of the goods. 72. Making credit and technical assistance available for re-equipment, training, or reorganization of firms which would be forced to lower costs and/or improve quality in order to survive would also help to ease the transition. Insuring availability of working capital, and a possible expanded role for government trust funds, as discussed below in paragraphs 116-122, would be of help here. 73. Assistance to displaced workers would be important. Mexico has no unemployment insurance and few other programs to help workers find new jobs. In this situation the authorities are understandably reluctant to permit firms to fail, even if the firms are inefficient, because the displaced workers could be badly hurt even though the economy as a whole would benefit. Provid- ing better assistance to displaced workers is therefore desirable for two reasons: it would reduce political pressure to protect inefficient firms, - 28 - and it would spread the cost of achieving greater efficiency over the society as a whole, where it belongs, rather than concentrating it on workers in the industry involved. Conclusions: Foreign Trade Policies 74. In conclusion, the mission's analysis suggests that: (i) An exchange rate around 20-22 pesos per dollar, in terms of December 1976 prices, plus restored CEDIs and a great liberalization or dismantling of the import licensing system, would promote rapid growth of employment and effi- cient production in Mexican manufacturing. This set of policies would promote both efficient exports and efficient import substitution, with not only growth in output and employment but also improvements in quality of Mexican goods. (ii) Promoting exports and import substitution mainly through the exchange rate, instead of through special protection or promotional schemes, will call forth growth in sectors where Mexico is most efficient--including tourism and efficient agriculture and mining, as well as manufacturing. This will increase use of Mexico's abundant labor force, and reduce or remove instances of very high cost or low quality import substitution. Such a strategy will make export profits closer to profits on domestic sales of a given product, and induce more specialization and scale economies, rather than the more diversified and less efficient production designed only for a captive domestic market now common in Mexican manufacturing. Mexico's manufacturing sector is now mature and developed enough to benefit from such a system, without suffering large-scale replacement of domestic production by imports. (iii) Promoting growth through stable rules, known in advance and applied automatically, instead of relying on many daily decisions about granting import permits, CEDIs, etc., will permit entrepreneurs to estimate their costs and profits with more assurance. Investment, for both exports and domestic sales, will be on more economical scales of production and more concentrated in sectors where it is most profitable. - 29 - II. Public and Foreign-Owned Enterprises 75. As in most other less developed countries, public, domestic private, and foreign-owned enterprises coexist in Mexico. Similarly to the situation in other countries they are each concentrated in different industries, they are ruled by different laws and they show different economic performance. Public Sector Enterprises 76. Magnitude: The public sector in industry increased rapidly during the last five years or so. Public investment in manufacturing is estimated to have been 57 percent of total investment in the sector during 1971-75. The impact of public investment in industry was strikingly unequally distributed over time; in 1971 the public sector was responsible for only 20 percent of total industrial investment; during 1974 and 1975 the percentage had increased to as much as 80 percent while private investment in industry had declined to an insignificant level. 77. As of 1976 there were some 400 enterprises in which the public sector held a majority interest in the economy as a whole, of which only 15 account for about 90 percent of total investment. In the manufacturing sector, of 34 large public sector enterprises only five 1/ have fixed assets above $100 million; these five account, together, for 88 percent of the total fixed assets of the group of 34. These enterprises increased their assets by $1.7 billion from 1973 to 1975 -- an increase of 140 percent. To finance this expansion they increased their indebtedness by $1.4 billion, which implies that 80 percent of the increase in their assets was financed by loans. Capital for the substantial sample of enterprises was distributed principally among five industries: sugar (5 percent), automobiles (22 percent), machinery (2 percent), chemicals (13 percent), and steel (58 percent). However, the public sector is also involved in smaller enterprises throughout the manufac- turing sector, from bakeries to bicycles. 78. Motivation: The Mexican Government invested substantial resources into a few key sectors of the economy mainly to control prices and/or to assure domestic development of the industry. Both reasons were important in steel and petrochemicals, where investments are large, lumpy and have long gestation periods; the price element is very important because the prices of these products play important roles in determining the price level of many other industrial goods in the country. The government entry into the auto- mobile industry is somewhat more difficult to explain, the motivation appar- ently being a desire to put up a national counterpart to foreign investment so as to have more leverage in pressing for decreasing the import content of automobile production, as well as to see this production located in an area which the Government wanted specifically to develop. In sugar it has been, in turn, the Government's desire to resolve a long standing conflict between 1/ Sicartsa (steel), Guanomex (fertilizer), Altos Hornos (steel), Diesel Nlacional (motor vehicles), and Constructora Nacional (railroad equip- ment). - 30 - the refiners and the cane growers in favor of the latter -- with the result- ing income distribution effect -- and its simultaneous resolve to maintain a low consumer price. 79. Performance: Because a large share of the Government's involve- ment in manufacturing is of very recent vintage, only a part has become fully productive and it is still too early for a full review of its efficiency. However, when important enterprises which did not go on stream before late 1976 (e.g., SICARTSA) are excluded, one could reasonably expect to see the investment performance of the remaining enterprises improve over time. This does not seem to have been the case. The mission has collected and analyzed the balance sheets and profit and loss accounts of over 30 industrial enter- prises with either exclusive or predominant public participation. This sam- ple was broken down into five large groups: sugar industries, which included 17 enterprises, steel with 3 enterprises, automobiles with 3 enterprises, machinery and appliances with 6 enterprises, and chemicals consisting of 2 public enterprises. W1hile the assets of these enterprises have doubled over the last three years, their cumulative losses increased by 70 percent. Their indebtedness increased sharply, with their debt-equity ratio growing from 1.9 in 1973 to 3.2 in 1975. Losses on production were incurred in practically all industries except for steel, where Altos Hornos retained an earning capacity, and for chemical industries, where Guanomex remained profitable. By 1975 the ratio of losses to net worth became 4.3 percent, an increase from 3.3 percent in 1973 and 2.6 percent in 1974. (See Table 1.5 for more details.) 80. Part of the cause of this poor performance lay in the division of authority over and lack of control of these enterprises. Responsible in theory to the Secretariat of National Property, with investment plans cleared by the Secretariat of the Presidency, and with the Secretariat of Finance paying the bills for whatever deficits appeared, the enterprises did not have strong incentives to hold down costs or to keep prices in step with whatever cost increases did occur. 81. The current government has shown its recognition of these problems, in enacting two laws which may help to remedy the situation (Ley de la Admin- istracion Publica and Ley de la Deuda Publica). Responsibility for the plan- ning of public investment and other expenditure programs, the budgeting of current expenditures, and the planning and control over state enterprises' expenditure has all been placed in the new Secretariat of Programming and Budgeting. After approval by that Secretariat, state enterprises must submit to the Secretariat of Finance their plans for expenditures, a schedule of the sources of funds to cover the expenditures, and where borrowing is planned they must specify sources of funds to cover future debt service. Moreover, the approval of the Secretariat of Finance is required before even preliminary exploration of new foreign borrowing may take place. These new statutes improve the government's ability to control capital expenditures and indebted- ness of state enterprises, and it is to be hoped that they will prove useful in this regard. - 31 - 82. Pricing: The pricing system is an important element in determining the financial performance of public enterprises. Comparison between domes- tic and international prices is a reasonably good indicator of efficiency of domestic production. Such comparison should, however, also take into account whether, at the present level of domestic prices, the enterprises are gener- ating any profit, and whether the exchange rate is close to an equilibrium level. 83. In steel, price ceilings have been imposed since 1956. The private sector has complained frequently that these ceilings were too low and that they were limniting its capacity to generate internal resources. However, there is little evidence that price ceilings, at least those prevailing in the 1960s, were any lower in Mexico that prices prevailing within the main steel-producing countries. At 12.5 pesos per dollar, price comparisons be- tween Mexico and other countries show that Mexican prices were roughly at the level of international prices at that time. This should be viewed against the backdrop of the public steel industry still generating some net profits -- the return on net worth in operating enterprises in the steel industry in 1975 was 4 percent, compared with 6 percent in 1973. With a rate of 20 pesos per dollar and a less-than-proportionate domestic price increase, Mexican steel prices became about 20 percent lower than inter- national prices. 84. Sugar prices had been kept unchanged since 1970, while the price paid to cane growers increased by 190 percent. Thus by mid-1976 the cost of cane to the sugar refiners became, at 2.15 pesos per kilogram, equal to the price of refined sugar charged by commercial enterprises for distribution to consumers. As a result, private owners of refineries could not operate their factories profitably; the government has taken over some factories and decided to set up a few new ones, to be operated by government corporations, so as to increase production capacity by about 460,000 tons of sugar a year. At 2.15 pesos per kilogram (about 8 U.S. cents per pound before August 1976), sugar was considerably cheaper in Mexico than in international markets, even at presently depressed prices in the latter. Low prices to the consumer were partly compensated by higher prices (about 13 U.S. cents per pound) to indus- trial users, but the government had to carry the brunt of the subsidization. Consumer prices for refined sugar have been increased to 6 pesos per kilo (14 U.S. cents per pound) in the last week of December 1976; this price is prac- tically at par with the price paid to industrial users, and compares favorably with current international prices of about 8 to 10 U.S. cents per pound. However, the price of brown sugar is still set at the 2.15 peso level, which has resulted in its virtual disappearance from the market. 85. In fertilizers the government granted GUANOMEX, a wholly owned government corporation since 1965, sole responsibility for marketing fertil- izers in order to avoid excessive price fluctuation and prevent the sale of fertilizer to final users at what has been considered to be a high price, mainly attributed to unreasonably high profits of intermediaries. Comparison with international prices shows that at least in the case of some fertilizer products there might have been a strong subsidy. Thus in 1975 even at the - 32 - exchange rate of 12.5 pesos per dollar, Mexican prices were in most cases below the U.S. prices. When the peso began to float, even after some domestic price adjustment Mexican prices were 11 percent below U.S. export prices. 86. In the automotive sector, most firms are private. Government sets price ceilings which have been roughly 50 percent above international levels, at the past exchange rate of 12.50 pesos per dollar. These price levels seem to be sufficient for most companies to earn satisfactory returns, but the public sector firm has been suffering losses. 87. Text Table 5 summarizes price policy in public sector enterprises in the important sectors just discussed. Text Table 5 Mexico: Prices and Profits in Public Sector Manufacturing Enterprises Sector Prices Profits Steel (excluding Equal to or Low but Sicartsa) below imports positive Sugar Low until Negative December 1976 Fertilizers Low Positive; diminishing Automobiles Above imports Negative 88. The government desire to hold down prices in steel, sugar, some chemicals, and fertilizers may by itself have been one of the major causes of unwillingness of the private sector to invest in these fields. These low prices are maintained in order to transfer resources to private users of the products. However, it is not clear how much such measures increase social justice--the subsidized fertilizer is used mostly on large highly profitable irrigated farms, for example. The policy seems even more questionable when its results in terms of higher public sector deficits, higher inflation, and erosion of purchasing power are also taken into account. 89. Pricing policy in public sector enterprises involves more than just raising prices to cover costs. Such action helps to reduce the overall public sector deficit and to bring relative prices more in line with produc- tion costs of different products. But it does little to discipline the enterprises to control their costs. Cost control, including economic cal- culations on new investments and their returns, would be improved if public - 33 - sector enterprises were required to cover a predetermined portion of their operating costs and investment requirements from operating revenues. Such a requirement cannot be applied 100 percent to all public enterprisess--for example, it would reduce the effectiveness of Conasupo, whose objective is to supply basic foods at low prices. The requirement could, however, be applied to most or all public sector manufacturing enterprises, with the percent of costs that must be covered set below 100 percent where appropriate. A requirement to cover even part of expenses would increase incentives for efficiency. 90. Purchases of equipment: Another important aspect of the role of public enterprises in industry is their reluctance to use domestically produced capital equipment. Until now the public sector in Mexico has been supplying its large and growing requirements for capital goods mainly from imports. The strong proimport orientation of the public sector is due to three factors: 91. First, public sector enterprises, while planning for future expan- sions and related capital requirements, were unable to predict the resources available to them from the Treasury, and were similarly unsure of Treasury approval for future foreign borrowing. Consequently, the public enterprises could. not negotiate long-term contracts with potential domestic suppliers of capital goods. Such contracts could have helped private domestic enterprises to prepare designs and technology well ahead of time and, after those were approved, to supply more equipment in required quality and within reasonable costs. The absence of such arrangements contributed to a situation when, each time the public enterprises realized that their financing capacity -- largely depending on Government decisions -- permitted them to purchase capital equip- ment, foreign suppliers proved to be more flexible in providing appropriate terms of delivery and quality of required goods because of their accumulated experience, while domestic suppliers were handicapped in this respect. 92. This situation changed somewhat recently with the creation of Com- isiones Mixtas de Abastecimientos, supposed to enquire not only into what the domestic industry currently produces but also into what it could produce. The positive result was best shown in purchases of goods from domestic indus- try by the Comision Federal de Electricidad. The CFE was purchasing as much as 63 percent of its supplies from foreign sources in 1973; by 1976 imports fell to only 39 percent. However, the CFE's long-term plan for 1978-1990 still foresees that out of the projected $8 billion purchases of plant and equipment, as much as 75 percent will be purchased abroad -- unless the domestic industry sets up new capacities or organizes its existing capacity so as to be able to supply the required equipment. 93. The second reason for the preference given by public enterprises to imported capital goods lies in high prices of domestically produced capital goods. The price differential can be quite large in some cases; domestic goods costing 80 percent more than imports (at the pre-devaluation exchange rate) are reported. To a certain extent, this situation is of a - 34 - classical vicious circle type, because the relatively high price level is in part attributable to the still relatively small scale of production of the private sector enterprises. Had they been given a full possibility to sell to the public sector, this disadvantage would no doubt diminish. Thus, in 1975 total Mexican imports of capital goods amounted to $2.4 billion of which over $1.2 billion were purchased by the public sector. In the same year, the gross value of production of domestic capital goods producing industries amounted to some $1.1 billion. There is little doubt that the latter could have produced some of the imported goods and that such addition to their production would have permitted them to reduce costs and prices. Another dimension of the same problem is credit terms, which are far more generous in the case of imported equipment than for that produced in Mexico. The over- valued exchange rate also made imports relatively cheaper. 94. Third, public enterprises are entitled to import without having to pay any customs duties. They are supposed to allow domestic producers a margin of 15 percent over import prices, but since the 15 percent is only an element in a calculation and never has to be paid in the case of imports, it provides less motivation for purchasing from domestic producers than would an actual tariff of the same amount. 95. To what extent the introduction of the new exchange rate, even if coupled with elimination of tariff exemptions and/or moderately higher tariff rates on capital goods, could equalize the chances of domestic manufacturing in bidding against imports for public sector orders is still unknown. It is also somewhat puzzling that the accords, negotiated late in 1976 between the private sector and the then Director General of IEPES do not include commit- ments as to the approximate extent of public purchases from private manufac- turers. The accords involve some $8 billion in private investment by 170 firms in 10 sectors over the next six years. Each agreement includes prom- ises from the private sector (usually investment and/or production goals), a quid pro quo from the public sector (usually unspecified promotion and/or incentive measures) and statements of joint action (usually joint groups or committees to study promotion problems and needs). These accords are im- portant instruments to restore confidence in a fruitful cooperation between the public sector and private industry. They should be considered, however, only as a first step, and their effectiveness may gradually wane if they are not followed by a change of industrial policies including specific inten- tions and, later, contracts of public sector enterprises for purchasing goods which private industry would be able to supply. 96. Conclusions: Three basic conclusions can be drawn from the analysis of the situation in the public sector: (i) Large amounts of private capital actually invested in steel, automobiles, sugar, and chemicals suggest that if the private sector had adequate access to credit in pesos (see paras. 113- 116), freedom to set prices at a reasonably remunerative level, and perceived a more favorable investment climate, then private - 35 - entrepreneurs might have continued to put their resources in the sectors where public investment has been large, alongside or even instead of public resources. The Government now appears more disposed to consider possibilities of either joint ventures with the private sector or, in cases where the private sector proves to be willing to invest enough of its capital, even to shift allocation of its resources to other sectors of the economy. In the latter case, public goals such as pricing, location, or domestic integration could be equally well achieved through the use of appropriate economic policies, some of which are discussed in this report, as through direct decisions of managers of public enterprises. (ii) If prices of goods produced by public sector enterprises covered production costs, returns on capital invested in the public sector could become sufficient to pay for a substantial share of new investment from internally generated resources. Indeed, imposing a requirement for at least partial self-financing of investment could help to induce a more appropriate pricing policy, as well as draw more attention to a need for cost reductions. To the extent to which, through a more flexible price policy and more exacting performance criteria the public enterprises would be able to finance a larger part of their expansion, the danger that capital allocation from the budget may not be forthcoming during years of budgetary squeeze would be- come less than it has been hitherto. (iii) The public sector could promote domestic production of equipment, at little cost to itself or the nation, if long- term planning of capital expansion became an explicit duty of public enterprises. The projections of physical expansion and resulting needs for equipment are often available, but the financing of the expansion and hence the actual purchase of equipment are left open. To make such plans more specific, more realistic, and thus more useful, the proportions of their own financing and of budgetary allocations should be set forth well in advance. In turn, if their expansion plans were made public, with detailed specifications of equipment requirements, the domestic equipment suppliers could prepare longer-term production plans and probably could supply a greater share of publicly purchased equipment, within acceptable quality norms and price, than they have been able to do in the past. Even where foreign financing and international tendering is involved, advance information of the specifics of the future demand may somewhat equalize the chances of domestic firms in competing with other, more experienced bidders. - 36 - Foreign-Owned Enterprises 97. Magnitude: Foreign-owned enterprises account for a substantial share of investment in Mexico. During the years 1972-75, total foreign investment represented 5 percent of total economy-wide private investment in Mexico. Seventy-two percent of total foreign investment came from the United States, 6 percent from the Federal Republic of Germany, and 5 percent from the United Kingdom, with various countries sharing the remainder. Some 4,100 enterprises in Mexico have some foreign participation, and in almost two-thirds of these enterprises the foreign participation exceeds 50 percent. 98. About one-half of these enterprises and three-fourths of the total amount of foreign capital were located in the manufacturing sector. This share has changed but little over time, growing from 74 percent of the total in 1970 to 75 percent in the total in 1975. Of the total of some $1.2 bil- lion invested in enterprises with private foreign participation during 1972-75, $930 million went into manufacturing. This amount is, roughly, about 25 percent of total new investment in manufacturing. 1/ The share of public investment is estimated at 57 percent, so the residual of 18 percent represents investment undertaken by private capital without public or foreign affiliations. Thus foreign private investment is quite important just for the amount of investment it represents, apart from other aspects. (See Table 1.6 for more details.) 99. The importance of foreign private investment can be partly ex- plained by its greater access to financing (even in cases of only partly foreign-owned firms), from parent companies or other linked foreign sources. Foreign liabilities of Mexican enterprises in which foreign'participation exceeded 5 percent (manufacturing enterprises represented 67 percent of this statistical universe) grew by almost $600 million in one year: from $3.2 bil- lion in 1974 to $3.8 billion in 1975. Of this amount 17 percent was obtained directly from the parent companies abroad, and a substantial part of the bal- ance came from financing sources related to these parent companies. 100. Thirty-five percent of the foreign financed investment in manufactur- ing went to the machinery and transport equipment industry, compared to only 13 percent of the total channeled into this sector by all investors combined. The chemical industry was another large beneficiary, accounting for 26 per- cent of the total private foreign industrial investment, compared to only 17 percent of the total for all investors combined. Thus foreign investment is particularly important in sectors with complicated technology. 101. Export performance: It has been a long-standing belief that for- eign investment in developing countries is bound by restrictive clauses, im- posed by parent companies, to restrict exports so as to avoid infringing 1/ Based on approximate estimates by the mission of total investment in manufacturing. - 37 - upon the parent companies' markets in other countries. The extent to which this practice was followed in Mexico is unknown, but since 1973 the practice has been forbidden by Mexican law. The automobile industry, in which foreign investment participated substantially, 1/ has proven to be one of the most dynamic exporters of manufactured products, with a total value of exports of about $150 million in 1974 (see paragraph 27). 102. An ECLA study, using data from an IMCE sample survey, reports that in 1975 multinational corporations accounted for 55 percent of Mexico's manufactured exports (ECLA, 1976, page 80). Industries where these foreign enterprises were reported to be most important were petroleum products, transport equipment, chemicals, machinery, rubber, electrical equipment, and the miscellaneous category, in all of which multinationals accounted for more than three-quarters of all exports. The mission's estimate of the share of manufactured exports coming from firms with foreign participation is 37 percent, rather lower than the ECLA estimate. Nevertheless, both estimates suggest that foreign investment is important for manufactured exports. 103. A review of foreign private investment in Mexico carried out by Stanford Research Institute in late 1975 shows that gross sales of manufac- turing enterprises with foreign participation amounted to $4.2 billion in 1974, compared to $26 billion of total industrial sales in this year or to about 16 percent of total sales of industrial goods. In the same year these foreign-owned enterprises exported industrial goods worth some $250 million, equal to 6 percent of their output, while the manufacturing sector as a whole exported only 5 percent of its output (both figures exclude assembly plants). The share of particularly difficult exports like those of the engineering industries is much higher for enterprises with foreign participation than for the others. Thus, for instance, as much as 63 percent of total exports of electrical machinery originated in enterprises with foreign participation. 104. Foreign private investment can facilitate cooperation with multi- national enterprises in other countries. The best example here would be Brazil, which is included in ALALC and therefore has easy access to the Mex- ican market. Exports of Brazil to Mexico have been increasing at 46.5 per- cent per annum, during 1972-75, compared with the overall increase in Mexican imports of 34 percent per annum (all in current terms). The total stock of foreign direct investment in Brazil ($7 billion in 1972, of which 70 percent is in manufacturing) is much larger than that in Mexico ($3.6 billion, of which 77 percent was in manufacturing--also for 1972), and as much as 41 percent of total imports by Mexico from Argentina and 35 percent of total imports from Brazil consisted of goods produced by foreign-owned enterprises (Fajnzylber and Martinez, 1976). This trade should be viewed as an important part of a continent-wide specialization of ALALC countries, in which foreign enterprises play a substantial role -- as they did at the onset of the Euro- pean Common Market specialization. 1/ Out of 14 automotive plants, 5, including tractor plants, had a 100 percent foreign ownership, 2 firms had a substantial foreign participa- tion, 2 firms had a 100 percent government ownership, and the remain- ing 5 firms, mainly producing tractors, trailers and buses, had a 100 percent private MIexican capital. - 38 - 105. Regulation: New foreign investment legislation introduced in Mexico in early 1973 consists of the Law on Promotion of Mexican Investment and Regulation of Foreign Investment, and the Law on Technological Transfer and Use and Exploitation of Trademarks and Patent Rights. The former introduces the general rule of 49 percent foreign capital ownership of manufacturing enterprises as a maximum. This is considered as the most important regulation in this field. It appeared to have created some uncertainty among foreign investors who had become used to the idea that companies could be set up in Mexico with 100 percent foreign capital and still be considered national companies. Prior to this law, although restrictions had been imposed on certain fields of activity for companies with foreign capital majority, joint ventures were in the minority. Even stricter limits than this now generalized rule of 49 percent foreign capital exist in certain fields such as mining (34 percent) and manufacture of automotive parts (40 percent). However, the National Commission of Foreign Investment established by this law is empowered to authorize increases or decreases of foreign capital par- ticipation, and on occasion has authorized creation of companies with 100 percent foreign capital as exceptions, without stating any general criteria. 106. The law is not retroactive and therefore does not change the legal situation of the foreign investment which existed before it went into effect, although it applies to their expansion, which tends to create some compli- cations. In the period since the law has entered into operation, until April 1976, about 345 enterprises with foreign participation were established in Mexico and the inflow of foreign private investment in manufacturing apparently has not decreased. Of these 345 enterprises, foreign participation in 338 has not exceeded 49 percent. Most of these enterprises were small; only 28 of the 345 had capital exceeding $800,000. 107. In the technological transfers and trademarks and patent law, the royalty section presents the greatest interest, because it introduces a clas- sification which should protect users of foreign technology from being charged more than competitors in the same sector. Also, payment for trade- marks is limited to 1 percent on net sales. Another important prescription is that all foreign trademarks should be used together with their domestic acronyms. 1/ The licensee acquires rights to improve on or develop his own technology in addition to the acquired one. Finally, almost all restrictions on exports of manufactured products which incorporate foreign technology were ruled out. 108. To seek benefits from the existing system of fiscal incentives in Mlexico, the foreign investor must not only conform to the 49 percent partici- pation rule but as another condition, his entire transfer payment for foreign technological assistance and use of trademarks should not exceed 3 percent of net annual sales. 1/ The importance of foreign trademarks is considerable; of about 6,000 foreign contracts signed within the last few years, 2,900 included the use of trademarks. - 39 - 109. Conclusions: The mission's analysis of foreign investment suggests five general conclusions: (i) Regulation. There is little doubt that a developing country could only benefit from foreign investment legislation that is designed, on the one hand, to help the country to derive from this investment the greatest possible advantage for its national economy and, on the other, attract as much as possible, serious foreign investors. However, undue com- plexity of the law or limitation of foreign entrepreneurs' freedom to decide on their product mix or expansion plans quickly and on their own, has sometimes led, in other countries, to a slowdown in the inflow of foreign investment. It has also been proven elsewhere that instead of very specific measures, across-the-board industrial policies which apply to the entire manufacturing sector without making a distinction between public, domestic private and foreign private enterprises, can be so designed as to extract the largest benefit possible for the economy from each of these groups. (ii) Industrial strategy exceptions. Exceptions from the 49-51 participation rule are not made clear in the law and were left to the discretion of officials of the Secretariat of Industry and Commerce. Large established firms may know how to deal with the bureaucracy, and large new potential investors can afford to hire advisors who know this. But the need to negotiate and the uncertainty of the outcome may constitute barriers to entry, especially for smaller firms. Therefore, while flexibility can be useful, the lack of clear and explicit rules of the game may deter some new investors. Clear stipulations as to specific areas of national interest where exemptions could be made to permit majority participation of foreign capital might be useful (e.g., enterprises located in undeveloped regions; export- oriented industries where foreign investors may have better knowledge of or access to foreign markets; producers of complex products, such as engineering industries which appear particularly attractive to foreign investors and in which they enjoy the advantage of the research and development results of their parent companies). (iii) Financing. In a situation where credit is scarce it may prove difficult for domestic entrepreneurs to finance the 51 percent or more of capital in enterprises which foreign investors want to set up or to expand. Moreover, foreign investors seem to be willing to invest in industries that domestic investors are sometimes not very interested in. (Capital goods is an outstanding example in Mexico.) - 40 - For both these reasons if the 49-51 percent participation rule remains as the strict letter of the law, some valuable investments may not materialize for the lack of domestic counterpart investors. It may be advisable to consider some specific provision in the law, designed to avoid losing oppor- tunities in this way. (iv) Quality. The law of trademarks may impede those foreign investors who identify their own trademarks with a high quality of product and who would like to remain responsi- ble for the quality of the product. One way to avoid this problem could be a clause in the contract permitting the foreign investor to forbid use of his trademark some time after he withdraws from the enterprise, if he can prove that the quality of production is deteriorating. Also, the limitation of transfers of royalties and making fiscal incentives conditional upon their amount may deprive Mexico of the benefits of some technology that might be worth considerably more than its cost in royalties. These provi- sions should be reviewed with special care in cases of industries where exports are potentially important, or where research and development is particularly costly. (v) Joint ventures. The last and probably the most important point is that of the tripartite agreements between foreign investors, private domestic investors and public investors to set up joint ventures, in which the Mexican authorities have shown considerable interest. This type of agreement has been tested in some other semi-industrialized countries and has proven to provide substantial benefits to the coun- try. Foreign investors contribute their managerial experi- ence and technological achievements, domestic private inves- tors their knowledge of the country's internal market, of its manpower, and of the financial and legal complexities, while the public participation was usually considered a guarantor of the country's interests in the project, and also helps to navigate the country's regulations and admin- istrative procedures. Experience of other countries has shown that the presence of a strong negotiator, such as one of the local development banks considered as being both competent and impartial by all interested parties, is often important to the success of such ventures. - 41 - III. Industrial Financing 110. The inflation and lack of confidence of the last few years have had drastic effects on Mexico's well-developed financial sector, and on the credit situation facing Mexican industry. Savings have been directed away from Mexican financial intermediaries and into foreign assets and cash. The issues for policy are how to restore the flows of domestic savings, maintain access to foreign credits, and provide an institutional framework in which these resources will be allocated to productive uses. Ensuring adequate domestic supplies of credit may be even more important than it was earlier because Mexican business firms may now be more reluctant to borrow abroad than they were before the recent devaluations. The main direct instruments available are the exchange rate, the yields on both financial and real assets, and the government's fiscal management (or more precisely, reduction of the public sector deficit). The Financial Situation 111. The Mexican economy has a relatively well developed financial sys- tem. This development has taken place almost entirely through the growth of the banking system, consisting of public, private, and mixed enterprises in both commercial banking and development banking. Direct financing of firms through the public issue of stocks and bonds is much less important. In the years leading up to 1973, the resources available to the banking system grew substantially faster than GDP and allowed credit to the private sector to grow about twice as fast as GDP. 112. Non-monetary deposits made up more than half of the financial system's liabilities, and amounted to 27 percent of GDP in 1973. These shares are typical of developed countries and of developing countries with relatively well-developed financial systems (see Table 1.7). This achievement was made possible by economic stability, a positive interest rate differen- tial with respect to foreign interest rates (spreads of 2 to 5 percentage points during the last ten years) accompanied by full convertibility of the peso, and until 1972 a positive real interest rate in the domestic market (see Table 1.8). However, a large portion of those non-monetary deposits were held in very short-term instruments, the longest being a one-year promissory note; about one third of the value of the deposits was completely liquid. 113. The liquidity of the liabilities of the banking system, which was of course one reason why they were attractive to savers, also made the system highly vulnerable. In 1973 and 1974 non-monetary deposits fell in real terms, as real interest rates turned negative and the interest rate differential between foreign and domestic rates was reversed. Rising inflation relative to Mexico's trading partners induced devaluationary expectations, thereby further reducing the incentive to invest in peso-dominated financial - 42 - assets. This trend was reversed in 1975, due to the decline of inflationary and devaluationary expectations and the fall in interest rates prevailing in international markets. However, resumption of speculation against the peso early in 1976 coupled with the effects of the subsequent devaluations are estimated to have caused non-monetary peso deposits to drop by almost 30 percent in real terms in 1976. Increases in the money supply and the increase in externally financed credit (associated with the rapid increase in public sector external borrowing) acted as somewhat compensatory elements, but total credit still fell by an estimated 10 percent for the year. As a consequence of the drop in intermediation, the real stock of domestically financed credit to the public and private sectors combined is estimated to have been 12 percent less at the end of 1976 than at the end of 1972, while GDP grew by an estimated 23 percent over the same period. (See Table 1.9, where these trends are shown in relation to GDP.) 114. Availability of domestic credit to the private sector was even more squeezed because the public sector preempted part of the already scarce re- sources to help finance its rapidly increasing deficit, which is estimated to have gone up from 4.5 percent of GDP in 1972 to 9 percent of GDP in 1976. As a consequence, the real stock of domestically financed credit to the pri- vate sector was 23 percent less at the end of 1976 than at the end of 1972 and the private sector's share in total credit dropped from 50 percent in 1970 to 35 percent in 1976. (See last two columns of Table 1.9.) 115. Thus credit to the private sector has been doubly restricted, first from the fall in mobilization of savings and second by the increasing share of available credit being preempted by the public sector. Within the private sector, special credit programs helped maintain the share of agri- culture and livestock in the total, while industry's share has been eroded. (See Chart 1.2.) 116. The direct effects of the credit squeeze have been felt most on working capital; financing for new investments or expansions of production capacity has been more readily available. However, the squeeze has affected new capacity creation indirectly, as additional capacity also implies addi- tional working capital requirements. Some firms have been reluctant to invest in new capacity, even if financing was available, because of present and/or anticipated future problems in financing working capital needs. 117. The negative effects of this credit squeeze on private industry's cash situation have been aggravated by recent cost increases. Some of these increases are due to the devaluation (interest on foreign borrowing, replace- ment and expansion costs of imported equipment, some current inputs) while others are not (electricity). Thus, during recent months many private busi- ness firms are facing a severe cash squeeze because their costs have risen much faster than their sales prices. These problems beset efficient firms as well as inefficient ones. At the time of the mission's visit to Mexico in October-November 1976, the Banco de Mexico was aware of this situation and was taking steps to assure that some credit would be available to help - 43 - firms through this period. Moreover, information obtained from the private sector suggested that only a relatively small percentage of firms were in immediate danger of bankruptcy. Nevertheless, the situation clearly deserves continued attention and prompt action where necessary. Not only bankruptcies but also lay-offs and reductions of scale of operations for lack of working capital represent a wastage of Mexico's resources. While restricting credit may be a necessary part of a general anti-inflation program, care should be taken to make it felt in restriction of non-essential consumption and invest- ment, while providing access to working capital for ongoing enterprises. 118. Financial institutions themselves have been seriously affected, as increases in their own administrative costs and interest payments have ex- ceeded growth in their income. Financial institutions have not been able to pass on all these costs to borrowers through increased interest rates, in part because they were facing competition from dollar-denominated loans from foreign banks (at least through 1975), and partly because increased reserve requirements have reduced the portion of their resources which the banks could lend. Consequently the financial sector's rate of return on equity, and profits expressed as a percent of average assets, have both declined -- even though the gross spread and debt-to-equity ratios have increased. (See Table 1.10.) Finance: Conclusions 119. The recovery of financial intermediation in Mexico, perhaps already begun in the first half of 1977, requires confidence above all. Thus it can be achieved only in the context of a serious and comprehensive stabili- zation effort aimed at a gradual reduction in inflation, a gradual and sus- tained increase in public sector savings, and a reduction in the balance of payments deficit. 120. One key to achieving these goals is avoidance of another cycle of overvaluation of the peso. This has already been dealt with at length in relation to trade in goods and services; its importance for mobilizing financ- ial resources is obvious. Confidence in a stable nominal value of the peso is not absolutely necessary, although of course it is very desirable. What is necessary is confidence that a large devaluation will not occur again, combined with interest rates high enough to compensate for any expected small devaluations. 121. Another crucial ingredient to assist recovery of financial savings and to make them available to Mexican intermediaries would be the provision of a number of attractive alternatives for financial investments by both Mexican and foreign private savers. The essential characteristics of the strategy would be to provide a full range of different types of instruments so as to attract the largest possible number of savers, and to permit interest rates to go high enough to attract the amount of resources needed to finance both the government's and the private sector's needs. - 44 - 122, An example of the new type of obligations just mentioned are the "Petrobonos" (bonds indexed to the price of petroleum) recently issued by the government. A further possibility during the period of price stabiliza- tion in Mexico would be to permit banks to issue peso-denominated obligations indexed to the exchange rate. Issuing such obligations with maturities of one year or more would help to reduce the vulnerability of the Mexican financ- ial system by lengthening its term structure. If these obligations were easily transferrable, liquidity could be provided by a secondary market, which would also be a useful barometer of credit conditions. The possibility of issuing government notes similar to US Treasury bills could also be explored. 123. The higher real interest rates paid to depositors would have to be passed along to borrowers. In the present situation, this would tend to in- crease the role of the price of credit in its allocation. Also, because more savings would become available, borrowers would be less dependent on contacts or inter-company relations with bankers. Moreover, some firms would turn again to foreign sources of credit, thereby reducing some of the demand for peso financing. A strategy of generally higher. interest rates does not, of course, preclude promotion of specific sectors or types of industry (for example, capital goods producers, or small-scale enterprises) through special low-interest loans. (See paragraph 125.) 124. The combination of less demand from the public sector, recovery of resource mobilization by the banking system and managing demand for credit by higher interest rates rather than by lending only to favored clients, would benefit smaller firms even more than larger ones. Availability of credit is even more important than is the interest rate, and when lenders face excess demand for credit the smaller firms tend to be the ones that are unable to obtain loans. 125. There is an important role for special purpose trust funds ("fideicomisos") in reducing the credit squeeze. Given their small size in the total credit picture, their primary contribution cannot be the amount of credit channeled through them; rather, they could help in allocation and they can call the banking system's attention to new sectors and new ways of doing business. They can help channel credit to where it is most hard to get (e.g. smaller enterprises), or to sectors that the government wants to promote. They can help soften terms by lengthening maturities. They can allocate credit to efficient firms, by lending to exporters (as FOMEX and FONEI already do) and by lending on the basis of project appraisal rather than just on collateral (as FONEI already does). They can, perhaps, demonstrate (in the case of FOGAIN) that lending to small business can be both useful to the borrowers and profitable to the lenders if higher interest rates yield an attractive spread for taking the higher risks. In sum, the special purpose funds can help to provide funds in areas or in ways where private financial intermediaries' actions are not sufficient to achieve public economic objectives. - 45 - IV. Technological and Sectoral Issues Employment and Wage Policy 126. Employment issues, and the more basic issues relating to income distribution, can best be analyzed in an economy-wide study. In this report the analysis is limited to manufacturing and hence is unavoidably partial. 127. The most important employment-promotion measure that is dealt with in this report is maintaining an appropriately high real exchange rate. Increasing the price of foreign exchange, relative to domestic costs and prices, creates incentives in favor of both import substitution and exports in all sectors of the economy. Thus, if foreign trade policy is changed as suggested earlier, without allowing the effects to be offset by higher costs and prices, employment should increase. As noted earlier, this can probably only be achieved with the help of fiscal restraint and monetary and credit policies that hold down inflationary pressures. 128. Other important policies dealt with in this report that would affect employment and income distribution are assistance to small-scale industry. (See paragraphs 141-160 below.) Small-scale firms are more labor intensive and, moreover, they directly affect a higher percentage of poor people than does employment in larger firms becaue not only their workers but often the entrepreneurs have lower incomes than some industrial workers in larger firms. 129. In addition to these two topics, reducing the large distortion in factor prices in Mexico, which lowers the rate of job creation in manu- facturing, merits consideration. Mexico, as many other developing countries, achieves much of its promotion of industrial expansion by measures that make capital equipment cheaper, while it uses measures that increase the cost of labor to achieve part of its income distribution and other social goals. Import duties on capital goods are low, and are often waived completely. The overvalued exchange rate of the last few years has further lowered the peso price of machinery and equipment. Loans for the purchase of fixed plant and equipment are more readily available than for working capital. These measures lower the cost of equipment by 25 to 50 percent, compared to what it would have been in their absence. On the wage side, the "seventh day" payment, the 5 percent INFONAVIT contribution, the vacation bonus, the one percent edu- cation tax and the 9 to 12 percent social security tax, the year-end bonus, and the eight percent of profit sharing levy raise the cost of labor by as much as 50 percent (see Table 1.11). Moreover, the minimum wage itself raises labor costs above a free market level. For workers earning the minimum salary, these charges are borne completely by the employer, or by his cus- tomers if he can pass on the increased costs in the form of higher prices for his products. For workers earning above the minimum salary, however, a part of these charges may be reflected in lower nominal wages, and thus be borne by - 46 - the workers. The net result of the increased labor charges and the subsidies to equipment purchasing is that the cost of labor, relative to the cost of equipment, is perhaps twice what it would be in the absence of these measures. Such a large distortion must affect the relative amounts of the two factors used, and is especially inappropriate in view of the explosive growth of Mexico's labor force. 130. Since these fringe benefits, and the minimum wage, are not available to many workers, distortions and inequities are created. Migration from rural to urban areas, always sensitive to wage differentials, is increased by the higher wages available in cities. On the other hand, higher labor costs in the manufacturing enterprises that must pay these fringe benefits may protect smaller, "informal" enterprises which do not pay them, and thus are better able to compete. This last effect may be important even if the smaller enter- prises do not produce the same products, because higher prices for products from larger enterprises may shift consumption to some extent to other products, or services, that are produced by informal enterprises or by individuals, and therefore have lower unit wage costs. 131. A shift toward neutral promotional instruments would increase job creation to some extent. Income tax rebates or temporary tax holidays could be used instead of subsidies that keep equipment prices low. Benefits for workers could be financed from general public revenues (which could be increased by the elimination of tariff waivers recommended earlier and general fiscal reforms), rather than through wage taxes. On the wage tax side, the prime candidates for change are the INFONAVIT contribution, the education tax, and the social security tax. Financing these benefits from general revenues would reduce wage costs to employers by about 11 percent for workers on the minimum salary; for higher-paid workers the change would be divided between lower wage costs for employers and higher nominal wages for employees. 132. Such changes would also require modifications in laws and insti- tutions. In the case of INFONAVIT, only workers who contribute are eligible for INFONAVIT housing. The kind of change proposed above could perhaps be approached through an intermediate step of financing the fringe benefits through a special, "set-aside" portion of a value added tax, if such a tax is substituted for the present sales tax, with eligibility for INFONAVIT housing limited to workers in firms that pay the tax. Legal changes would also be required to change Social Security Financing. Such changes could be made in ways that would increase equity, through broader coverage, as well as effi- ciency. 133. Other labor policies such as job security and annual wage negotia- tions should also be studied for possible modification, so as to reduce the burden to employers without unfairly hurting the employees. 134. Trends in wages and productivity are difficult to analyze because of problems in data; only approximate estimates are possible. As measured by all wages in a sample of firms, average real wages grew by about 7.5 percent per year from 1960 through 1964, by about 2.7 percent per year from 1965 through 1970, and by about 3.6 percent per year from 1971 through 1976. The annual rate for the period as a whole was about 4.2 percent (see Table - 47 - 1.12). The real minimum salary, which some economists believe to be a more accurate measure of trends in average wages, grew slightly faster: 4.9 percent per year over the period. Increases in fringe benefits further increased the real income of workers in the organized urban sector. Output per worker in manufacturing grew somewhat faster than wages, at 5 to 7 percent per year. This productivity growth, which is much faster than ,hat observed in most countries, may be mostly attributable to the shift to more capital- intensive frocesses in a given sector. This slift is itself a result of the distorted faster prices discussed in paragraph 126. A smaller part of the productivity growth is due to a shift towards more capital-intensive prodUcts; however this shift, as shown in Tables 3.1 and 3.2, has been small. 135. Most of the increase in real wages during 1971-76 occurred in 1976. The annual increase had been under 2 percent through 1975 (as measured by either the average wage in the sample of firms or average minimum wages); in 1976 the minimum wage rose 13 percent while average wages rose about 16 per- cent, in real terms. (All these estimates are based on annual data, and thus approximate mid-year to mid-year changes. The December-toDecember increase in 1976 was even greater.). 136. From 1972 on, wages in Mexico rose relative to those in the ''nited States, along with the increasing overvaluation of the peso (see Table 1.13). Devaluation in 1976 brought the relative wage back to the level of the late 1960's. 137. Returning to a pattern of moderate wage increases should contribute to economic stabilization, export promotion, and employment growth. This could be done by a combination of measures, including systematic efforts to reduce the size of any official nationwide wage increases, revision of the Federal Labor Law to allow contracts of two or three years with pre-agreed wage adjustments, rather than re-negotiating contracts each year, and guidance of informed opinion toward a recognition that a realistically modest wage cost of production is essential for rapid economic growth. Employment growth could be further promoted by shifting costs of some fringe benefits from payroll taxes to general tax revenues, as already noted. Further, current rules in regard to temporary layoffs and termination of employment greatly increase the costs and risks of hiring workers. Their effect is to make labor costs almost fixed rather than variable, and hence they add to other biases favoring using more equipment instead of more labor. These rules might be revised, and the apparent reductions in workers' real wages and job security that might result would be more than offset by the beneficial effect of increased output and employment throughout the economy. It is recognized that many of these changes would be difficult to make. 138. In considering the equity aspect of industrial wage policy, it is well to remember that organized industrial workers are at or above the middle of the income scale in Mexico. A worker earning the average minimum salary in 1975 received a monthly wage (including profit sharing and the year-end bonus) of about Mex$ 2,056 -- about $160 in US dollars. Since a famil with - 48 - income of Mex$ 2,056 per month was at the 50th percentile of Mexico's income distribution in 1975, and since most families have more than one worker, a typical family headed by a worker earning the minimum wage was above the median income in Mexico. 139. If real wages were to fall without other suitable measures being taken at the same time, there would be a redistributive effect in favor of employers which might well lead to increased consumption by the well-to-do. This could be prevented by (a) liberalization of import controls and increasing export profitability so that inefficient enterprises are squeezed while those that are efficient or can become efficient, and whose productivity is liable to grow quickly making future wage increases possible, would increase their share of the market; (b) designing real interest rates, tax measures, and credit policies so as to induce employers to channel their income into savings and productive investments rather than into consumption, making future employ- ment increases possible. 140. To keep product prices from unwarranted increases, a non- inflationary macroeconomic policy is obviously a sine qua non. Two addi- tional measures might help achieve stable product prices. The more important and effective measure would be to eliminate import licensing. This would limit price increases to those in world markets. It would create a structure of prices among products that would induce greater growth in products Mexico produces efficiently, and would also increase the motivation for efficiency and quality improvements by providing a reasonable amount of competition from imports. A second measure that might also be temporarily useful during a period of stabilization would be price agreements -- widespread voluntary limits on allowable price increases -- rather than spotty controls on selected products. Experience in other countries shows that this form of restraint, although virtually useless by itself, can be useful as part of a stabilization program. If monetary policy, fiscal policy, and wage policy are designed to form a coherent package to reduce inflation to a given target level, then price restraint set at or near the same target can help to condition expecta- tions and generally help achieve the target. However, experience worldwide also shows that controls without the rest of the package, or over too long a period, are not effective in restraining the overall price level and induce distortions in relative prices, investment, and eventually in output. It is possible that the "Alianza para la Produccion" and price-control agreements for wage goods may furnish a useful part of the framework for effective price controls in Mexico. Small and Medium-Scale Industry 141. Overview: Small and medium sized enterprises (SMI), defined as those with up to 250 employees, play a very important role in manufacturing in Mexico. According to the 1970 Census, they accounted for 45 percent of - 49 - production and 60 percent of employment in the sector; moreover, these esti- mates almost certainly underestimate their true importance because the Census misses smaller enterprises proportionately more than larger ones. In nearly all industrial subsectors SMI enterprises predominate by number of firms, and outside the three largest cities the vast majority of industrial enterprises are small or medium sized. 142. Small enterprises (those with up to 25 employees) represent over 90 percent of the total enterprises comprising the SMI range, and on average have only 4 to 5 employees. Medium sized enterprises (those with 25 to 250 employees), although they represent only 10 percent of SMI enterprises, con- tribute 80 percent of value added by SMI firms and provide 60 percent of employment. They average 75 workers per firm. Small scale firms are gen- erally much less capital intensive than medium and large scale firms but in many branches of industry they operate efficiently and competitively. 143. On average SMI enterprises appear reasonably profitable, reporting earnings of 10 percent on sales and 18 percent on equity capital. Neverthe- less, in common with enterprises of similar size in other contries, many SMI in Mexico have deficiencies in accounting, administration, production and marketing which stem from the small size and lack of specialization of their management team. They also encounter greater difficulties in securing ade- quate financing from normal commercial sources, and suffer from the tendency of industrial policies to favor the larger enterprises. All these problems are particularly serious in the smallest firms. 144. Given their great number, their labor intensiveness, and their predominance outside the main cities, SMI can clearly play a major role in achieving national goals for employment generation and regional development. Furthermore, SMI enterprises provide the seed bed for the development of new entrepreneurial talent and for upgrading the skills of the labor force, and by their linkages with large industrial enterprises and with other sectors such as agriculture they affect the entire economy. They also produce a large percentage of mass consumption goods, and thus increasing their output and keeping their costs down could be a useful ingredient in a price stabilization package. Some strengthening of the support provided to SMI is therefore clearly justified. 145. Industrial policies and SMI: Certain aspects of present policies of industrial protection, export incentives, employment and wages, and decentralization are relatively disadvantageous for SMI in general and small firms in particular. 146. The existing system of protection in Mexico, which depends more on quantitative controls than on tariffs, tends to create difficulties for SMI enterprises mainly in the supply of inputs. In a recent CANACINTRA survey of SMI, a high proportion of firms complained of high raw material prices and excessive price fluctuation, shortages of production inputs, and delivery delays causing production losses. Production problems were also caused by poor and inconsistent quality of available raw materials. While these problems also affect some larger enterprises, they are more acute for - 50 - small firms which have weak purchasing power, limited resources to finance large inventories and more difficult access and lower leverage when applying for import licenses. Some smaller firms also have more difficulty than do larger firms in assuring protection for their output through denial of import licenses. A shift in the protection system towards tariffs and away from quantitative restrictions, particularly for industrial raw materials and intermediate goods, would be welcomed by the majority of SMI firms. 147. While many SMI firms are producing goods that are potentially competitive in international markets, comparatively few are taking advantage of existing export incentives (tax credits, concessionary credit, duty exemp- tions, etc.). The bulk of Mexican exports come from the larger firms, al- though in some subsectors such as food, textiles, apparel and footwear a significant proportion of exports do come from smaller firms. Despite recent efforts by the Instituto Mexicano de Comercio Exterior (IMCE), many small enterprises find it difficult to get adequate information regarding demand, prices, qualities and styles in external markets and to make appropriate contacts with foreign buyers and distribution channels. In addition, smaller enterprises often do not have sufficient volume of production to interest prospective buyers and lack confidence in their own ability to operate suc- cessfully in the more sophisticated and exacting world of export markets. To help overcome these natural disadvantages, further attention might be given to strengthening the system of intermediation between SMI enterprises and their potential export markets. This might be done through the establish- ment of trading companies which would develop export contracts and sub- contract production to a number of small firms -- as is already being done in the shoe industry. Alternatively, more active cooperative efforts among the SMI enterprises might be encouraged through chambers of commerce and trade associations, as has been achieved in exports of fruit and vegetables. 148. With regard to employment and wages policies SMI firms also have some disadvantages. Here the disadvantages are greater for medium-sized firms, becaute many of the smallest firms do not always follow the regula- tions (and in fact thus gain an advantage against their larger competitors). Because SlI firms are generally much more labor intensive than large firms, labor costs and employment regulations have a much more important influence on their production costs. However, there is a tendency for employment and wage policies to reflect the situation and payment capacity of large firms, because of the latters' greater visibility and the greater strength of the unions which represent their workers. 149. In general minimum wages do not appear to present serious difficul- ties to small enterprises, especially since the legal minima are differentiated by region. However, a high proportion of SII enterprises consider that the level of social and other non-wage benefits for workers, which add about 50 percent to direct labor costs, is becoming a serious burden. In addition, other aspects of labor legislation such as job security, severance pay and minimum hours of work substantially reduce the flexibility of smaller firms. The severity of these regulations tend to make firms cautious about expanding their labor force to take advantage of market opportunities, and tend to encourage them to switch to more capital-intensive methods as they expand. - 51 - 150. Incentives to promote the decentralization of industry, including fiscal incentives, direction of credit through trust funds and the establish- ment of industrial parks, do not appear to have achieved much success with SMI. Fiscal incentives available are mostly subsidies to capital and thus favor the larger, more capital intensive firms. About one third of SMI firms located outside Mexico City, Monterrey and Guadalajara have been able to benefit from the fiscal incentives, but it is doubtful whether many small enterprises based their locational decisions on the availability of these incentives, or whether the existence of these incentives significantly influenced their creation or viability. 151. Loans discounted with FOGAIN bear slightly lower interest rates to the borrower outside Zone 1 (the 3 big cities) but the differences are too small (1 or 2 percentage points) to provide any significant incentive. Fur- thermore, because banks and financieras receive the same margins, irrespective of zone, on loans discounted by FOGAIN, SMI firms located outside the 3 major cities have had no advantage in access to credit. FOMIN has also tried to give priority to Zone 3 (the least developed part of the country) in its equity investments, but overall its impact has been small compared to the needs of StMI in the Zone (only 24 new enterprises in Zone 3 have been assisted so far). Under FIDEIN's industrial estates program, most lots sold have gone to medium- sized and larger enterprises primarily because the costs of purchasing even the smallest available lots and building a simple factory are beyond the scope of most small enterprises, and adequate financing for such projects has not been available to the smaller firms. FIDEIN has developed proposals to con- struct factories for lease to small enterprises and to construct common service facilities for groups of small enterprises, but has lacked the funds to implement these proposals. 152. Finance: The accelerating inflation and tighter monetary control since 1973 have reduced credit availability, especially for the smaller firms. By 1974 about )0 percent of SMI firms had to supplement credits obtained from normal banking channels by funds borrowed from money lenders and other un- official sources, and less than half of SMI firms were able to meet all of their credit requirements. The existence of rediscounting institutions like FOGAIN has not made a substantial contribution to improving the access to credit for the smaller firms. In its 22 years of operation to date, FOGAIN has discounted credits to about 10,700 SMI enterprises, which represents only 14 percent of eligible enterprises. Even more important, FOGAIN has financed 60 percent of eligible medium sized enterprises but only 9 percent of eligible small enterprises. This difference arises because the intermediary banks who normally carry the credit risk prefer to lend to medium-sized enterprises under prevailing conditions of FOGAIN discounts. Transaction costs are higher for small loans, but FOGAIN lending margins are the same regardless of loan size. It is usually easier to obtain adequate collateral from medium-sized borrowers and FOGAIN's guarantee facility is not sufficiently attractive to encourage banks to lend to firms with inadequate collateral. Moreover, the rudimentary accounting systems of many small enterprises make it difficult for the firm and costly for the banks to prepare the basic information required to obtain a FOGAIN discount. For FOGAIN to facilitate access to its discounts by - 52 - small firms, it may be desirable to review the appropriateness of its lending conditions and to increase the attractiveness of its guarantee arrangements. In this context it is important to recognize that access to credit is even more important to small enterprises than concessionary interest rates. 153. FOMIN, the other trust fund that mainly serves SMI, is not a redis- counting mechanism but assists SMI enterprises directly through minority investments in equity capital. The majority of FOMIN's investments are in companies whose sizes fall in the middle to upper portion of the SMI range. The scale of FOMIN's operations is relatively small compared with those of FOGAIN, and the transaction time and costs of its individual financing oper- ations are inevitably very much greater. While FOMIIN is providing a useful service in facilitating the creation of new enterprises and the expansion of certain existing enterprises, its impact on the problems of small enter- prises is likely to remain limited, unless it can find ways to reduce the associated transaction costs and develop new forms of financing that are more in keeping with the characteristics of small enterprises. Generally small enterprises find it advantageous to be structured as family concerns or partnerships for fiscal reasons, but this makes it difficult for FOMIN to participate as an investor in common stock. In addition, small enterprises would find it uneconomic to employ external auditors and provide the reporting information normally required by FOMIN. Perhaps some form of quasi-equity loans with low interest and principal repayments in the early years, or subordinated loans with convertibility features, would be more appropriate instruments in the case of small enterprises. 154. Smaller enterprises have similar problems when dealing with other trust funds. With FOMEX (concessionary export financing) they must first find a bank prepared to accept them as a credit risk; with FOTEP (direct loans for preinvestment studies) they must also satisfy conventional commercial lending criteria. 155. Technical assistance: SMI enterprises in Mexico suffer from a wide range of problems apart from limited access to finance. These problems are associated with enterprise size and the background of the entrepreneur and are not fundamentally different from those facing small industries in other indus- trialized and semi-industrialized countries. They may include: (a) inadequate financial management and accounting; (b) poor production planning and control; (c) poor quality control; (d) difficulty in keeping abreast of technological developments; (e) limited knowledge of how to expand into new markets; and (f) management weakness in areas such as purchasing, inventory control, labor administration, etc. 156. Despite the availability of numerous public and private institutions offering specific types of technical assistance, comparatively few SMII enter- prises, particularly small enterprises, are receiving appropriate technical assistance to help overcome their particular problems. The reasons for this are several. Many small enterprises are unaware of their managerial and tech- nical deficiencies and therefore do not seek outside help in diagnosing these deficiencies. In addition, the entrepreneur is often unaware of what sources - 53 - of technical assistance are available, and has not readily available funds or sources of credit to finance such assistance. Existing technical assistance institutions direct their attention mainly to larger companies, and the con- tent and cost of their services is often inappropriate to small enterprises. Furthermore, most of these institutions concentrate their activities in the three largest cities and are not known to, or easily accessible by SMI enter- prises located in other regions. 157. To combat this lack of information and lack of access, a strategy of linking the provision of technical assistance more closely to the provision of financial assistance might help. Banks and financieras would try to help identify the non-financial problems of their smaller clients when they are evaluating a loan application. Nacional Financiera, acting as a direct lender to SMI, and other public sector and mixed institutions, could be encouraged to take the lead in such a program. 158. The system for delivering technical assistance to SMI needs to be strengthened in terms of geographic coverage and relevance to SMI. This might be accomplished by (a) creating a regionally based extension service for small industries; (b) establishing a network of small business advisory centers at provincial locations in association with local chambers and trade associations; and (c) coordinating much more closely the scope and content of the ongoing technical assistance programs, to develop a more comprehensive and appropriate package of technical assistance support for SMI. 159. Better arrangements are also needed for financing the cost of tech- nical assistance. While part of the cost of technical assistance could perhaps be capitalized with the loan and recovered from the enterprises assisted, the cost must be spread over a reasonable time to make these services affordable. In practice, a package of financing arrangements involving contributions to the lenders' overhead costs by public and private sector bodies, higher interest rates to the borrowers on loans to SMI discounted with FOGAIN, and fees charged directly to recipients of technical assistance with appropriate credit arrange- ments being made available, will probably be required. 160. Need for coordination and planning: While many institutions are providing support to SMI, there are still gaps in the availability of some important services, and the means by which recipients can pay for services offered are deficient. Insufficient contact between the institutions and SMI enteprises often results in a lack of awareness of what services are available and in inappropriate design of these services from the point of view of the recipients. To concentrate improvements where they would be most useful, it may be advisable to establish a high level body responsible for formulating and monitoring the implementation of national policies and programs in support of SMI, and for coordinating the activities of the numerous institutions in- volved. A small committee composed of high officials from the most important institutions involved, and from SHIs, might be more effective than a larger or separate agency. Special emphasis is needed on the closer integration of financial and non-financial assistance, and on design of programs aimed specifically at the smaller enterprises in the SMI range. - 54 - Capital Goods 161. The situation: The capital goods sector in Mexico is less well developed than the economy's size and level of development is able to support, and the kinds of goods that are produced seem to be poorly related to Mexico's likely comparative advantage. A wide variety of heavy machinery and equipment is produced by about 80 large and medium-scale enterprises, but production is largely limited to import substitution of relatively unsophisticated final products. Production of intermediate inputs is poorly developed. 162. Progress in domestic production of capital goods can help provide Mexico with technology appropriate to her needs, and can promote technological and scientific advance in general. However, these relations are two-edged swords: just as efficient development of capital goods production could bring many advantages to Mexican development, inefficient growth of this sector would spread high costs and technological backwardness throughout the economy. 163. A recent study found that Mexico's production of capital goods was considerably retarded (Gomez Palacio, 1976). Unfortunately international com- parisons are difficult because consumer durables are often lumped with capital goods in the data. The study notes, for example, that in 1970 Mexican produc- tion was less than that of Brazil or Argentina in each of four major branches of metal products industries, except for electrical equipment where Mexico's production exceeded Argentina's, partly because of assembly plants. Mexico's lag was most marked in non-electric machinery, where her output was only one- third that of Argentina and one-fourth that of Brazil, while apparent consump- tion of capital goods in Brazil was only about twice that of Mexico. MIexico's lag in non-electric machinery and in transport equipment is also shown in Table 1.14, which shows imports in relation to domestic consumption. In machine tools, Mexico produced only 7 percent of her consumption in 1970, while Brazil and Argentina produced 54 and 49 percent, respectively--even though Mexico's consumption of machine tools was higher than that of Brazil or Argentina. Trends since 1970 have put Mexico even further behind. 164. The problems of capital goods manufacturing in Mexico can be seen through examination of four sub-sectors: the foundry industry, which provides inputs to a wide variety of products; the machine tool industry, which is a good indicator of the level of manufacturing sophistication achieved by a country; metal fabrication, which includes many labor-intensive special order goods; and heavy electrical equipment, which supplies goods mainly for public enterprises. 165. The amount of foundry capacity in Mexico, above 470,000 tons/year of cast iron and 100,000 tons/year of cast steel, is reasonable for a country of Mexico's size. However, well over 50 percent of this capacity produces vir- tually exclusively for the automotive sector. The remainder of the foundries are not only small in their total capacity, in relation to the size of the economy, but also have major problems of cost and quality, which can be traced to: - 55 - (i) Raw materials - Efficient foundry operation requires raw material supply of consistent and proper grades. Foundry- grade pig iron, used in most industrialized countries, is not available in Mexico. (Pig iron from the steel mills would have to be treated further to be suitable for foundry use.) The foundry industry is therefore heavily reliant on scrap for making both iron and steel castings. Domestic scrap is in short supply and substantial quantities must be imported, and thus the price is at least somewhat above that in the US. More important, deficiencies in ability to adjust to the varying quality of scrap are a major cause of excessive casting defects in Mexico, which often cannot be detected until some or all of the subsequent machining has been done, thereby adding further to the cost. (ii) Productivity - Except for the more efficient automotive foundries, Mexican foundries average over 100 man-hours/ton of products, as compared with 40-50 man-hours/ton in compa- rable foundries in the United States. Also, rejection rates are high, above 10 percent. 166. The few Mexican foundries outside the automotive sector that can produce high-grade castings enjoy a near monopoly, and give delivery dates on orders as long as 6-8 months. Their prices are pegged to the landed price of imports; however, this suggests either large profits or very lax management -since labor accounts for about 30 percent of foundry costs in industrialized countries and Mexican wages, even at the overvalued exchange rate prevailing in early 1976, were only about one-third of US wages. Mexico should therefore have a substantial cost advantage in foundry products. The reasonably high quality of that part of the foundry sector that produces automotive castings testifies further to Mexico's ability to produce efficiently in this sector. 167. The underdevelopment of the machine tool industry illustrates the difficulties surrounding both high precision and simpler machinery manufac- ture in Mexico. From 1970 to 1975 production stagnated at around $5 million; by comparison, in Brazil production over the same period increased from around $30 million to over $100 million. In metal cutting machine tools there are 3 or 4 makers of simple lathes and milling machines -- all small and engaged essentially in the assembly of simple imported models. The price of a medium-sized center lathe produced in Mexico was over $12,000 (at 12.50 pesos per dollar), which was roughly double the world price for such a product. Devaluation would not make this machine much more competitive, since most parts are imported. While the simpler products are produced at high cost, higher precision machine tools are almost all imported. Mexico could produce cheaper simple machine tools, as well as some higher precision pro- ducts, if more appropriate technology could be imported and selective backward integration pursued. 168. The situation in the fabrication subsector is better, because fabricated products (heat exchangers, pressure vessels, cranes, drill rigs, - 56 - etc.) generally do not require complex production processes. Nevertheless, before the recent devaluations, local producers had some difficulty competing against imports because of difficulties in getting assured access to certain imported inputs, and inadequate specialization and scale of production caused by excessive vertical integration in the industry -- no one likes to rely on subcontracting to local suppliers, and hence firms produce many inputs them- selves. Moreover, the international trend towards increasingly larger-capacity equipment that demands more exacting designs and manufacturing methods may further reduce the competitiveness of local manufacturers. 169. The electrical equipment subsector has not developed according to Mexico's comparative advantage. Domestic production is concentrated on smaller "off the shelf" items, while almost all heavy equipment is imported. Many of the smaller items are produced at high costs, mainly because of grossly sub-optimal scales of production, while Mexico probably could be competitive in some of the heavy items which are more labor intensive and more specialized -- for example, switchgear and transmission equipment. 170. The main reason for this inappropriate development within the elec- trical equipment sector is the way the import licensing system operates. The market for heavy electricals in Mexico is domirnated by CFE, which has virtually unlimited access to imported equipment, available with generous financing and (until recently) at the overvalued exchange rate. In contrast, the market for smaller equipment consists mostly of private sector firms which must purchase from domestic manufacturers because they cannot obtain licenses to import. The heavy electrical sector presents a classic example of how distorted incentives can distort a country's production structure, leading to domestic production of some items at high costs while other items that could be produced more efficiently are being imported. 171. In summary, the capital goods sector in Mexico is less developed than it might be, its structure may not be appropriate and in many parts its productivity is low. Underlying this situation is a basic problem: Uncertainly about future demand, caused by lack of planning by customers and the arbitrary nature of import licensing, leads to sub-optimal scale of production. The licensing system also leads in some cases, to i-nappropriate choice of products for import substitution. The sub-optimal scale, combined with inappropriate choice of products, in turn inevitably mean high costs of production. 172. On a more detailed level, the reasons for the low level of develop- ment are: (a) Fairly low protection for capital goods on the average, and high exposure to imports of both new and used equipment from the United States. For private sector importers tariffs are low and often waived while government agencies, which form a large part of the Mexican market for capital goods, have virtually unlimited access to imported equipment and import it at zero tariff rates. Until September 1976 the overvalued exchange rate contri- buted further to the low protection against imports. (b) Lack of credit facilities through which Mexican producers could compete with favorable financing offered by foreign suppliers. The inappropriate structure and low productivity of the sector are due to the following factors: (a) the access of government agencies to imports, mentioned above, drastically reduces - 57 - domestic demand for larger and more special equipment, much of which Mexico could probably produce fairly efficiently. The private sector, which faces more difficulty in importing equipment, tends to use smaller and more stan- dardized equipment; in industrialized countries this is produced either in mass or in very large series, and hence at lower costs, than can typically be achieved in Mexico; (b) denial of import licenses (to the private sector) for goods that are available domestically, with virtually no regard to relative cost or quality, results in import substitution which often bears little or no relation to comparative advantage; and (c) lack of analysis of demand, com- parative costs, and resulting definition of priorities among different types of capital goods and intermediate products used to make capital goods. 173. The demand for capital goods can be expected to grow in pace with the anticipated investments in oil, petrochemicals, steel, power generation and other capital-intensive sectors. While the total demand is large, it is highly diverse as to type and specification of equipment and includes some sophisticated items beyond the capacity of local makers to produce. Local demand alone for some lines of capital goods may not be large enough to sus- tain economic of efficient production; the market therefore needs careful identification. 174. Policy issues: Promotion of efficient capital goods production in Mexico could be pursued, in part, by (a) provision of an adequate rediscount- ing facility that would enable Mexican capital goods producers to finance sales (both domestic and export) on competitive terms; (b) ending preferential access of public enterprises to imported equipment (this means treatment equal to private sector importers both as to licensing, if it continues to exist, and as to tariff payment); and (c) ending or restricting the import licensing system, as already recommended for manufacturing as a whole, so as to increase specialization, scale of production, and quality and price competitiveness. 175. The fairly low overall protection afforded capital goods, and the access to used equipment from the United States, represent important advanta- ges to the Mexican economy. To dispense with them in order to promote produc- tion of capital goods would be very expensive. A reasonable level of protec- tion -- say 15 to 20 percent in a system where the all-manufacturing average is around 10 or 15 percent and the exchange rate is realistically valued -- could be afforded capital goods and imposed on all importers of such goods. Slight additional protection against imports of used equipment might be justified, especially for sectors with strong export potential, to promote the use of modern technology. 176. If development of the sector is to be efficient, the strategy for identification and promotion of capital goods production in Mexico must eventually aim at export markets and not limit itself to import substitution. This approach is important not only to take advantage of scale economies, but also to assure continuous contact with world standards of quality and price. Reasonable cost, good quality, and appropriate technology are perhaps even more important in capital goods production than in other manufacturing, because of the pervasive effects of capital goods on production throughout the - 58 - economy. Foreign private investment is often important for access to tech- nology and to export markets. Thus persistent high protection, restrictions that unduly impede foreign investment, and weak export incentives would be most serious impediments to development of the capital goods sector. 177. Other measures to improve efficiency in capital goods might include: (a) comprehensive analysis of the likely growth in demand for various finished capital goods and intermediate inputs, taking into account the complex inter- dependence among products and from this estimating possible scales and costs of production, and competitiveness; (b) government promotion of R and D, qual- ity control and testing, and norm setting for the sector; and (c) assistance to existing firms (especially in intermediate products), to solve problems such as obsolete equipment in some lines (e.g., in certain types of casting). Different firms will need different packages of technical assistance and credit; such help to established firms may be at least as important as crea- tion of major new capacity. Regional Development and Manufacturing 178. The World Bank has recently issued three studies relating to spatial policies in Mexico. IJrban Development in Mexico (IBRD 1449-M1E, January 31, 1977) is about the historical development of Mexican cities and regions; Spatial Development in Mexico (IBRD 1081a-ME, January 31, 1977) analyses policy options, and The Economic Development of the Isthmic Region of Mexico (IBRD 1080-ME, March 30, 1976) treats in more detail the possibilities for development in the Isthmus of Tehuantepec. Because of these recent studies, the mission did not study regional questions. However, the conclusions of the earlier Bank reports most relevant to the manufacturing sector, as well as those of other studies, may be summarized here. 179. Manufacturing is highly concentrated, spatially, in Mexico. The Federal District and the surrounding State of Mexico accounted for 52.1 per- cent of all manufacturing production in 1975; Nuevo Leon and Jalisco, the States where Monterrey and Guadalajara are located, had 9.7 and 5.2 percent respectively, and none of the other 28 states had more than 4.0 percent of the national total. 180. Such concentration is not unusual for a semi-industrialized country such as Mexico. Industrialization does not and cannot occur everywhere at the s pace. The capital city, and occasionally a few others, get a head start, ai,d subs 7quent development of the transport system, a large local market, a large and relatively skilled local labor pool, relatively better economic and social infrastructure, concentrated sources of information and both public and private decision makers, and simple "follow-the-leader" risk-minimizing location decisions all reinforce each other to strengthen the initial advan- tages. 181. As development proceeds, the largest centers start to experience growing diseconomies such as high land and labor costs, congestion, pollution, - 59 - etc., while infrastructure, information, and other positive factors become more and more developed in other cities. Thus a counter-trend of decentral- ization starts, usually with industries that can take advantage of low-wage, low-skilled labor in highly routinized activities. This counter-trend is already evident in Mexico. For example, Monterrey and Guadalajara, the nation's second and third cities, are growing even faster than Mexico City, and over forty other cities with populations over 100,000 are growing faster than the national average. However, Mexico City and its metropolitan area is still not only dominant but continues to attract much additional manufac- turing. 182. Experience throughout the world shows clearly that the forces behind these trends are very strong, and that even strong policies can have at best small impacts on the trends. Nevertheless, a number of governments, including the Mexican government in recent years, have tried to develop backward regions and/or to slow down growth of major cities, and in general to reduce regional inequalities. Such attempts have many motives, usually including a concern for poor people in backward areas and a desire to reduce the negative effects and management difficulties of fast-growing large cities. In all too many cases, however, the policies adopted are weak, poorly planned, and at odds with each other. 183. Mexico, like most other nations, has not had a strong explicit spatial policy, and instruments that affected spatial development have not been well-coordinated--at least until very recently. Among the most important measures (both those designed with regional objectives in mind and those where regional effects are unintended by-products) have been the following: (i) A number of tax exemptions have been available, since 1972, to industries that located outside the three largest metro- politan areas. Unfortunately, these incentives are both too small to affect most location decisions, and are available for plants that locate in several high-income, highly urbanized, rapidly growing areas as well as in poorer and more backward areas. Therefore, the main effect of the incentives has been to increase profits for firms that would have located in these growing areas in any case. (ii) Industrial parks, complexes, and cities, and commercial centers, have been set up in many places, especially since 1971. A few of these have been quite successful, such as the Queretaro "industrial city." However, most have met with but little success because the main benefit provided --physical infrastructure--is not sufficient to attract most business firms to the unfavorable locations chosen for many of the sites. (iii) A number of programs assist small and medium-sized industries (see paragraphs 141-160). Almost all business firms outside the three largest cities are small enough to be eligible, and thus the programs may have helped develop some of Mexico's - 60 - poorer regions to some extent. About one-third of all eli- gible firms outside the three major cities have been helped by these programs. (iv) The maguila program, in which goods are assembled at least in part from inputs imported in bond and then re-exported to the United States, has increased growth in several border cities. (v) Low prices in Mexico City for water, and nationally for gasoline, diesel fuel, and rail transport of raw materials, have all promoted growth in Mexico City. Many of these prices have been raised in the last few years, but to help induce even a little decentralization they would have to be con- siderably higher. 184. Within the last year Mexico has moved towards establishing a stronger, coherent and explicit spatial policy. The Law of Human Settlements, adopted in June 1976, provides the legal authority for planning and implement- ing programs to achieve spatial goals. The stated objectives of the law in- clude improved rural-urban integration, more balanced growth as among cities and regions, promotion of growth in medium-sized cities as alternatives to the largest cities, greater citizen participation in solving urban problems, land use control, better provision of urban services, and improvement of the housing location-job location-commuting situation within cities. The major powers of the law are the Federal government's responsibility to make a national urban development plan and to coordinate all public sector investment with that plan, local governments' responsibility to plan for and control land use within cities, and the creation of "conurbation commissions" to deal with local gov- ernment problems in urban areas that cross state boundaries. It is still too soon to tell how these powers will be used, but they do have significant potential. 185. As explained in the Bank's report on spatial development, cited earlier, it would probably make very good sense for Mexico to adopt spatial policies with the two goals of (a) bringing access to markets, jobs, educa- tion, and all the other "urban amenities" closer to people who live far away from the country's three major cities, and (b) slowing down the growth of Mexico City. Affecting the location of manufacturing activity is crucial for the success of such an attempt. In turn, the key elements of policies to induce manufacturing to locate away from Mexico City, and also in certain backward but promising areas, are: (i) Incentives must be limited to a few areas, and those areas must have significant potential for development. Experience in country after country, including recently in Mexico, shows conclusively that incentives that are either not focused or are focused on the most backward regions almost never work. Recent studies by the Bank and others have identified some promising regions: selected places in the Isthmus of Tehuantepec, the Gulf Coast, and the Lazaro Cardenas area on the Pacific Coast. - 61 - (ii) Incentives must be strong. Tax exemptions such as those now in force in Mexico are seldom sufficient, although if concen- trated on a few appropriate locations they would be of some use. An effective package could include (a) tax exemptions; (b) labor taxes on new plants that locate in Mexico City combined with labor subsidies for those that locate in the areas targeted for growth; (c) government investment in both industry and infrastructure in the target areas; and (d) increased prices for public utilities in Mexico City, especially in the form of high initial connection charges for new industrial users. (iii) Incentives must be enforced for a long time. Regional policies need at least 5 or 10 years to take hold. (iv) Finally, the explicitly spatial policies must be comple- mented by other policies needed to help the target popula- tions. Promoting growth in selected urban centers will help the local and surrounding rural population only if these people also have access to credit, prices for their labor and their products, education, training or technical assistance, land, and the other ingredients of economic development. ANNEX I TABLES AND CHARTS Table 1.1 MEXICO: Imports as Ratios to Domestic Demand, 1967-1974 S e c t o r 1967 1968 1969 1970 1971 1972 1973 1974 la Petroleum extraction and refining and coal products .027 .022 .023 .019 .031 .036 .085 .089 lb Basic petrochemicals .069 .055 .063 .067 .085 .101 .085 .079 2 Meat products .014 .009 .009 .015 .014 .010 .020 .035 3 Flour milling and baking .002 - - .001 .002 .002 .004 .002 4 Other food products* .021 .018 .018 .025 .024 .031 .040 .054 5 Beverages .007 .006 .005 .005 .011 .014 .014 .013 6 Tobacco .005 .007 .001 .003 - .001 - - 7 Textiles .020 .016 .018 .016 .015 .015 .022 .027 8 Clothing, including footwear .009 .012 .015 .015 .018 .019 .021 .019 9 Wood products .035 .037 .043 .044 .041 .046 .055 .051 10 Paper products .120 .115 .133 .139 .112 .109 .133 .179 11 Printing and Publishing .047 .051 .046 .044 .061 .112 .088 .071 12 Leather and products .011 .011 .013 .013 .012 .019 .017 .014 13 Rubber and products .060 .048 .047 .043 .036 .038 .044 .050 14 Basic chemicals .460 .438 .401 .363 .362 .390 .398 .486 15 Synthetics and plastics .181 .154 .148 .128 .098 .095 .090 .207 16 Fertilizers and insecticides .097 .120 .076 .057 .076 .081 .070 .099 17 Perfumes and soaps .009 .009 .008 .006 .008 .010 .011 .016 18 Pharmaceuticals .126 .170 .136 .122 .141 .129 .119 .110 19 Other chemicals .146 .144 .135 .146 .120 .139 .156 .181 20 Non-metallic mineral products .036 .037 .036 .031 .026 .025 .029 .027 21 Basic metals .078 .065 .063 .072 .056 .050 .077 .120 CNetal products n.a. n.a. n.a. .053 .047 .058 .057 .053 Non-electric machinery n.a. n.a. n.a. .514 .467 .474 .509 .550 22-26 lElectric machinery n.a. n.a. n.a. .147 .144 .207 .178 .171 lTransport equipment n.a. n.a. n.a. .218 .176 .196 .216 .215 27 Miscellaneous .153 .169 .170 .173 .162 .171 .186 .209 Total .105 .105 .100 .100 .099 .102 .114 .131 * Includes sugar Source: Nacional Financiera, 1977. Table 1.2 Mexico: Manufactured Exports Not Including Assembly Plants: 1965-1976 (Million dollars) Ist Hal-f 1965 1970 1971 1972 1973 1974 1975 1975 1976 Food products ar '......rn5 ages a/ 42.0 53.6 70.7 77.0 105.2 142.9 116.3 66.2 80.0 Twftiles, clothirn,- -= footwear 28.8 38.3 50.8 77.7 167.7 247.8 141.1 74.8 80.0 Che icals 43.5 81.4 90.0 102.4 152.3 259.0 204.2 106.3 107.3 I>.chinear and tr^-=- cr, equipmentY' 5.8 61.0 75.3 106.0 241.4 249.3 269.8 126.6 126.6 Iron and Steel 23.8 33.6 57.9 67.6 37.1 50.0 38.1 19.9 16.8 Printed books 5.7 17.4 11.8 12.8 18.0 23.9 23.0 11.9 15.9 Glass and glass n--W-cts 4.8 8.5 13.3 18.3 21.6 25.9 25.9 6.1 17.2 Other manufactures- 48.6 59.7 84.4 113.1 94.6 243.5 250.9 118.6 14 I 203.0 353.5 454.2 574.9 837.9 1,242.3 1,069.3 530.4 590.6 Excludes sugar but includes tabacco. Includes electronics and parts. c/ Excludes primary nonferrous metals and petroleum refining. Source: Banco de Mexico, Indicadores Econ6micos, varios issues; data from 1974 on are provisional, Table 1.3 MEXICO: Exports as Ratios to Gross Value of Domestic Production 1967-1974 S e c t o r 1967 1968 1969 1970 1971 1972 1973 1974 la Petroleum extraction and refining and coal products .036 .029 .031 .026 .020 .012 .012 .040 lb Basic petrochemicals - - - - - - - - 2 Meat products .049 .070 .075 .073 .067 .076 .062 .028 3 Flour milling and baking - - - - - - - - 4 Other food products* .119 .122 .115 .109 .106 .116 .126 .125 5 Beverages .004 .004 .004 .004 .005 .008 .009 .011 6 Tobacco .031 .024 .040 .046 .048 .054 .081 .080 7 Textiles .033 .024 .026 .022 .028 .038 .063 .075 8 Clothing, including footwear .006 .006 .007 .010 .010 .015 .012 .018 9 Wood products .031 .036 .043 .038 .050 .066 .052 .056 10 Paper products .043 .041 .042 .041 .034 .036 .031 .031 11 Printing and Publishing - - - - - - - - 12 Leather and products .014 .013 .017 .014 .009 .010 .004 .008 13 Rubber and products .008 .017 .023 .027 .041 .041 .027 .030 14 Basic chemicals .100 .088 .096 .093 .092 .096 .118 .171 15 Synthetics and plastics - - - - - - - - 16 Fertilizers and insecticides .016 .022 .048 .078 .081 .105 .0,,9 .107 17 Perfumes and soaps .027 .024 .020 .017 .018 .017 .018 .022 18 Pharmaceuticals .109 .099 .098 .082 .089 .067 .086 .110 19 Other chemicals .033 .028 .024 .030 .021 .028 .035 .057 20 Non-metallic mineral products .021 .027 .027 .025 .036 .044 .041 .039 21 Basic metals .022 .032 .042 .032 .052 .058 .029 .030 'Retal products n.a. n.a. n.a. .009 .010 .014 .015 .019 2 Non-electric machinery n.a. n.a. n.a. .043 .054 .059 .094 .102 2 Electric machinery n.a. n.a. n.a. .008 .017 .017 .020 .027 Transport equipment n.a. n.a. n.a. .032 .037 .042 .060 .059 27 Miscellaneous .024 .027 .028 .033 .040 .040 .048 .055 Total .036 .037 .040 .039 .043 .045 .045 .050 * Includes sugar Source: Nacional Financiera, 1977. Tab le 1.- roxico's ma.r.of.et'roh : , As3rrhJor Pso,s) bLEleati..r ti8 99 SITC Other Otho 0Qe. v . r 2< Code Des-t. -otIon 3m.?> 'J2.t lfA3, Io ..it 4: rrv3ci LAPOF Lti .... -, o,, . - 5 Chemicals 16'i 414.0 32>9 3? 1o6 3sl ? 4233 21i 2,; C cl 6, Manufactvrnes Ol 'seified hainly by mirnus 68 nattrisal2 to. oo4rxoos l4etal2,, I5> 5 2Zc'J 93 10e 32,9 21 . 2 7 !flI4aca.inery *-'- :ozansport Eq'JiAP'nt 6720 '7. 0 ." ,7 '? 18 loS 14533 U,.> ZO 4 8 MisootL'o"rv ,rf.tro33 . 3- 4 i. 102 3i.7 25Go 6ri S8, Total -3 6?, 10140,?7 0I3 F 35.3 266 i," 31,? . 25 7,- Las68 P.'Žr,?,:oeot,P&I, Dit.rThotio'± <(3 , 74 7 2. Lr9> 0<2 J il1 31,N3I?Aj wq )3Qx, , ..i '. .< L 4 ,.i 65 Scepsa, 0tnfvM¢ ."" n 7 3 0 021 ? (,14 56 Fartili%er 35,3 634 .5 7 3.4 59 03lc o- fo'- 'flr0f'0<' ? 414 .41 4 2 (2 I 4., 61 .t.13',i I. (13 . 522 5o 0.5 71 5' C) 2 1~~~~~~~~~~~~~ 9( O.& P8,0 6: 0' 3'4 72 Eloctrico?. ":h-r', , Occtro.si $ 2.. 395.3 , e5 l e i 73 Trar.port Vuipussot . 03.7 73.7 y'7 3y'7 2Ž4 0Oo 7'38 930 0 33 81 Plumbing avld Mghting FBqipment 31432 13.4 0.6 0,1 (ol 82 Furniture 14.9 14.2 0.1 0.1 0.4 0,1 83 Travel Goods and Handbags 13.4 12.3 0,1 0.9 0.1 84 Clothing 106.4 99.6 3.3 0,1 2.2 0.7 0.1 0.4 85 Footwear 15.4 14.5 0.8 0.1 86 Instrwments and Film 18.4 8.7 0.2 0.1 8.6 - 0.8 89 Miscellaneous Manufactures 99.1 68.9 2.8 0.8 2.9 2.0 16.4 5.0 0.3 Note: Data for developed countries are imports from Mexico, from World Trade Annual. Supplement, volume 2, United Nations. Data for other countries are exports by Mexico, from Cosdity Trade Statistics, United Nations. Resulting totals are somewhat higher than those in Mexican export statistics adjusted for assembly industry exports. kiLU; Public Sector Enterprises in Manufacturin. - Iverteat acd F , anc,ia Perior-sance I N V E S T M E N T P E R F 0 R M A N C E Gross Fixed Investment Loss (-) or Profit (t) Loss (-) or Profit C+) to net worth to total assets Debt/Equity Real Growth 1971 1972 1973 1974 1975 ger year 1973 1974 1975 1973 1974 1975 1973 1974 1975 (million pesos) (percent) Ipercent) 1. Petrochemicals a! 354 436 541 795 1,760 32.6 2, Fertilizers 236 189 200 279 663 14.9 7.3 4.3 5.9 2.9 1.4 1.6 1.57 2.02 2.66 3. Subtotal-Chemicals (590) (625) (741) (1,074) (2,423) (25.2) 8.6 7.1 11.5 3.6 2.6 3.5 1.41 1.77 2.26 4. Steel 636 397 1,794 3,992 7,944 66.9 b/ 4 7 5.3 1.8 2.1 2.1 o.6 1.21 1.55 2.15 5. Sugar & other food & beverages 241 590 1,187 1,397 2,349 56.9 -375 .621.0 -18.9 -18.5 211.3 -10.9 1.03 0.86 0.74 6. Pulp, paper, wood products 73 83 208 428 757 59.5 n.a. n.a. n.a. 7. Transport equipment & machinery 240 411 422 676 1,216 33.2 -51.2 -86.5 -49.1 -4.8 -5.3 -4.4 9.80 15.41 9.93 8. Others 271 461 201 369 427 -o.6 b. a3 0 -2.0 -2.8 -1.1 -0.6 -0.8 1.80 2.21 2.64 A. Total 1,859 2,567 4,553 7,936 15,116 &(_33 -2.6 -4.3 -1.1 -0.7 -1.0 1.95 2.72 3.24 B. Total in 1970 prices 1,793 2,408 3,687 5,249 9,o46 49.9 Year-to-year grolwth (percent) 34.3 53.1 42.4 72.3 C. Total investment in manufacturing (in 1970 prices) 9,047 8,056 7,729 8,898 n.a. D. Total public investment 28,112 38,217 46,848 53,985 56,127 18.9 (in 1970 prices) E. Total private investment 51,524 50,660 52,829 54,418 54,796 1.6 (in 1970 prices) F. Total investment (private & 79,636 88,877 99,677 108,403 110,923 8.6 public) (in 1970 prices) Relative Shares (percent) Public investment in manufacturing/ total public investment (B/D) 6.4 6.3 7.9 9.7 16.1 Public investmnent in msmafactm'tng/ total investment in manufacturing(B/G 19.8 29.9 47.7 59.0 n.a. Total public investment/total investment (D/F) 35.3 43.0 47.0 49.8 50.6 Public investment in manufacturing/ total investment(B/F) 2.2 2.7 3.7 4.8 8.2 / Investent in expansion of oil excluded b/ Including SICARTSA c/ Excluding SICARTSA Source: Mission's calculations on the basis of information received from the Ministries of Presidencia and Patrinonio Nacional. Table 1. 6 MEXICO: Foreign ITvestment Enterprises in Manufacturing - Investment and Performance ____________ _ LNVE;, v:in mesilionpes", cwrrent prices) P E R F 0 E M A N CE -/ 1971 1972 1973 197ll 1975 Public 1,859 2,567 4,553 7,936 n.a. a. Tax Contribution (1973) Private 7,523 6,021 4,993 5,518 n.a. Poreign investment taxes paid or withheld (million pesos) 5,271 Total manufacturing taxes paid or withheld (million pesos) 20,297 Total 9,382 8,588 9,546 13,454 15,472 Share of foreign enterprise in total taxes by manufacturing Foreign private (1,838) (1,779) (2,670) (3,396) (3,734) b. Exports (1974) Foreign All Enterorises S/o enare of foreign - as share of the total (9) 19.6 20.7 28.0 25.2 24.1 enterprise enterprise foreign particip. total (%) - as shore of the private (S) 24.4 29.5 53.5 61.5 n.a. Gross sales (billion pesos) 52,219 324,087 271,868 16.1 Exports (billion pesos) including maquiladoras 10,530 b/ 28,440 17,910 37.0 excluding maquiladoras 3,090 15,536 12,446 19.9 Sectoral distribution of fixed capital ( a) s of 1975 (US$ million) Foreign Investment All Investment including mquiladoras 842 2,275 1,433 excluding maquiladoras 287 1,285 998 Agriculture 0.2 Share of exports in sales (%) Mining 6.3 including maquiladoras 20.2 8.8 6.6 excluding maquiladoras 5-9 4.8 4.6 Manufacturing 75.1 100.0 100.0 Of which: c. Le7mnt (1973) Share of foreig Foreign enterprise All enterprises enterprise in Food & beverages 5.8 7.2 19.2 (manufacturing) (manufacturing) total (t) Textiles, shoes, clothing 2.8 3.7 9.5 Total employmet (thousands) 401.9 2,550 15.8 of which - manual workers (260.4) n.a. n.a. Wood prod. furniture o.6 0.8 1.7 Total wage bill (billion pesos) 17.6 75.9 23.2 Paper, printing 3.3 4.4 7.7 Average income (in US$ equivalent) c/ Ieather, rubber 2.1 2.9 3.9 per capita of manual worker 3,508 2,769 d/ Chemicals 19.2 25.6 16.6 per capita of white collar 5,294 6,829 e/ Non-metalic minerals 1.9 1.4 9.3 d. Productivity (1970) Hlto 9 Basic metals 2.6 1.9 17.6 Value added (billion pesos) 21.8 94.6 23.0 Non-electrical machinery 5.0 6.7 3.0 Valu added per worker (pesos) 69,870 43,40o 161.0 Electrical machinery 10.3 13.8 2.9 Transport equipment 11.3 15.0 7.2 Others 10.1 13.5 1.4 Comnerce 11.4 Transport 0.1 Services 6.9 Total 100.0 7 "Estimates made on the basos of different, often unrelated ntudies. b/ Figure includes also some minor exports from sectors othec dhon manufacturing. c/ Includes services. 3 Mexico City only. Sources: Estimated by the Mission on the basis of the following sources: The_Impact of MultinationalJCounoration3 gIlont nod Incone - Ihe Cas nj , by Vicior Manuel Bcrnal Sahagnn and Angeli-a Gutierrec AriS aer-i Serr4orio 0O14edo 0-att&(Sn Insttiuto de Investigaoiones Economicas, Universidad Nacinnal Autnonoa de Mexico, ILO, Geneva, 1976. The Is ct of oreipgn Private Invesoment on the Mexican Econoo0'j, by Harry Rohinson and Timothy Smith, Stonford hesearch Institutc, Mlenlo Park, Cii. 1973, LaE snreEag t sion tl Vpd;Ly noe on en_l _ i a Mex_cana, by r,'naondo Fajozylber & Trinidad dMatinen foarago, c . - in Cu nrn Economica, Mexico, 1976. La nversi_Extron54ro Mexico, by Bernardo Sepulveda & Antonio Ch-macero, Fond. da Cultura Econonica, Mexico, 1973. onmercio Exterior, Mexico, Suplemento del Vol. 26, Hm. 7, Julio de 1976. Information received from the Direc-ion General del Registro Nacional do Inver-ionen Extranjeran in COt. -No,. 1976. Table 1.7 Indicators of Financial Development (1973X Non-monetary Deposits Non-monetary Deposits as percentage of GDP as percentage of money Country Supply Japan 88 244 Germany 57 399 United States 55 252 Singapore 42 -163 Taiwan 37 175 Korea 30 206 Mexico 27 208 Venezuela 17 109 Pakistan 14 46 Argentina 13 61 Philippines 10 101 Ecuador 8 60 Colombia (1974) 6 44 Bolivia 6 32 Haiti 5 46 Cameroon (1970) 3 21 Dahomey (1970) 2 -10 Source: Calculated from International Financial Statistics. Definition of non- monetary deposits: Country - line no. from IFS Japan - 35, 45, 46a; Germany X 35, 36a; United States - 35, 45; Singapore 35, 45; Taiwan - 35, 45; Ecuador - 35. 46a; Bolivia - 35, 45; Haiti - 35; Cameroon 35, 45; Dahomey - 35, 45; Philip- pines 35, 46. Table 1.8 Mexico: Nominal & Reel Interest Rates Interest Rate Inflation Real Interest Year (highest deposit rate) rate 1969 10.00 3.5 6.2 1970 11.00 5.0 5.7 1971 10.08 5.4 4.4 1972 9.01 5.0 3.8 1973 9.51 12.1 - 2.3 1974 11.53 23.7 - 9.9 1975 12.00 15.0 - 2.6 1976 11.99 16.0 - 3.4 Average 1969-72 10.0 4.7 5.0 Average 1973-76 11.2 16.7 - 4.6 Table 1.9 Mexico: Monetary System Balances, 1970-76 (end-of-year stocks as percentage of GDP) Total Domestically Allocation of Total Credit Money Non-Monetary Financed Foreign Total To Public To Private Year Supply Deposits Credit Liabilities Credit Sector Sector (1) (2) (3) (4) (5) (6) (7) 1970 12 27 38 8 44 22 22 1971 12 29 41 9 47 24 23 1972 13 30 42 8 48 25 23 1973 13 27 40 10 47 25 22 1974 12 24 36 10 43 25 18 1975 12 25 37 11 46 27 19 1976 12 18 31 13 42 27 15 Source: Bank of Mexico through 1975; staff estimates for 1976. Minor items are omitted and therefore figures do not add across. However, columns (I)and (2) are the main determinants of changes in column (3); likewise columns (3)and (4) are the most important influences on column (5). Table l.lC Mexico: Private Banking System Income Statements, 1966-75 Financial Gross Labor Other Adminis- Year Income eFinancial Gross Lor al trative Expens- Net Year Incoe Expenses Spread Costs - es Spread Profits Profits (percent of average total assets) (a percentage of equity) 1966 10.5 5.7 4.8 2.3 1.7 0.8 0.8 16.8 1967 10.5 5.9 4.6 2.1 1.8 0.7 0.6 13.8 1968 10.7 6.1 4.6 2.0 1.8 0.8 0.8 18.7 1969 11.2 6.6 4.6 2.0 1.8 0.8 0.8 18.4 1970 12.1 7.6 4.5 2.0 1.8 0.7 0.6 14.6 1971 12.3 7.8 4.5 2.0 1.9 0.6 0.6 14.9 1972 11.7 7.2 4.5 2.0 1.7 0.7 0.5 13.9 1973 11.6 7.1 4.5 2.1 1.8 0.6 0.4 10.9 1974 12.5 7.4 5.1 2.4 2.0 b/ 0.7 0.2 7.0 1975 13.3 7.9 5.4 2.5 2.4 b/ 0.5 0.1 3.4 Source: Comisi6n Nacional Bancaria y de Seguros a/ Does not include workers participation in profits. b/ Acc'lminting prohh'ms rplato'd to taxes and portfolio losses may overstate ttiis item. Table 1.11 Mexico: LABOR COSTS, 1976 (Example: Baja California) Weekly Charge (pesos) (1) Minimum wage (122.80 per day, times 7 days) 859.60 (2) Housing tax 42.98 (3) Vacation bonus (122.80 x 15 . 52)a/ 35.42 (4) Education tax 8.60 (5) Social security tax (category "R") 75.47 (6) Profit sharing (approx. 859.60 . 52) 16.53 (7) Year-end bonus (122.80 x 30 , 52) 70.84 Tlqotal | 1,109.44-/ a/ 15 days of vacations; 52 working weeks. b/ Divided by 6 working days per week = Mex$ 184.91/day Excess over the 184.91 1.51 (51%) minimum wage 122.80 Source: Staff estimates. Table 1.12 Mqxico: Manufacturing Sector Wages Indices of Nominal Wages Indices of Real Wages Year Average Minimum Prices Average Minimum Wage Wage Wage Wage (in current prices) (in constant prices) 1960 100.0 100.0 100.0 100.0 100.0 1961 102.6 100.0 100.9 101.6 97.3 1962 112.8 120.7 102.8 109.7 117.4 1963 133.3 120.7 103.3 129.1 116.8 1964 143.6 148.3 107.7 133.3 137.7 1965 153.8 148.3 109.7 140.2 135.2 1966 161.5 172.4 111.1 145.4 155.2 1967 171.8 172.4 114.3 150.3 150.8 1968 179.5 194.8 116.5 154.1 167.2 1969 189.7 194.8 120.6 158.8 161.5 1970 197.4 220.7 126.6 156.0 174.3 1971 215.4 220.7 133.5 161.3 165.3 1972 228.2 259.5 140.1 162.9 185.2 1973 256.4 273.1 157.0 163.3 173.9 1974 328.2 368.1 194.3 168.9 189.4 1975 371.8 425.6 223.4 166.4 190.5 1976 500.0 a/ 558.0 a/ 258.7 193.3 a/ 215.7 a/ a/ IBRD staff estimates. Sources: Average wages from Banco de Mexico, unpublished estimates. Minimum wage: 1960-70: Urban wages in Federal District, from Secretariat of Industry and Commerce Statistical Yearbook. 1970-75: Arithmetic average for the nation, from Minimum Salary Commission. Prices: 1960-68: Wholesale prices in Mexico City, Banco de Mexico. 1968-76: National Consumer prices, Banco de Mexico. Table 1,13 Mexico: Average Wages in Manufacturing compared with the United States: 1960-1976 Year Mexico USA Mexican wages as percent- age of USA wages (dollars per hour) (percent) 1960-1965 .49 2.43 20 1966-70 .70 3.03 23 1971 .84 3.57 24 1972 .89 3.81 23 1973 1.00 4.07 25 1974 1.28 4.40 29 1975' 1.45 4.79 30 Dec. 1976 a/ 1.18 5.40 22 Source: Bank of Mexico and staff estimates. a/ salaries are preliminary estimates; exchange rate of 20.21 pesos per dollar. Table 1.14 Capital Goods Imports in Relation to Domestic Consumption in Argentina, Brazil and Mexico 1973 Argentina Brazil Mexico (Imports as percentage of internal demand) Metal products 2.4% 5.6% 5.7% Non-electrical machinery 19.4 29.7 45.6 Electrical Machinery 18.0 17.2 17.6 Transport equipment 3.8 5.6 21.6 Total 9.9 15.8 21.8 (billion dollars) Total Internal demand for machinery and equipment 2.3 4.4 3.3 Source: Nafinsa (1977), pp. 212,236. CHART I MEXICO: REAL EFFECTIVE EXCHANGE RATES FOR IMPORTS AND EXPORTS OF MANUFACTURES CASE A 27~~~~~~. 22 ~ ~ ~~~~~~~~~~~O EQUILIBeRIUM ~ 25 17MPORTSv~'EXPORTS 2470 Dec, 197 1980 0 O CASESB a 2 0~~~~~~~~~~~~~~ 24 _- I MO - ul 2 3 1 F FREE TRADE __ < 22 - 0 o EQUILBRIUM - - - - - fn -b/ - - - - - - - - - - - - - - - - - tu 201- ~__~~ ^ wx x~~ ~~~~~~~~~~~~~~~~~~~~~ ir is~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~. w EXPORTEXSRT 1970 1975 Dec. 1976 1980 Wordd Bank-1 7393 CHART 1. 2 MEXICO: SECTORAL DISTRIBUTION OF CREDITS FROM BANKING SYSTEM 80 70- Industry r5 0 40 4 30~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 Government 20 N.Commerc ........ ~~~~~~~Agriculture- - 10 I. . ..~~~~~~~~.......... Mining 0 - ~~~~ - - -~~ - - - S S-i - - S - 4* - 1950 1955 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 Source: Bank of Mexico World Bank-1 71 55 ANNEX II Page 1 Annex II Free Trade Equilibrium Exchange Rates 1. The object of this annex is to estimate the exchange rate at which Mexico's balance of payments on current account would have been in equilibrium in the absence of any taxes, restrictions, or incentives on imports or exports. 2. The main use of this "FTE" exchange rate is to permit comparisons of levels of protection and export taxes at different times. As is well under- stood, protection and/or export subsidies can substitute for devaluation (i.e., compensate for overvalution). Thus, if in a hypothetical case a country's exchange rate becomes overvalued by 50 percent (due, say, to domestic infla- tion), and the observed rate of protection increases by 80 percent, in one sense protection has increased only about 20 percent (1.8/1.5) because the rest of the increase is not really protection but rather a substitute for devaluation. (In an extreme case, observed protection in Brazil in 1957 was 890 percent: the average cost of a dollar's worth of imports was Cr $164.66 while the official exchange rate was Cr $18.50/dollar. Protection relative to an estimated equilibrium rate, however, was only 89 percent.) 3. In the present case, "equilibrium" will not be defined in terms of the current account only: i.e., imports precisely equal to exports. Rather, it will be recognized that Mexico, as a developing country, will normally run a current account deficit in order to receive net foreign savings. In the three years in question, Mexico's current account deficits were 11, 14, and 43 percent of exports. The first two values do not seem unreasonable, while the third does. "Equilibrium" in 1960 and 1970 is therefore defined as in- cluding the actual deficits in terms of percent of exports, while for 1975 the deficit is limited to 15 percent of exports. These deficits imply net capital inflows of $147 million, $403 million, and $945 million in the three respective years -- equal to 1.2 percent of GDP in each of the three years. 4. The essence of the present method is to use estimates of actual protection, export incentives, and price elasticities of demand and supply of imports and exports to calculate the exchange rate that would induce "equilibrium" in the current account if the protection and incentives were removed. For present purposes, this method is superior to estimating "over- valuation" by the amount of higher domestic inflation, because the latter ignores relevant changes in the structure of Mexico's production and trade. An estimate of overvaluation that is the same whether Mexico is a net importer or a net exporter of petroleum is not relevant to adjusting observed protec- tion for overvaluation -- even though it may be relevant for other purposes, such as estimating changes in living standards or changes in competitiveness of manufactured exports. 5. There is nothing in the estimated FTE exchange rate that makes it the right or best rate. In particular, its level depends on fiscal policy, growth of output and changes in economic structure. The correct rate depends on economic objectives and on what other measures are taken that affect domestic growth, demand for imports, and supply of exports. ANNEX II PaCr, 2 6. To derive the estimat.or. formula, define: M, M' = imports before and after the hypothetical move to free trade; X, X' = exports, similarly; eSM, eDM =price elasticities of supply and demand for imports; eSX, eDX = price elasticities of supply and demand for exports; r, r' - exchange rates (peso per dollar), before and after the hypothetical move to free trade; tM, tX taxes on imports and on exports (in oiLr case, t I will be measured by comparisons of domestic-anid international prices and for convenience,lefine the ratios of exports to imports: k = 1 x k' = MI Assume that the price elastidity of supply of imports is infinite. For exports, note that: _a= e A(pr) = e 2 (1) Q SX pr _DX p where Q = quantum of exports p = foreign price of exports From eq. Cl) an expression br the change in value of exports in foreign prices can be derived: A X CSx (eDX + 1) (r' - r) X eDX -e r (2) and therefore for convenience define: ex= esx (eDX+ 1) e -e eDX - SX ANNEX II Page 3 From the definitions of the elasticities: M' - M e r' - r (1 + M) ' = eM r(l +t) 1 (3) XI - X= eX [r - r(l - X)] 4 x x ~r(l - t~ j4 Dividing (3) by (4) gives: r' - r (1 + M) M' - DM r(l+ ,(5) ___ e t X *- X ex r r r(1 - X) x) Transforming the LHS of (5): MN k'X' X'___ ___X_ ____1 (6) - r t - 1 - 1 k l+ X r(l - Substituting the transformed LHS (6) into (5), and solving for r'/r gives: e /k') (l D\I (7) r eDM e x IC _____ x ANNEX II Page 4 7. Import protection estimates are taken from Balassa (1971) for 1960, from ten Kate (unpublished) for 1970 and from preliminary Mexican govern- ment estimates for 1975. For the elasticity estimates, use the mean values of two sets of estimates made by Gerardo Bueno: GB1 GB2 Nean M -3 -1 -2 SX 3 3 3 eDX -10 -4 -7 Source: GB1: Balassa (1971). GB2: Bueno (1974). In this article Bueno does not settle on one set of elasticities; the ones used here are those he seems to think most appropriate. 8. Estimates of the "export taxes" (tx in equation 7) have been made in a more complex way than is usual. The calculations for 1970 and 1975 are shown in Table 2.1. In addition to taking CEDIs and export taxes into account (lines 2 and 4), adjustments for a number of other incentives and disincentives are made. The general principle has been to try to include all government measures that affect after-tax profits in export sales, relative to what they would be in a free trade situation. 1/ The effects of each measure are converted into an equivalent percentage change in before-tax, taxable, export revenue; this enables direct comparison with each other and with the effects of a change in the exchange rate. 1/ Most of the estimates do not pretend to be more than approximations; these are preferred to the larger error of simply ignoring the effect completely. Firmer estimates would require better data. ANNEX II Page 5 9. The methodology can best be explained line by line: Line 1: For most measures, estimates are made separately for exports of primary products, secondary products, and services. For services incentives are assumed to be zero, so no figures are shown. (Actual incentives for ser- vice exports, other than the exchange rate itself, are small and in opposing directions.) For the others, the weights shown are their share in total exports. Assembly industries are treated here as services, since most of the measures do not apply to them in any significant way, if at all. Line 2: Typical value of 11 percent is used. Line 3: Assumes that two-thirds of CEDIs go to firms that pay income tax (the rate for which is taken as 42 percent). Line 4: Estimates of actual receipts: SHCP. Allocated 100 percent to primary products. Line 6: For manufactures in 1975, Balassa's 1973 estimate of 1.5 percent was increased to 2 percent to account for the greater spread between FOMEX rates and market rates. For primary exports and for manufactures in 1970 (before FOMEX existed), 0.5 percent is assumed. Line 7: Measures here include price controls on inputs, QRs on exports of inputs, and the somewhat easier access to import permits when the imports are to be used to pro- duce exports. It is estimated that these measures reduce input costs by 2 percent, on average, and that input costs are 62 percent of gross value of output for manufactures (1975 Census) and 35 percent for primary goods (Staff estimates). These input cost ratios are used in lines 9 and 11 as well. Line 8: This is simply the sales tax, taken as 4 percent. Line 9: Assume the sales tax, on average, is applied 1.5 times to inputs. The shares of inputs in output are as explained above under Line 7. Line 11: Here the problem is not only to estimate the effects of protection on input prices, but also to net out the as-yet-uncalculated effect of overvaluation. (Over- valuation must not be counted here because under free trade, the exchange rate would increase by the amount of over- valuation and this would increase input costs.) Overvalua- tion was estimated by iteration until the assumption was ANNEX II Page 6 consistent with the result; only the final iteration is shown. For the level of protection on inputs (relative to the official exchange rate), economy-wide levels of protec- tion of 13 and 28 percent in the two years were used. Line 12: This algebraic total shows the estimated net effects of all incentives and disincentives, except over- valuation. Economy-wide these were -6.6 percent in 1970 and -4.0 percent in 1975. These are the figures used for t in equation 7. Line 13 is the estimated overvaluation as calculated by equation (7), including the restrictions of the current account deficit to 15 percent of exports in 1975. Thus the estimated free trade equilibrium exchange rate would have been about 12.85 in 1970 -- only 2.8 percent above the actual -- and about 14.80 in 1975; 18.4 percent above the actual. 10. These estimates of overvaluation are smaller than most others. The two main reasons are: (a) The effect of protection on raising the cost of export production is quite significant as a share in total export "taxes" -- about half in 1970 and one-third in 1975. Taking this effect into account implies that moving to free trade would increase exports more than would otherwise be estimated, and thus a lower devaluation is required to maintain (or restore) balance of payments equilibrium. (b) Any estimate for Mexico based on import and export elasticities (as these are), will give a lower estimated of overvaluation (for 1975 at least), than will an estimate based on purchasing power comparisons using some time earlier as a base. (See, e.g., Villarreal, 1976.) This is because Mexico handled her balance of pay- ments in part by import substitution and export growth (e.g., petroleum and products; tourism). Given these events the exchange rate did not have to keep up completely with relative price changes to maintain equilibrium. Nevertheless, failure to keep the real exchange rate constant in terms of relative prices implies decreasing competitiveness for given export products. 11. For 1960 the data needed to follow the same method are not avail- able. Bueno's estimate of overvaluation in 1960 was 9 percent. Comparing the present estimate of 3 percent for 1970 to an earlier estimate of 6 per- cent for the same year using a method similar to Bueno's suggests that an estimate for 1960 of 6 percent would be not too far off, and consistent with the present method. (1.03 x 1.09 = 1.06) (1.06 12. Thus overvaluation in 1960, 1970, and 1975 is estimated at 6, 2.8, and 18.4 percent respectively. ANNEX II Page 7 13. This implies that the net effects of protection on manufacturing were 17.9, 15.6 and 19.0 percent in the three years considered; the apparent rise in protection in the 1970's has compensated for a roughly equal increase in overvaluation. On the export side, there is no estimate for net export taxes in 1960. For 1970 and 1975 they are estimated at 14.4 and 9.5 percent, respectively. (Again this is for manufactures.) Thus the introduction of CEDIs had a significant positive effect, but was insufficient to make up for the overvaluation and other disincentives. 14. In the estimates for December 1976, reported in Text Tables 3 and 4, a different method was used. Instead of using import and export taxes and elasticity estimates to calculate a FTE exchange rate, the FTE rate was assumed to remain constant in real terms from its 1975 level. In nominal terms it therefore is 44 percent above the 1975 level, reflecting the amount by which wholesale prices in Mexico increased relative to wholesale prices in the US. Net export incentives to manufacturers were then calculated as in Table 2.1, the numerical differences from 1975 being that in December 1976 CEDIs were zero (lines 2 and 3 in Table 2.1), the effect of protection on input costs, net of overvaluation, was -.048 (line 11), and the resulting net incentives other than overvaluation were -.093 (line 12), also shown in line 3 of Text Table 3. ANNEX IT Page 8 Table 2, 1 Mexico: Export Incentives and Disincentives, and Estimated Overvaluation, 1970 and 1975 Item 1970 i 1975 Primary Secondary Average Primary Secondary Average 1. Weights for Sectors .340 .096 1.000 .239 .191 1.000 (ratios o exports) 2. CEDIs .110 3. Adjustment for non-taxability .053 4. Export taxes -.079 -.151 5. Sub total -.079 .000 -.027 -.151 .163 -.005 Adjustments: 6. Low interest export financing .005 .005 .002 .005 .020 .005 7. Reduction of input costs; better access to imports .007 .012 .004 .007 .012 .004 8. Sales tax on export sales -.040 -.040 -.017 -.040 -.040 -.017 9. Effect of sales tax on input prices -.021 -.037 -.011 -.021 -.037 -.012 10. Sub total, all incentives above -.128 -.060 -.049 -.200 .118 -. 11. Adjustment for effect of protection on input costs,net of overvaluation. (Assume overvaluation of 3% in 1970 and 19% in 1975) -.034 -.060 -.017 -.026 -.047 -.015 12. Net export incentives, excludirg overva luation, -.162 -.120 -.066 -.226 .071 -,040 13. Estimated ~r' 1.184 13. Estimated overvaluation,- 1.028 r Incentives are expressed here as ratios to export revenue at actual prices. In the text, however, net incentives have been converted to percentages of what export revenues would have been in the free trade equilibrium situation. ANNEX III Page 1 Mexico's Industrial Structure 1. The structure of production (by type of product) of Mexican manu- facturing, and its level relative to the size of the country, are essentially similar to other countries in more or less similar circumstances--at least at the 2-digit level of aggregation. Mexico's industrial structure and output per capita is strikingly similar to that of Brazil, while both countries are somewhat less industrialized than Spain but rather more than Turkey, which were chosen as countries roughly similar to Mexico in size and level of devel- opment (see Tables 3.1, 3.2, and 3.3). 1/ 2. In labor productivity, however, Mexican manufacturing was behind both Brazil and Turkey in 1970, and more similar to Spain (see Tables 3.4 and 3.5). However, Mexico's labor productivity rose considerably between 1970 and 1975, as measured by the industrial census. 3. Comparison to an international pattern based on income level, popu- lation, and economic structure shows that in 1970 the Mexican manufacturing sector was smaller than the predicted level; the only subsectors above their expected values were food, beverages, and tobacco; chemicals and petroleum products; and metal products: 2/ Similar analysis for 1975, using preliminary results of the 1975 census, shows manufacturing again below its expected value; those groups above predicted levels are (again) food, beverages, and tobacco; and chemicals and petroleum products. Metal products, however, dropped to 5 percent below its predicted size (see Table 3.6). Thus deviations from the international pattern increased from 1970 to 1975; the sector most above its predicted level in 1970 is also the one that grew most relative to its predicted growth (food, beverages, and tobacco), and two of the other three sectors above or near predicted values in 1970 ranked second and third in relative growth (chemicals and paper). The exception is metal products, which fell from slightly above to slightly below its predicted level. All sectors below predicted levels in 1970 were even further below 1/ The 1975 population and income per capita of the four countries are: Mexico Brazil Spain Turkey Population (millions) 60 107 35 40 GDP per capita (dollars) 1,190 1,010 2,700 860 2/ Comparisons are to values predicted from equations estimated by Chenery and Taylor (1968). Intepretation of the total figures is in doubt because ISIC sector 39, miscellaneous manufacturing, is not included, and also because of questions as to the appropriate exchange rate. Comparisons among 2-digit sub-sectors or over time, however, are less subject to these problems. ANNEX III Page 2 predicted levels in 1975. The rapid growth in food products was a response to Mexico's natural advantages in these lines, and probable strong growth in domestic final demand. Growth in chemicals and in paper also responded to natural advantages, but in these sectors import substitution and export markets were the major sources of growth in demand. 4. There seems to be a relationship, albeit a weak one, between labor productivity and growth, and also between labor productivity and Mexico's deviation from the international pattern. Sectors 27 (paper and products) and 31-32 (chemicals, and chemical petroleum and coal products) especially, show high labor productivity, high growth rates, and are large relative to other Mexican sectors when compared to values predicted by the international cross-section. At the other extreme, sectors 23 (textiles), 24 (clothing), 28 (printing and publishing) and 29 (leather and products) show low labor productivity, low growth in output, and large negative deviations from inter- national patterns. These differences are also shown, to some extent, in the export share of each sector (see Table 1.3); chemicals, petroleum refining and wood products export more of their production than mexican manufacturing on the average, while clothing and leather export about average levels. (The textile sector, with an above-average export share, is not consistent with this pattern.) This tendency, as well as the export figures in general, suggest that Mexican manufacturing performs well in products that are capital intensive, high-technology, and/or based on mineral resources, while it performs less well in traditional, low-technology, labor-intensive processes. This is similar to the conclusions of an earlier study, that Mexico's manufac- tured exports are concentrated in mature, high-technology sectors (Boatler, 1975). However, this same pattern is also consistent with the reasoning of this report that protection inhibits exports (hence the relatively good export performance of sectors not dependent on imported inputs) and that more gener- alized export incentives could result in increased exports of more labor- intensive products. ANNEX III Table 3.1 Page 3 MEXICO - Value added in Manufacturing (million 1975 dollars) 1965 1970 1975 Sector (ISIC) Value Percentage Value Percentage Value Percentage Annual Growth Added distribution Added distribution Added distribution Rate 1965 - 75 20 - 22 Food, beverage, and 1,380 Tobacco 22.7% 2,213 22.7% 2,906 22.7% 7.7% 23 Textiles 574 9.4 773 7.9 818 6.4 3.6 24 Clothing, footwear, and made-up textiles 319 5.2 424 4.3 396 3.1 2.2 25 - 26 Wood products and furniture 140 2.3 245 2.5 317 2.5 8.5 27 Paper and paper products 211 3.4 325 3.3 459 3.6 8.1 28 Printing and publishing 216 3.5 325 3.3 345 2.7 4.8 29 Leather and leather products, except wear- ing apparel 31 0.5 54 0.6 60 0.5 6.8 30 Rubber products 120 1.9 221 2.3 261 2.0 8.1 31 - 32 Chemicals and chemical petroleum and coal products 881 14.4 1.442 14.8 2,057 16.0 8.8 33 Non-metallic mineral products 331 5.4 515 5.3 675 5.2 7.3 34 Basic metal 522 8.5 865 8.9 ' 1,259 9.8 9.2 35 - 38 Metal products 1,303 21.3 2,219 22.7 3,065 23.9 8.9 39 Miscellaneous 91 1.5 137 1.4 203 1.6 8.4 20 - 39 Total 6,128 100.0% 9,758 100.0% 12,821 100.0% 7.6% Source: Industrial Censuses. ANNEX III Page 4 Table 3.2 Industrial Structure in Brazil, Turkey and Spain, 1970 (million 1975 dollars) BRAZIL TURKEY SPAIN Sector (ISIC) Value Percentage Value Percentage Value Percentage Added distribution Added distribution Added distribution 20 - 22 Food, beverage, and tobacco 2,920 17.1% 1,274 32.6% 1,275 13.5% 23 Textiles 1,595 9.3 757 19.4 870 9.2 24 Clothing, footwear and made-up textiles 571 3.3 14 0.3 478 5.0 25 - 26 Wood products and furniture 788 4.6 31] 0.8 521 5.5 27 Paper and paper products 437 2.6 58 1.5 301 3.2 28 Printing and publishing 628 3.7 52 1.3 296 3.1 29 Leather and leather products, except wearing apparel 110 0.6 8 0.2 139 1.5 30 Rubber products 333 2.0 51 1.3 235 2.5 31 - 32 Chemicals and chemical petroleum and coal products 2,871 16.8 753 19.3 1,576 16.6 33 Non-metallic mineral products 1,004 5.9 156 4.0 795 8.4 34 Basic metal 1,974 11.6 313 8.0 241 2.5 35 - 38 Metal products 3,483 20.4 426 10.9 2,699 28.5 39 Miscellaneous 360 2.1 15 0.4 46 0.5 20 - 39 Total 17,074 100.0% 3,908 100.0% 9,472 100.0% Source: United Nations Yearbook of Industrial Statistics, 1974 Edition; International Financial Statistics. ANNEX III Page 5 Table 3.3 Value Added Per Capita in Mexico, Brazil, Turkey and S,pain, 1970 (US dollars per person) Sector (ISIC) MEXICO BRAZIL TURKEY SPAIN 20 - 22 Food, beverage, and tobacco 43.67 31.47 36.16 37.90 23 Textiles 15.26 17.19 21.48 25.85 24 Clothing, footwear and made-up textiles 8.37 6.15 0.40 14.20 25 - 26 Wood products and furniture 4.84 8.49 0.88 15.49 27 Paper and paper products 6.41 4.71 1.65 8.95 28 Printing and publishing 6.41 6.77 1.48 8.80 29 Leather and leather products, except wearing apparel 1.07 1.18 0.23 4.13 30 Rubber products 4.36 3.58 1.45 6.98 31 - 32 Chemicals and chemical petroleum and coal products 28.46 30.95 21.37 46.84 33 Non-metallic mineral products 10.16 10.82 4.43 23.63 34 Basic metal 17.07 21.28 8.88 7.16 35 - 38 Metal products 43.79 37.55 12.09 80.21 39 Miscellaneous 2.70 3.88 0.43 1.37 20 - 39 Total 192. 184.06 110.93 281.53 Source: Secretaria de Industria y Comercio - Censo Industrial 1971, Mexico. United Nations Yearbook of Industrial Statistics - 1974 Edition. World Bank Atlas, 1972. ANNEX III Page 6 Table 3.4 MEXICO - Value Added per Worker (thousands 1975 dollars) Percent Increase Sector (isic) 1965 1970 1975 1965-75 20 - 22 Food, beverage, and tobacco 4.14 5.83 6.94 67.6% 23 Textiles 3.41 5.00 5.82 70.7 24 Clothing, footwear and made-up textiles 2.94 3.40 3.22 9.5 25 - 26 Wood products and furniture 2.19 3.24 3.89 77.6 27 Paper and paper products 6.84 8.65 10.65 55.7 28 Printing and publishing 4.36 5.71 6.47 48.4 29 Leather and leather products, except wearing apparel 2.96 4.51 6.00 102.7 30 Rubber products 6.83 9.95 13.18 93.0 31 - 32 Chemicals and chemical petroleum and coal products 8.06 10.06 12.36 53.3 33 Non-metallic mineral products 4.30 5.64 6.45 50.0 34 Basic metal 10.37 12.36 15.89 53.2 35 - 38 Metal products 4.37 6.81 7.11 62.7 39 Miscellaneous 3.61 44.86 5.41 49.9 20 - 39 Total 4.56 4.09 7.50 64.5% Source: Industrial Censuses. A"NNEX III Page 7 Table 3.5 Value Added per Worker in Brazil, Turkey and Spain, 1970 (thotisasnd 1975 dnllars) Sector ISIC BRAZIL TURKEY SPAIN 20 - 22 Food, beverage, and tobacco 8.51 10.70 4.65 23 Textiles 5.06 6.30 4.10 24 Clothing, footwear and made-up textiles 4.02 2.80 3.88 25 - 26 Wood products and furniture 4.12 3.10 3.72 27 Paper and paper products 7.66 5.27 6.27 28 Printing and publishing 8.72 5.77 4.70 29 Leather and leather products, except wearing 1.52 2.66 6.61 apparel 30 Rubber products 11.89 5.66 6.52 31 - 32 Chemicals and chemical petroleum and coal 18.64 20.91 9.43 products 33 Non-metallic mineral products 5.45 4.33 4.94 34 Basic metal 8.69 6.02 2.34 35 - 38 Metal products 9.02 3.00 4.87 39 Miscellaneous 7.05 5.00 3.83 20 - 39 Total 7.86 7.04 4.95 Source: United Nations Yearbook of Industrial Statistics, 1974 edition. International Financial Statistics. ANNEX III Table 3.6 Page 8 MEXICO - Actual and Predicted Valued Added in Manufacturing - 1970 and 1975 1970 1975 Actual as Actual as Actual Predicted percentage Actual Predicted percentages deviation deviation above (+) or above (+) or Sector (ISIC) below (-) below (-) predicted predicted (million pesos) (percent) (million pesos) (percent) 20 - 22 Food, beverages and tobacco 17,948 15,721 14 43,110 36,299 19 23 Textiles 6,266 13,187 -53 12,126 35,141 -65 24 Clothing and footware 3,441 8,237 -58 5,869 22,159 -74 25 - 26 Wood products 1,985 5,511 -64 4,698 15,188 -69 27 Paper and paper products 2,633 2,650 - 1 6,803 6,882 0 28 Printing and publishing 2,633 6,496 -59 5,109 17,253 -70 29 Leather products 434 853 -49 895 2,102 -57 30 Rubber products 1,793 3,703 -52 3,869 9,240 -58 31 - 32 Chemicals and petroleum coal products 11,697 10,834 8 30,507 27,872 9 33 Non-metallic mineral products 4,180 6,647 -37 10,011 17,355 -42 34 Basic metals 7,019 11,767 -40 18,677 34,306 -45 35 - 38 Metal products 17,996 16,969 6 45,449 47,720 - 5 20 - 38 Total excluding miscellaneous 78,025 102,575 -24 187 123 271,517 -31 Sources: Actuals: Mexican Industrial Censuses; preliminary results for 1975 Predicted: Staff estimates based on Chenery and Taylor (1968) References Aspra, L. Antonio, "Import Substitution in Mexico: Past and Present," World Development, January-February 1977. Balassa, Bela, et al, The Structure of Protection in Developing Countries, Johns Hopkins, 1971. Bergsman, Joel, "Commercial Policy, Allocative Efficiency and 'X-Efficiency,' Quarterly Journal of Economics, August 1974. Boatler, Robert W., "Trade Theory Predictions and Growth of Mexico's Manufactured Exports," Economic Development and Cultural Change, April 1975. Bueno, Gerardo, "La Paridad del Poder Adquisitivo y las Elasticidades de Importacion y Exportacion en Mexico," El Trimestre Economico, April-June, 1974. Chenery, Hollis B. and Lance Taylor, "Development Patterns: Among Countries and Over Time," Review of Economics and Statistics, November 1968. Connor, John N. and Willard M. Mueller, "Market Power and Profitability of Multinational Corporations in Brazil and Mexico," Report to the Subcommittee on Multinational Corporations, United States Senate, June 1976 (revised). Economic Commission for Latin America, "La Exportacion de Manufacturas en Mexico y la Politica de Promocion," Version provisional, August 1976. Fajnzylber, Fernando and Trinidad Martines Tarrago, Las Empresas Transnacionales, Fondo de Cultura Economica, 1976. Gomez Palacio, Bernardo, "La Industria Metalmechanica y los Bienes de Capital en Mexico," El Mercado de Valores, November 29, 1976. Nacional Financiera, S.A., La Politica Industrial en el Desarrollo Economico de Mexico, 1971. Nacional Financiera, S.A., Mexico: Una Estrategia para Desarrollar la Industria de Bienes de Capital, 1977 Solis, Leopoldo, "Mexico" in Selection of Industrial Priorities, United Nations Industrial Development Organization, to appear. Vazquez Tercero, Hector, Una Decada de Politica Sobre Industria Automotriz, Editorial Technos, 1976. Villarreal, Rene, El Desequilibrio Externo en la Industrializacion de Mexico, 1929-1975, Fondo de Cultura Economica, 1976.