Mexico - Future directions of industrial strategy

Report No. 4313a-ME Mexico Future Directions of Industrial Strategy May 31, 1983 Projects Department Latin America and Caribbean Region FOR OFFICIAL USE ONLY Document of the World Bank This docunent has a restricted distribution and may be used by recipients only in the performance of their ofticial duties. Its contents nmay not otherwise be disclosed without WA/orld Bank authorization.  CURRENCY UNIT - PESO (MEX$) Since December 20, 1982, Mexico has the following exchange rates: (i) a sliding controlled rate for priority imports, most exports and service payments on foreign debt; this was set at 95 pesos per dollar on December 20 and was about 155 on May 20, 1983; and (ii) a free exchange rate for all other transactions; in recent days the free peso has been trading around 148 pesos per dollar. FISCAL YEAR January 1 to December 31 Principal Acronyms CEDI - Certificado de Devolucion de Impuestos (Export Tax Rebate Certificate) CEPROFI - Certificado de Promocion Fiscal (Investment Tax Rebate Certificate) FERTIMEX - Fertilizantes Mexicanos, S.A. (Mexican Fertilizer Corp.) FOMEX - Fondo de Fomento a las Exportaciones (Export Promotion Fund) GDP - Gross Domestic Product IMCE - Instituto Mexicano de Comercio Exterior (Mexican Institute for Foreign Trade) IMF - International Monetary Fund NAFINSA - Nacional Financiera S.A. (National Development Bank) PEMEX - Petroleos Mexicanos S.A. SECOM - Secretaria de Comercio (Ministry of Commerce) SEPAFIN - Secretaria de Patrimonio y Fomento Industrial (Ministry of National Patrimony and Industrial Development) SHCP - Secretaria de Hacienda y Credito Publico (Ministry of Finance and Public Credit) SIDERMEX - Siderurgia M,exicana S.A. SPP - Secretaria de Programacion y Presupuesto (Ministry y Programming and Budget)  FOR OFFICIAL USE ONLY MEXICO FUTURE DIRECTIONS OF INDUSTRIAL STRATEGY TABLE OF CONTENTS Page No. SUMMAR'Z OF CONCLUSIONS AND THE POLICY RECOMMENDATIONS ............... i-x I. THE MACRO-ECONOMIC SETTING ................ ... . . 1 II. MAIN ELEMENTS OF SECTORAL POLICY, 1971-1981 .............. 5 III. SUBSIDIES AND PROTECTION........ ..... ....... . 11 IV. INDUSTRIAL PERFORMANCE AND STRUCTURE ....................... 31 V. THE LABOR MARKET............,,.....1................. 56 VI. THE POLICY FRAMEWORK FOR FUTURE INDUSTRIAL GROWTR......... 76 STATISTICAL APPENDIX ............. .............. 87 Annex 1 Investment a-d EMloMent Incentives in the Mexican Industrial Sector. An Evaluation of their Possible Distortionary Effectse.a .............. ............ * a f 103 This report is based on the findings of a mission which visited Mexico in February 1982. The mission comprised Mr. I. Baskind, Ms. Judith Press, LCPI2; Mr. A. Pastor, LC1; Ms. M. Cortes, EAP (formerly DED); Messrs. A. Zabalza-Marti, D. Morawetz, and P. Ruof, Consultants. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   PREFACE This report is based on the findings of a mission which visited Mexico in February 1982. In June 1982, a preliminary outline was discussed with Government officials and, in view of the developing crisis, basic information was also updated. A draft of this report was discussed with the autlhorities of the new Mexican Administration in February 1983. The main focus of the analysis is on sectoral developments in the period 1970 through 1981 and the influence of both macro- and sectoral polLcies. The objective is to provide a framework for elaborating industrial deveslopment strategy in the medium term given the changed circumstances in which the country now finds itself. Measures adopted by the new Administra- tion after January 1983 are not fully reflected in this study. A country economic report now in preparation will review in more detail the adjustment policies adopted and currently in place.   SUMMARY OF PRINCIPAL CONCLUSIONS AND RECOMMENDATIONS 1. At present, Mexico is experiencing the mst severe economic and finamcial crisis in its modern history. GDP growth has virtually stopped, and investment has declined substantially. The country must overcome both real constraints--deriving on the one hand from the continuing need to provide satisfactory standards of living for a population which is still growing rapidly, and on the other hand from the foreign debt problem--and a criEsis of confidence; this will require a subistantial measure of social discipline and mobilization of political forces, as well as the support of the international community. While this report focuses on the industrial sector, the recommendations which derive from the analysis of Government policies and sectoral responses attempt to take into account as they must the linkages between macroeconomic and sectoral behavior. Thus, the recommendations address both the issues of short-term adjustment and the longer-term needs for ,development. Past Policy Framework and Performance 2. In the contexKt of the Government's main macro-economic objectives to accelerate income and employment growth and to improve income distribution, the essential objectives of industrial policy in the previous sexennium were (a) to contribute to the achievement of overall goals (e.g. employment growth, increased foreign exchange earnings from exports of manuLfactures); and (b) to promote a broader and deeper industrial structure similar to that of the more advanced countries. At the same time, sectoral policy had to concern itself with preserving Lndustrial growth in the face of macroeconomic policies detrimental to its competitive position. Underlying the macroeconomic strategy were certain basic assumptions concerning, on the one hand, external conditions (the maintenance of high and growing demand for Mexico's exports, partLcularly for petroleum and related products) and on the other hand, domestic conditions (availability of resources which could respond rapidly to demand growth). Both assumptions proved to be only partially correct. 3. The rather imipressive economic performance of Mexico during 1978 through 1981 can be seen from the data on the main parameters shown in Table 1. Rapid growth was largely made possible by the foreign exchange earnings from its petroleum resources and the concomitant availability of external financing. These external resources enabled the Government to embark on an ambitious expenditure program and aggregate demand growth rates rose beyond historic levels; these measures together with the expansion of private investment were intended to solve the employment problem. By 1980, the rapid rise in aggregate demand was beginning to exert pressure on prices; since the nominal exchange rate lhad remained practically stable since 1977, the higher Mexican inflation resulted in an appreciation of the peso in real terms. By 1980, the competitive edge gained after the 1976 devaluation had been lost; non-oil exports became progressively less attractive, and import growth accelerated to over 30% a year in real terms. In domestic production, this was reflected in non-tradeables being increasingly favored over tradeables.  - ii- Main Economic Parameters Average Annual Real Growth Rates, 1978-81 GDP 8.5 Aggregate Demand 10.5 Imports of Goods & Services 25.6 Manufacturing 8.7 Export of Goods & Services 8.9 Total Employment 5.7 Imports of Manufactures 28.1 Employment in Manufacturing 6.8 Exports of Manufactures 7.4 4. The policy framework for the industrial sector from 1976 to 1981 was composed of two main components, the first deriving from the macro-economic strategy, the second specific to the sector. The first component included the relatively stabLe nominal exchange rate regime, the use of subsidized prices for energy and for certain other basic inputs (mainly food and primary petrochemicals) to moderate domestic inflationary pressures and expansionary demand management; the public sector deficit rose frorn 5.7% of GDP in 1977 to 14.8% in 1981. The second component included import liberalization (which, however, was reversed in mid-1981) and export promotion, public sector investment in "strategic" industrial branches, and incentives for private investment and employment generation iLn manufacturing to promote accelerated growth in priority industrial branches, regional decentralization of industrial activities and small industry expansion; in addition sector specific energy and input price discounts were provided. 5. Industry benefitted substantially from the system of transfers and subsidies although it was not the principal recipient. The subsidized prices for energy and for basic petrochemical and agricultural inputs had their major impact on intermediate goods-producing :Lndustries such as basic metals and secondary petrochemicals and on food processing industries. There were also substantial direct subsidies to parastatal enterprises producing intermediates and assembling vehicles. Fiscal subsidies through investment incentives benefitted a broad spectrum of industrial branches. 6. Partly reflecting import liberalization measures which had been initiated in 1977, the protective system for manufacturing output (excluding petroleum refining) in 1979 was characterized by a rather low level of average nominal protection as compared to other industrializing countries, and lower than that recorded in Mexico in 1970. However, due largely to the subsidized inputs mentioned above, effective protection in 1979 was highly dispersed, more so than in 1970; a number of items, particularly among foodstuffs, were subject to negative levels while the technicologically advanced manufacturing sub-sectors (mainly engineering products) were subject to relatively high protection levels. Particularly high levels were found among consumer durables; paradoxically, the priority industrial branches, including capital goods, had considerably lowetr rates of effective protection.  - iii - 7. The combined impact of commercial policy and subsidies has been analyzed for 1979 through the measurement of "effective subsidies". Among the traditional industries, processed foodstuffs were found on average to have negative effective subsidies as a result of price and other controls, in spite of substantial input and other subsidies; consumer non-durables such as textiles had relatively low positive effective subsidy levels mainly through tariffs. Intermediate goods were the principal recipients of input and diretct subsidies with an average effective subsidy of 25%. The advanced manufacturing industries had an effective subsidy rate of 79%, mainly through tariff protection as well as mderate levels of fiscal, input and direct subsidies. 8. Subsidies appear to have induced substantial new investment by the private sector in the industries producing intermediate goods which are energy-intensive as well as capital-intensive. In addition, the traditional non-durable consumer goods industries undertook considerable capital expetnditure oriented towards modernizing and improving existing plant, presumably in response to import liberalization. Public sector investment in manufacturing industry was mainly directed towards the capital-intensive intermediate goods sub-sectors. 9. Beginning in 1978, industrial output experienced four years of exceptionally high growth, more than 10% p.a. in 1978 and 1979 declining to some 7% per annum in 1980 and 1981. There is little evidence to support the view that import liberalization after 1977 prevented the expansion of output; rather, reduced manufacturing output growth rates in 1980 and 1981 (as compared to 1978 and 1979) reflected a number of constraints on domestic supply. Although import liberalization permitted the sharp rise in imports, the fundamental cause of that growth was the expansionist demand management combined with the exchange rate policy which permitted the peso to appreciate in real terms. 10. The volume of exports of manufactures, which had fallen in 1975 and 1976, rose sharply following the devaluations at the end of 1976; however, the increasing real appreciation of the peso and the pull of the domestic market led once wre to a decline in 1980 and 1981. Mexico's exports of manufactures are highly diversified, reflecting the relatively broad industrial base which its market size has permitted and the recent growth in the development of technical skills; this has extended to exports of diversified engineering design and technical services. Many Mexican enterprises have shown that they can penetrate external markets. Moreover, many firms are increasingly planning productive investments which will require a significant proportion of output to be exported in order to reach efficient levels of output. 11. The increase in total employment in 1977-1981 exceeded by a considerable margin the increase in the labor force. However, most sectors experienced a decline in productivity growth during this period. This decline in productivity growth is not surprising, given that almost one in four of the manufacturing work force was hired after 1977; however, these rather low growth rates cannot be expected to persist. There are expected to be over 800,000 new entrants to the Mexican labor force each year throughout the 1980s; past policies will not be adequate to absorb these  entrants even when more normal output growth rates resume. In order for manufacturing to play its part in employment generation under current policies, manufacturing value added would have to grow at an annual rate in excess of 7% throughout the 1980s; given only eight or seven years of growth, in the remaining years of the decade, this figure increases to over 9% or 10.5% respectively. 12. In mid-1981, the external environment experienced a dramatic reversal with the emergence of a "glut" in the world petroleum market; the resulting balance of payments difficulties welre compounded by sharply rising interest rates on external borrowings and by the adverse impact of the recession in developed countries on non-petroLeum exports. The constraints on aiggregate domestic supplies had already manifested themseLves in the acceleration of inflation at the end of 1979 well in excess of external inflation. Continued use of subsidies and other expansionist public policies resulted in the growing public sector deficit. Failure to take the necessary measures to adjust to these new conditiLons thus underlies the current crisis. B. The Macro-economic Policy Framework_ 13. The Mexican authorities are nIow in the process of elaborating both short-run and long-run strategies designed to deal with the critical situation confronting them. In general, the rnature of both strategies must be congruent, essentially dealing with appropriate resource utilization and allocation. However, it must also be recognized that the severity of the current difficulties, the need to undergo a sharp contraction with resulting increases in unemployment, could lead to serious social problems which might jeopardize the recovery itself. In these circumstances, some actions may have to be taken in the short run that would later need to be. modified or phased out to encourage efficiency in the longer run. At the same time that such short-term actions are taken, the authorities should enunciate the longer-term program so as to minimize the risk of sub-optimal resource allocations and poor investment decisions being made. Moreover, a clear statement of long-term policy could contribute to the successful achievement of the short-term objectives. 14. In elaborating new strategies, it must be emphasized that an adequate macro-economic framework is fundamental for achieving the objectives of sectoral policy. This has been confirmed by the Mexican experience in the period 1979 through 1981 when many appropriate sectoral policies were followed, but inappropriate macro-economic management rendered them ineffective. The main concerns of future longer-run policies (and of short-term policies with the proviso above) must be (a) the maintenance of a real exchange rate which will favor domestic production of tradeables and (b) achieving a structure of factor prices reflecting resource availabilities. Such policies must lead in the first place to a sufficiently large surplus on trade account to meet Mexico's current external liabilities. In addition, since the depletion of Mexico's petroleum resource base, which currently constitutes a significant portion of its exports, is comparable to dis-investment, it must be compensated Eor by an expansion of investment in the tradeables sector. 15. To achieve these objectives, it will be necessary to ensure that real wages which have been reduced by recent measures do not rise in excess of increases in productivity or more than that permitted by fuiture improvements in the external resource balance. It will also be necessary to  v exercise restraint on public sector expenditure to avoid the competition for real resources with the private sector which characterized the recent past. 16. The necessity to develop an externa:Lly-oriented industrial sector sterns not only from macro-economic considerations, but also from sectoral conditions. At the beginning of the 1970s, itports of manufactures represented less than 10% of domestic demand; while that ratio rose in 1980 and 1981, the increases were due to unsustainably high levels of aggregate demand and the overvalued exchange rate and did not reflect a major change in the relatively low levels of import dependence which had been achieved. Although at the product level import substitution is a continuous process, it is not likely that the aggregate ratio can be reduced further without creating distortions adversely affecting the longer-term growth of the sector. Thus, industrial output would be limited to the growth in aggregate demand unless exports of manufactures are increased. Moreover, higher output levels in manufacturing are required to obtairL the increases in employment opportunities necessary to absorb the growing labor force. Policies for Adjustment Period 17. The framework for macro-economic policy in the next three years is largely defined in the agreement signed by the! Mexican Government and the International Monetary Fund for the granting of an Extended Fund Facility (EFF). The economy will have to undergo a severe contraction in 1983. This program is supported by a set of policies designed to achieve an improved fiscal balance and to keep Mexican exports competitive. Recovery is expected to be led by private investment in 1984-85. Under the EFF guidelines, public investment cannot play the role it played in 1.977; nor can the external sector, under current assumptions for moderate! OECD growth, have as decisive an impact on aggregate demand as in the past. 18. The re-establishment of a healthy irLvestment climate will depend, on the one hand, on the extent to which entrepreneurial confidence in economic management is restored; and, on the other, on expectations concerning the prospects for social and political stability in the future. Compliance with the EFF guidelines should help overcome the effect of last year's events on private sector confidence; but, at the same time, care must be taken to avoid excessive social pressures deriving from a prolonged recession. Although the EFF projections imply that in 1982-84 Mexico will grow at substantially less than the past trend rate, actual performance could be even worse than the projections if recovery from the world economic recession is slower than anticipated and if oil prices weaken further. 19. The Mexican economy will be facing a major unemployment problem. Although precise estimates are hazardous, it is clear that, in 1982 and 1983, the ,economy will have failed to absorb the 1,600,000 new entrants into the labor force, and unemployment may keep on growing beyond 1985. If actual GDP growth is negative in 1983-84, the unemployment problem may become very serioDus with consequent implications for the maintenance of social and political order.  - vi - 20. Specific sectoral policies for employment maintenance and generation should be pursued to influence the real wage rate, the effective cost of labor to employers and the relative prices of capital and labor. In the short term, while high layoff costs constitute some incentive to formal sector firms to mainta:Ln employment levels wherever possible, the mDst important measure the administration can take for employment: maintenance is likely to be intervent:Lon to support firms which are sufferiLng from short-term liquidity problems. Support has already been given to firms through the increased availability of credit for working capital and through assistance with foreign debt payments; such support may need to be expanded and other, more direct forms of intervention may have to be found if these are not sufficient. In addition, in order to protect employment, consideration could be given, for example, to short-run measures to offset the impact of indirect labor costs; futrther study of this issue is required. In the case of firms in the informal sector, as well as smaller enterprises in general, mechanisms which have been developed for longer-term financial and technical assistance must be adapted quickly to meet these new needs. 21. The current "controlled" exchange rate, which is applied to most import and export transactions, represents a significant real devaluation relative to the previous periods of good export performance; the stated policy is to continue to adjust this rate in line with domestic inflation so that it will converge with the free rate. As such the policy represents a major shift in favor of domestic production of tradeables, promoting exports and encouraging a decline in imports; aside from its contribution to improvement in the trade balance, this can make a contribution to supporting existing employment levels. 22. However, maintaining employment at Ea given level--say, that of early 1982-is not sufficient; the economy must find a way to absorb the large number of new entrants to the labor force each year. The Government has proposed job programs which may be able to, assist in the short run; the informal sector will also clearly have a very important role in absorbing labor. Nevertheless, policies to promote employment generation in the formal sector, and in particular manufacturing, must also be devised and implemented. 23. There are two measures which, taken in conjunction, could encourage significant formal sector employment generation; both require a mre flexible approach than at present to contracting new labor. The first measure would be the introduction of a system allowing, at least for the next two or three years, new employees to, be hired on a short-term contract basis, possibly up to one or two years, with no layoff penalties at the end of the contract period. The second meaLsure would be the revival of the "aprendizaje" (apprenticeship) system; however, the system would have to be designed to avoidl hiring trainees when there is little chance that they will be retained after the training period. 24. In this connection, there is need to strengthen the institutional arrangements for the formulation and iraplementation of employment policies. An improved system of collection and analysis of labor force, productivity and wage statistics would also help support more timely employment policy  - vii - development. Of particular importance in the current situation is better knowrledge of developments in the labor market to permit rapid adjustments to emerging difficulties. 25. Finally, to reduce private sector uncertainty and thus to encourage its necessary revival, there is urgent need to clarify the issues relating to control and ownership of the equity holdings of the nationalized banking system and to assure the private sector of equal access to credit from that system. Main Components of Longer-Term Strategy Trade Policy 26. The fundamental objective of trade policy must be encouragement to the growth of exports. Because import protection leads to (or is used to support) an overvalued peso, it discourages exports. Rationalization of protection is therefore a crucial element of any strategy to increase manufactured exports. As soon as the current crisis has been overcome, there is need to return to the process of import liberalization which had been initiated in 1977. It is, therefore, important that the present Government renew the commitment to dismantle over time the quantitative import controls which have been imposed, together with the reaffirmation of the intention to follow a congruent exchange rate policy as not:ed in para. 14. 27. There is also need to review the protective structure which is provided by the present tariff system and to eliminate the excessive dispersion of effective protection; in particular, very high protection has been given to consumer durables, the economic justification for which is not clear. There may be some justification, however, for some product lines, particularly among the capital goods, to receive temporary protection, combining tariffs and production incentives, on the basis of the infant industry argument. Careful selection of specific product lines to receive such protection is necessary and manufacturers must clearly understand that they would be required to achieve internationally competitive price levels for these items within reasonable periods of time (say, 5 years). 28. Residual anti-export biases also need to be removed. Policy related measures for this purpose can be identified through the studies which are now available on effective protection and competitiveness. It is also important to ensure that the controls which haLve been imposed on exports and imports are efficiently administered. A fundamental prerequisite for the efficient operation of these systems would be the close coordination with the private sector and particularly with those involved in exporting. In the longer run, as such measures represent basic disincentives, they should be gradually dismantled. 29. There is a need for a more thorough analysis, at the product level, of the impact of various indirect taxes so as to determine whether a new tax rebate system can be constructed to replace the CEDI system, which would be compatible with internationally accepted practices.  - viii - 30. To penetrate external markets, it will be necessary to strengthen the efforts at export promotion. This will require close ccoperation between the private sector and the relevant official institutions in order to define priorities and to assign responsibilities in accordance with areas of competence, as well as to take advantage of scale economies in external promotional campaigns. The possibility of making increased use of joint export marketing and distribution mechganisms such as export consortia, for example, should be further examined. Efforts at improvinlg the availability of market information meed to be strengthened, especially under current difficult marketing conditions. Fiscal Policy and Subsidies for Industry 31. This is clearly an opportune time to review the impact of direct subsidies and fiscal incentives provided to Mexican industry. Such policies have distorted industrial growth and hiave represented a drain on the public fisc, contributing substantially to the recent inflation. To avoid those distortions and reduce expenditure, the system of subsid:Lzed prices for certain agricultural products, energy and petroleum-related inputs should be dismantled. With regard to energy, better criteria are needed for the determination of the optimal pricing pattern for domestic oil and related resources and the search for such criteria is now being umdertaken by the Mexican authorities. While recent measures have been taken to increase the pric,es of some of these products, further increases are required and it will be necessary to adjust them periodically to maintain the appropriate real levels. 32. Inappropriate pricing policies have resulted in the need for direct subsidies to non-petroleum manufacturing activities controlled by the public sector. Adjustments of these prices are required to eliminate this source of unnecessary public expenditure. However, the longer-term viability of a number of these enterprises will have to be examined carefully; but, in the immediate future, it is desirable to maintain production and thus employment in t'hese particular activities. 33. There is alsc, the need for a longer-term review of the Government's direct industrial investment policy particularly from the point of view of employment generation aLnd capital intensity. It is important that, as part of the investment decision-making process, economic evaluations be undertaken whicih use appropriate prices for labor and capital under existing market conditions. 34. The present nieed to reduce public sector expenditure may result in a significant delay in implementing infra-structure investments, particularly those related to the efforts to promote decentralization of industry and the development of the new industrial ports. Priority should be given, however, to the completion of thLe facilities needed to handle export shipments, given limited existing capacity. 35. During the period 1977 through 1981, interest rate policy had not been a source of substantial subsidies for the industrial sector. The authorities pursued an open capital market policy which resulted in longer-run lending rates which were generally positive in real terms,  - ix - although highly subsidized credit lines were made available to the agricultural sector. 'rhe current situation has led to the imposition of capital controls and credit restrictions whiclh have caused interest rates to lag behind inflation. As the stabilization program takes effect, it is important that the authorities ensure that interest rates be set at levels which would be positive relative to expected inflation. In addition to encouraging domestic savings, such a policy would discourage inappropriate factor proportions (see para. 37). At the sale time it is important to ensure that interest rate differentials between sectors (such as agriculture and industry) be eliminated or reduced as much as possible. Employment Policy and the Cost of Labor 36. Growth in employment opportunities and the reduction in unemployment remain fundamental objectives of economic and social policy which must be achieved within the constraints imposed by macro-economic conditions. Demographic developments suggest a decline in the rate of expansion in the second half of the 1980s, but: there is nevertheless the need to absorb each year over 800,000 new entrants into the labor force; while industry cannot be the principal source of new employment, its role needs to be expanded. In the earlier sections, referertce was made to measures to maintain, and where possible, to generate industrial employment during the next few years of adjustment. Other policy changes are necessary for the medium and longer-term when new investment is expected to accelerate. 37. Past subsidies and incentives have favored investment in capital intensive industries; their elimination would contribute to reducing such distortions. It is recognized that the process of industrial development will require increasing productivity and new investment whose costs per job created and the degree of capital intensity will be higher than in the past; the main concern is that measures be taken to ensure that those costs and the degree of capital intensity will be no higher than required by technology. 38. While the 1979 industrial investment incentive decree has been modified to eliminate the bias against investment in the most labor-intensive activities, the system still retains characteristics favoring capital for those industrial branches with intermediate capital/labor intensities. Further study of the system is required to reduce that bias. 39. In addition, there are distortions arising in the formal sector tend:Lng to raise the real cost of labor to producers and these need to be revised. The most direct means would be to decrease payroll taxes and levies that increase overall labor costs. An alternative might be to increase the general corporate taxation rate at the same time as increasing the subsidies to employment. A study of the effect of labor legislation on the real cost of labor to employers, attempting to quantify these effects and to devise ways of mitigating them, would be very useful. The degree to which the relative price of labor should be reduced through subsidies, and the extent to wlich subsidies to capital should be reduced--if at all--will also depend on the incentive level, if any, that is felt to be required to stimulate investment in manufacturing as opposed to other sectors. There is need for a further review of this issue.  - x 40. It is also essential to design viable policies to support job creation in, and appropriate development of, the informal sector. Again, the determination of the precise mechanisms warrant further study. The principal constraint on increasing the productivity of labor in the informal sector is the high effective cost of capital to the sector which results from limited access to capital sources; this could be reduced through increasing the availability of credit to the sector through a mechanism such as a credit guarantee fund. In addition, and perhaps more importantly, organizational forms, different from those in the formal sector, should be preserved and reinforced where they exist and created elsewhere. 41. Finally, for the longer run, a better trained labor force is essential. It is important that training remaLin very much at: the "grass roots" level, responding to entrepreneurs' needs rather t'han to a predetermined plan based on predicted nanpower needs. This implies a system based on either on-the-job training or trainirLg through a local center controlled, in part, by local firms, or some combination of the two. The revival of the "aprendizaje" system, proposed above, would provide a good start to an on-the-job training system.  I. THE MACRO-ECONOMIC SETTING The Goals of Industrial Policy 1.01 The manufacturing sector is often felt to be an area in which an appropriate policy framework has been lacking during the Lopez Portillo Administration. This perception cannot be justified by the absence of state intervention in industry for, by any reasonab:Le measure, the role of the public sector in manufacturing was as important in 1981 as it was in 1976; if at all valid, it must be based on the realization that industrial policy goals either have not been clearly defined, or have not been achieved to a satisfactory degree. 1.02 In very general terms, the main goal. of industrial policy in a market economy should be that of correcting market imperfections so as to promote efficient growth of the sector: hence the adoption of spatial decentralization policies in order to eliminate congestion externalities, or the imposition of a certain degree of protection with the purpose of letting infant industries take advantage of decreasing costs. However, while the short-term efficiency objective is usually a part of any industrial policy, it is never its only, and seldom its most important, component: in most cases--and certainly in that of Mexico--efficiency arguments are overlaid, and sometimes contradicted, by considerations of a different nature: first, the industrial sector is often asked to contribute its share to the attainment of general policy goals; second, it. is commonly a subject for planning exercises, often with non-economic objectives in mind; third, it must operate within the constraints set by overall macro-economic policy. Elements of these three kinds can be found in Mexican industrial policies, and it may be useful to indicate the extent to which they have influenced the role of the basic efficiency criterion in recent years; in particular, the extent to which industrial policy has been shaped: first, by the desire to attain certain overall economic targets; second, by the wish to make the structure of the industrial sector conform to a general norm; and, finally, by the need to preserve its growth in the face of macroeconomic policies detrimental to its overall competitive position. The first two kinds of considerations are well illustrated in the Industrial Development Plan of 1979; as for the third, it may be said that, since 1979, many industrial policy measures have been inspired by the need. to maintain the growth of the manufacturing sector - and, in particular, that of manufacturing exports - against the effects of high growth in domestic demand together with an appreciating real exchange rate. The Industrial Development Plan (1979) 1.03 The Plan, prepared in the first two years of the Lopez Portillo Administration, provides a good example of the way in which industrial policy is thought of as a tool to achieve overall economic and social objectives. Industrial development was meant to address the main structural problems of the economy perceived by the Mexican administration: under-utilization of resources, inequalities in income distribution, a precarious external balance and an excessive dependence on foreign technology, as well as those exacerbated by the 1976-77 recession. Consequently, the Plan included among  2 its objectives: accelerating the generation of employment opportunities, promoting industrial investment in accordance with regional priorities, increasing the real income of the popuLation, orienting industrial production towards external markets, and promoting indigenous technological growth in industry. The connection between any one of these objectives and efficient growth is at best indirect; yet, before a judgement is passed on the degree to which the Administration succeeded in achieving its purposes, it must be remembered that these goals were often given greater weight in industrial policy formulation than were efficiency considerations. 1.04 The Plan was based on a review of output targets for each indus- trial subsector; and implicit in these quantitative targets was the wish to bring the structure of Mexico's manufacturing sector closer to that of more industrialized countries. Thus, for example, a greater degree of sufficiency was expected in steel and cement, while the production of capital goods was given a high priority in the Plan, Such targets were not based on any notion of static efficiency; rather, they reflected t:he perception of certain defi- ciencies or disequilibria in the industrial sector. In general, such defi- ciencies appear in a comparison betweern a given country and similar ones having a higher level of income per capita; and, even though the adoption of such "norms" can hardly be justified on the grounds of interrnational special- ization and comparative advantage, the fact remains that economic growth seems to give rise to certain patterns of structural change which, in broad terms, turn out to be fairly predictabl.e. Hence the tendency to single out subsectors which appear as relatively underdeveloped by compEarison to some average structure. 1.05 Applications of the best-documented of these "patterns-of-growth" models (the Chenery-Taylor model) have been done more than once in Mexico,l/ pointing out certain discrepancies between the. Mexican industrial sector and the norm: first, both in terms of value added/GDP and in terms of employment share, Mexico's industrial sector as a whole fell below the average; second, certain subsectors (especially basic metals and metal products) fell below the norm, not only with respect to more developed countries, but also in comparison with countries of a similar level of development, such as Brazil. Furthermore, these discrepancies showed no signs of disappearing; rather, they all increased significantly in 1970-1975. Although the Industrial Development Plan was not explicitly based on a "patterns-of-growth" model, its concern for faster industrialization and the establishment of some of its subsectoral priorities might be explained by the desire to correct such deviations from a broad norm, the more so since these peculiarities can be explained in part by Mexico's proximity to the U.S. This is another element to be brought in when appraising Mexico's industrial policy. Macroeconomic Policy and the Industrial Sector 1.06 Finally, the nature of Mexico's industrial policy in recent years cannot be properly understood without taking into account the effect of macroeconomic policies upon sector performance. To some extent, Mexican industry has been affected by the macroeconomic policies of its major 1/ See IBRD, Mexico - Manufacturing Sector: Situation, Prospects and Policies (March 1979).  - 3 - trading partners - insofar as these resulted in a slower growth of world trade after 1973, and in rising interest rates after 1980; much more important, however, has been the effect of domestic macroeconomic policy on the performance of the sector. 1.07 Following the 1976 devaluation, the Administration implemented a set of stabilizing measures which, by the end of 1977, had succeeded in decreasing the rate of inflation to 29% (it had exceeded 60%, on an annual basis, in the last quarter of 1976); in reducing the public deficit/GDP ratio from 9 to 6.2%; and in lowering the current account deficit/GDP ratio by half to 2%. At the same time, the lower rate of GDP growth had been achieved through lower aggregate demand and, in particular, through lower private and public investment, and unemployment had risen accordingly. 1.08 The following year was marked by the decision to speed up GDP growth in order to start absorbirg unemployment. Fast growth was made easier by the relaxation of the foreign exchange constraint, through the inflow of oil revenues as well as by the access to external finance provided by the existence of large reserves of crude. Thus, between 1978 and 1981, GDP grew at an annual rate of 8.6% in real terms, while imports grew at an annual rate of 25%. As a consequence, aggregate demand was rising at an average of almost 11% per annum. Furthermore, investment was the leading factor in GDP growth, its share in GDP growing from less than 24% in 1978 to 29% in 1981; finally, it is estimated that employment grew significantly, by some 5.7% per annum, during the same period. 1.09 Rapid growth subjected the economy to severe strains, which were beginning to be apparent by the second half of 1979: even though the proportion of imports in GDP grew from 10.3% in 1977 to 13.5% percent in 1981 - this increase being most noticeable in intermediate inputs and capital goods - the domestic resource base was under great pressure. As a result, domestic prices began rising faster than those of Mexico's major trading partners: even taking the 1976-1981 average as an "equilibrium" exchange rate, the 15% margin of undervaluation which still existed at the end of 1977 had been eroded by the end of 1979 and, two years later, the peso was over- valued by 17%; taking the period 1977iii to 1978ii as a benchmark, the overvaluation at end-1981 reached 26%. 1.10 Put another way, the financial transfers represented by oil revenues and foreign borrowing had to materialize in real transfers, that is (if one excludes increases in reserves) in excesses of imports over exports. The real transfer was effected, in part, by public imports, which had an inflationary effect only insofar as they were complements to non- tradeable goods in inelastic supply (e.g., PEMEX imports which required skilled labor for their operation). At the same time, foreign financial resources found their way, through expansionary fiscal policies, into the private sector; higher demand for both domestic and foreign goods resulted in a higher relative price for domestic goods, that is, in an improvement in the terms of trade; hence, the abundance of foreign financial resources translated itself into a real appreciation of the peso. 1.11 Such an improvement in the terms of trade had to affect the performance of the manufacturing sector: on the one hand, it became cheaper to iraport certain goods than to buy them domestically; this is one of the  - 4 - reasons why, for instance, capital goods imports went from 26 to 31% of total imports between 1977 and 1981. At the same time, it became easier to sell at home than to export; exports of manufactures slowed down, from 33% to 11% of total exports in the same period; they even decreased, in dollar terms, in 1981. In short, even though the symptoDms did not become apparent until late 1979 or early 1980, it is clear that, by relying on oil exports and on borrowing against oil reserves as a source of foreign exchange needed for GDP growth,' Mexico was following, after 1977, the path of "de-in,dustrialization", whose main trait is persistent current account deficits together with a shrinking base of non-oil exports (a hLgh domestic rate of inflation is, as has been seen above, a consequence of the way in which the real transfer of resources is made, and is not inherent to de-industrialization). 1.12 In the short run, a greater reliance on imports may well be an efficient way of making use of oil revenues, insofar as it lLberates resources which could be better used eLsewhere. In the long run, however, the trend towards de-industrialization will push the manufacturing sector away from its avowed policy goals by reducing the degree of diversification, both in production and in trade, by discouraging domestic production of certain strategic items, such as investment goods, by encouraging the substi- tution of imported inputs for domestic factors - such as labor - in produc- tion, and by creating persistant external disequilibrium. From the outset, there was a certain degree of awareness of the undesirable s:Lde-effects of excessive reliance on oil exports, and hence much of industr:Lal policy after 1978 is better understood as an attempt to ward off de-industrialization. 1.13 Finally, even though the problems faced by the manufacturing sector in the second half of the Lopez Portil]Lo Administration can be said to derive from the macroeconomic policies adopted, it would be simplistic to attribute them solely to the management of the extchange rate: rather, they stem from the growth strategy chosen--since a rapid expansion of public expenditures had to be met through deficit financing--and from the presence of resource bottlenecks in the economy. Put differently, a faster rate of devaluation in 1979-80 would have achieved little - beyond further increases in the domestic rate of inflation - so long as the pace of aggregate expenditures remained as high as it was; hence, although the recent problems in the Mexican economy have manifested themselves in macroeconomic disequilibria, these are more appropriately seen as symptoms of a failure to achieve a satisfactory degree of resource mbilization.  -5- II. MAIN ELEMENTS OF SECTORAL POLICY, 1976 - 1981 2.01 The main objectives pursued through macro-economic and sectoral policies have been described in Chapter I; this chapter examines the specific policies undertaken in the industrial sector in order to achieve these objectives. It is convenient to characterize the policy framework for the industrial sector which was established by the new administration taking office in December 1976 as composed of two elements - one which derived from overall economic (macro) strategy and one which was sector specific. The components deriving from the macro-framework described in Chapter I include: the relatively stable nominal exchange rate pu.rsued through early 1981; the use of subsidized prices for energy and for certain basic inputs (mainly food and basic petrochemicals) primarily as an anti-inflationary weapon (except for additional discounts on energy and petrochemicals prices which were sectoral specific mentioned below); and expansionary demand management. Those elements which derived from sectoral considerations include import liberalization and export promotion, public sector investment in "strategic" industrial branches, and incentives for private investment and employment generation in manufacturing to promote the accelerated growth of priority industries, regional decentralization of industrial activities, and the expansion of small industry. Commercial Policy 2.02 During most of the post-World War II period through the early 1970s, the primary focus of commercial policy was the promotion of the indus- trial sector.'/ Briefly stated, an inward-oriented development strategy was formulated which considered that the market size of Mexico could justify the production of most of the industrial products consumed in the country. Since local demand of manufactured goods in developing countries is normally first met through imports, the principal means for stimulating the growth of local production is protection from imports. In this formulation, exports of manufactures is a residual.2/ 2.03 The specific protective measures adopted by the authorities were characterized by moderate tariff levels (relative to other Latin American countries) but a rather comprehensive system of import licensing which covered an average of about 60% of the total value of imports during the 1960s. The structure of the protective system in 1970 has been studied by a private research group3/; the results, which were based upon direct price 1/ Virtually all primary agricultural trade has been in the hands of the public sector and thus subject to direct government control. See the forthcoming agricultural sector report. 2/ There is substantial documentation on this issue as it relates to Mexico; a brief summary of the discussion during this period is provided in a study by Antonio Aspra, "ImLport Substitution in Mexico: Past and Present", World Development, Vol.. V (1977). 3/ See Adriaan ten Kate and Robert Bruce Wallace, Protection and Economic Development in Mexico (1980).  - 6 - comparisons for a large sample of goods, indicated an average nominal prot,ection for manufacturing industry of about 20%. The study found that effe,ctive protection averaged 60% but was characterized by considerable dispersion. 2.04 It was not until the early 1960s that there was some recognition of the limits of the inward oriented industrial strategy. An export financing facility (FOMEX) was created in 1962 by the Bank of Mexico tc provide funds at preferential rates and an export promotion institution (IMCE) was created at the end of the decade. Export taxes, which as late as 1955 had accounted for 20% of total ordinary federal revenues, were substantially reduced in 1961 and a system of ircentives for exports was initiated involving primarily the reimbursement of certain federal taxes. IThis system, howrever, provided only small financial benefits and was replaced in 1971 by a more comprehen- sive set of tax reimbursements (CEDIs). These were subsequently modified on a number of occasions. 2.05 However, the fundamental nature of the protective system remained unchianged and was in fact in many ways intensified from the early 1960s through the early 1970s. The percentage of total items in the customs nomen- clature which were subject to licenses increased from 44% of total in 1962 to 60% :Ln 1966 and by 1973 reached 80%. The value of controlled imports rose to 72% of total imports in 1974, from the 60% figure prevailing in the 1960s. 2.06 In 1977 important changes were introduced in the policy framework. The number of items subject to controls was gradually reduced., although their share in the value of total imports did not decline proportionally. For the item; no longer subject to licenses (including many in the consumer goods category) moderate increases in tariffs were introduced. As regards export incenitives, although CEDIs were eliminated after the devaluation in late 1976, the system was restored in April 1977. In 1979 it is estimated that the GEDIs were equivalent to approximately 11% of the valule of eligible exports. Further details on these incentives are provided in, Chapter IV. In general, the intention of the authorities was to reduce tle anti-export bias of the protection system through a gradual rationalization of tariffs4/ while at the same time pursuing a more realistic exchange rate policy. In mid-:1981, however, as the peso became increasingly over-valued in real terms, the authorities moved to reverse the liberalization process. Industrial Investment Incentives 2.07 The main instrument of sectoral policy was the industrial investment incentive system which was put into place at the beginning of 1979. The system provided subsidies to investment by the private sector to meet certain objectives regarding employment, location and the promotion of prio;rity idustries. The rationale for the system was set out in the National Industrial Development Plan, described in Chapter I. 4/ In order to provide a basis for that process, the government organized an inter-ministerial working group to prepare a study of effective pro- tection. Some of the main results of this study are analyzed below.  -7- 2.08 In addition, the Plan indicated the expected role of public sector investment, including direct investment in strategic subsectors as well as in economic and social infrastructure to meet the objectives of decentralization and modification of the industrial sector's structure. However, the actual investment to be undert.aken was to be identified in the course of the annual budget exercises. 2.09 The fiscal incentives offered to private sector investment took the form of tax rebate certificates (CEPROFIs). Iable 2.1 summarizes the system and shows how the incentives (either for new facilities or expansion of existing plant) varied in accordance with regional and subsectoral priorities, from 10% to 20% of investment costs. Among the industrial branches, highest priority was given to expansion of capital goods producing industries, output of certain intermediates critical to the rest of the economy (steel and cement), and the development of agroindustries. These priorities were based on an analysis of input-output relationships in Mexico and other countries and were designed to capture backward and forward linkages both within the sector and with other sectors. The attention given to manufactured foodstuffs reflected the desire to increase the supply of these items for mass consumption. 2.10 The main objective of the regional decentralization effort was to achieve a better spatial distribution of industry and to control the growth of the main cities, in particular Mexico City, where the burgeoning popula- tion had already created serious social, economic and ecological problems. Top priority was given to the development of four "industrial ports" (Zone IA), embracing the areas around Lazaro Cardenas, Salinas Cruz, Coatzacoalcos and Tampico, and of selected other areas (Zone IB) whose development was considered important within the National Urban Plan. The second group of areas (Zone II), which are considered state priorities, were specified in agreements reached subsequently between the federal and state authorities. Growth in the Federal District and its immediate environs (Zone IIIA) and other nearby regions (Zone IIIB) was to be controlled and investments in those areas were excluded from receiving incentives with the exception of expansions of existing enterprises in Zone IIIB. Finally, as regards the rest of the country (i.e., the areas not specifically included in Zones I, II or III), the incentives offered to investment were lower than for the two priority areas (Zone I and II).  Table 2.1: SUMMARY OF PRINCIPAL FISCAL INCENTIVES FOR INDUSTRIAL INVESTMENT UNDER ORIGINAL 1979 DECREES Priority Sectors' All Industrial Activities Geographic Location Other Purchase of Employment Generated of Investment Small Enterprises bi Activity Cate ory I CL-ActivitX Category 2 d/ Industries Local Equipment by Additional Shifts Zone I Preferential 25% Investment 20% Investment 15X Investment - 5% 20% Additional 20% Employment 20% EMp loymeIt Emplo-,ent Zone II State Priorities 25S Investment 20% Investment 10% Investment - 5 i20% Additional 20Z Employment 20% Employment Employment Rest of the Country 25% Investment e/ 20% Investment 10% Investment e/ 5% 20% Additional 20% Employment 20% Emlfloyment e/ Employment Zonie I11I A. Controlled growth 5- Elo Consolidated orth 25X investment et 20% Iniveatment e! - O Investment el 20 Additional I 20% Employment el 20X Employment e/ - Employment a/ Geographic zones are defined in the decree published in Diarib Oficial, February 2. 1979. Fiscal credits can be used for the payment of any federal tax which is not imposed for a specific use. The percentages shown in the table for investment are applied to the total value of construction and installations and the purchase of new mnachinery and equipment directly related to the production process; it is granted at the time of undertaking of the expenditure. In the case of employment, the percentage is applied for a period of tvo years on thte new employment valued at the annual minimum wage. b/ Enterprises with fixed assets not exceeding 200 times the annual minimum salary in the Federal District. c/ Includes agroinduttries, capital goods producing industries and strategic inputs for the industrial sector (e.g. steel, cement) d/ Includes non-durable consumer goods, durable consumer goode and intermediate products specified it decree published in Diario Oficial, March 9, 1979. e/ Applied only to expansions of productive capacity in the same industrial activity. Source: Secretaria de Patrimonio y Fomento Industrial, Plan Nacional de Desarrollo Industrial, 1979-1982 (Mexico, 1979) p. 181.  9- 2.11 The objective of increasing employment was to be met by offering tax incentives for new jobs created by any investment, either for expansion or n,ew capacity, based on 20% of an imputed labor cost (multiplying the number of new jobs by the annual minimum legal wage in each zone), for a period of two years. Any enterprise which employed an additional shift was also eligible for a similar incentive (except in the Federal District). Through these measures, the authorities expected to offset any tendency for investment incentives to encourage inappropriate capital intensive production processes. 2.12 Fiscal incentives were also given in accordance with other criteria. Small industries, irrespective of product and location (except for the Federal District), received higher fiscal incentives (25%) as regards investment than any other type of industry, but no employment subsidy. FinaLly, as a means to promote the purchase (and therefore the production) of domestically produced machinery and equipment, all enteprises investing in such items were entitled to a tax credit equal to 5% of the value of the purchase. 2.13 In addition to the fiscal incentives, special prices for certain inputs supplied by public sector enterprises were offered to favor the location of new industries in the four industrial ports (cited above). All new industries established in those areas received a 30% discount on the prevailing prices of energy supplied in the form of electricity, natural gas or fuel oil. Moreover, new enterprises in those locations producing specific secondary petrochemicals received a 30% discount on their purchases of basic petrochemicals,5/ provided that they export at least 25% of their output. 2.14 Subsequent to the publication of the original decrees, some modifi- cations were introduced, which were mostly of a minor nature, the energy and input discounts were extended to certain other geographical zones and some of the zonal definitions were changed. An attempt to stimulate investment in export-oriented industries was made in June 1981 when a decree was issued which permitted an increase in the rates of both the investment and employ- ment incentives to a maximum of 30%, provided that the project would produce at internationally competitive levels and quality standards.6/ In addition, the rebate for the purchase of locally produced equipment was raised, under specified conditions, to 15%; the rebate was also extened to components. 2.15 In March 1982, however, a major change was introduced in an attempt to increase the employment incentives. The original system had represented a new approach to this particular problem recognizing that traditional invest- ment incentives had contained anti-labor biases. The system initiated in 1979 thus attempted to combine both employment and investment incentives. However, it became clear in implementing the system that the specific incen- tives offered to employment generation were inadequate. Part of this was due to the administrative difficulties in measuring new jobs and the 1982 decree 5/ These were in addition to the below world market prices which prevailed for these inputs other than electricity. 6/ Partly due to a delay in the formulation of the regulations for this provision, through the end of 1981 no projects had been presented for eligibility under this amendment.  - 10 - instituted new procedures to simplify the process. But there was recognition of a more basic weakness of the originaLl decree in that the relative incentives to capital and labor were substantially biased in favor of the former. The new decree therefore offered an alternative: an enterprise could choose between the old system or an incentive equal only to 40% to 80% of labor costs (as defined in the original decree) and for up to three years, depending upon sub-sectoral and regional priorities which had also been set in the original decree (see Table 2.2). The nature of the anti-labor bias of the original system and the extent to which the new alternative modifies that bias is examined in Chapter V; a detailed analysis of the system is contained in Annex I. Table 2.2: ALTERNATIVE INCENTIVE SYSTEM FOR EMPLOYMENI' GENERATION (as contained in. Decree of March 15, 1.982) Percentage of Incentives Priority Categorya/ (in relation to labor costs)b/ Category 1 A. In all geographic zones except Zone III 80 B. In Zone III B, only expansions 80 Category 2 A. In Zone I 60 B. In Zone II 40 C. In all other zones, except 40 Zone III A, only expansions Source: Diario Official a/ For definition of categories and zones, see Table 2.1 bI For method of computation, see Table 2.1  - 11 - III. SUBSIDIES AND PROTECTION A. The Growth of Transfers and Subsidies 3.01 One of the most important elements in the overall growth strategy pursuied by the Government, as described in Chapter I was its use of transfers and subsidies (T&S) to support productive activities (agriculture and industry) and to attempt to moderate the impact of the growth of domestic demand on inflation. In this section, the development of the T&S system is analyzed as regards its composition and importance for the various sectors of the economy. Table 3.1 summarizes the amount of tax incentives and subsidies for 1975 and 1981 for the whole economy. Table 3.1: TOTAL TRANSFERS AND SUBSIDIES BY TYPE, 1975 AND 1981 Type of Amount as % Transfer Amount Structure of Public Amount as % and (106 Mex $) (%) Expenditure of GDP Subsidy 1975 1981 1975 1981 1975 1981 1975 1981 Direct 45,316 402,682 64.8 51.0 11.3 17.3 4.1 8.2 Input Prices -1,821 298,240 -2.6 37.8 -0.5 12.8 -0.2 5.0 Tax I]ncentives 10,783 27,100 15.4 3.4 2.7 1.2 1.0 0.5 Finanicial 15,640 61,825 22.4 7.8 3.9 2,7 1.4 1.0 TOT'AL 69,918 789,847 100.0 100.0 17.4 33.9 6.4 13.7 Sourc.e: Mission estimates based on data from SHCP It is apparent from Table 3.1 that T&S represent a large part of the total budget and a sizeable fraction of GDP; this is an indication that the macroeconomic impact of the subsidy system on the economy must be considerable, and its eftfect on resource allocation not negligible. 3.02 Figure 3.1 shows how total transfers and subsidies have increased much faster since 1970 ithan both public expenditure and GDP; Table 3.1 indicates that transfers3 and subsidies have increased from 6.4% of GDP in 1975 to 13.7% of GDP in 1981, while they represented 33.9% of public expendi- ture in 1981, as opposed to 17.4% in 1975. 3.03 Figure 3.2 shows this evolution of total transfers and subsidies relative to public expeniditure, and highlights the increasing importance of two transfer and subsidy7 mechanisms: direct and input prices, with financial T&S and tax incentives playing relatively minor roles.  TOTAL TRANSFERS AND SUBSIDIES, EXPENDITURE AND GDP, 1970-1981 TRANSF$ ANii .JSlDIE. 700- PUBLIC EXPENDITURE --6GROSS DOMESTIC PRODUCT 400B 300- / 2G - ------ so - ~ ~ ~ ~ ~ ~ ~ ~ tee___ _ ,__- --- --- B- I I I I I I I 1970 t171 1972 1973 1974 1976 1970 1977 1978 1979 1G8B 1981  TOTAL TRANSFERS AND SUBSIDIES BY TYPE, AS PERCENTAGE OF PUBLIC EXPENDITURE 484 25- 25~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ---- ....--- -_ 15~~~~~~~~ *_- ,-'- X~~~~~~~~~~- 4e * -4- a~~~~~~~ -.-. -- 9 .. 18 - S ~~~~~~~~~~~~~~~~~~~~TAX DICENTZYES VMRCT Iv i1  - :L4 - 3.04 Total transfers and subsidies are brcken down by sector for 1975 and 1980 in Table 3.2 below. Table 3.2: TRANSFEBts A!ND SUBSIDIES BY SECTOR PERCENTAGE STRUCTURE 1975, 1980 and 1S98i Ty;e of Transfer Agriculture Manufacturi.g Other Industry Services Final Denand Total anid Subsidy ________indiustry-/ _____ 1975 1980 1975 15901 997 iT980 1975 1980 1975 1980 1975 1980 1981 Dltert 18.1 6 6 3 3 2.7 3.3 9.7 39.8 25.6 0.4 0.1 64.8 44.7 51.0 Ir.put PrIces 0.:6 2.0 3,4 10.03 3.4 5.3 3.0 13.2 (13.2) 13.5 (2.6) 44.0 37.8 Tax Incentives - - 11.4 3.2 2.7 0.2 1.3 0.2 - - 15.4 4.1 3.4 Firnacial 18.3 2.6 2.8 1.6 0.4 1.9 1.1 1.1 - 0.1 22.4 7.3 7.8 TaTOA- 36.7 11.2 21.1 17.5 9.8 17.2 45.2 40.5 (12.8) 13.6 100.0 100.0 100.0 *2 !i~nufacturing industry here excludes petroleum and derivatives, and basic petrochemicals 3.05 From Table 3.2 and Figure 3.3(a) it can be seen that services remains the sector receiving the most transfers and subsidies, while transfers and subsidies to Final Demand shift from negative in 1975 to positive in 1980, along with a sig- nificant drop in T&S to Agriculture. Industry's share of T&S has increased from 30.9% in 1975 to 34.7% in 1980 (and 36.9% in 1981). The growth of T&S to Industry as a percentage of public expenditures is shown in Figure 3.3(b). 3.06 The share of Manufacturing Industry in transfers and subsidies fell slightly from 21.1% in 1975 to 17.5% in 1980; in absolute real terms however, T&S to Manufacturing Industry increased significantly, more than doubling between 1975 and 1980, as shown in Table 3.3 and Figure 3.4. The share of T&S in sectoral value added rose from 5.8% to 8.8% between the two years. ?dble 3.3: MANUFACTURING INDUSTRY: TRANSFERS AND SUBSIDIES BY SEBSECTOR, 1975 and 1980 (in millions of constant 1975 pesos) Direct Input Prices Tax IncentIves Financial rotal Total as X of Subsector's Value Added 1975 1980 1975 1960 1975 1980 1975 1980 i975 1980 1975 1980 Food Products 1,466 3,406 850 1,965 !68 181 991 1.594 3,475 7,147 6.7 10.9 Beverages and Tobacco 48 - 3 350 723 154 7 12 .81 515 3.4 1.6 Textilea 160 170 11b 273 121 228 19 29 418 701. 2.5 3.2 Clothing, Shoes and Leather - - 115 (77) 50 67 66 91 231 81 1.2 0.3 Wood and Wood Prodaucts - 84 16 (11) 57 57 17 21 90 152 1.1 1.2 Paper and CsrdboarA 22 269 152 963 404 321 22. 34 600 1,588 8.6 14.4 Printing and Publishing 49 2 25 (15) 78 35 31 41 183 62 3.3 0.8 Chealcal tndustry!/ 448 653 310 8,605 358 556 90 143 1,206 9,956 4.3 23.5 Rubber Products - - 29 122 21 38 1 1 51 161 1.2 2.3 Nom-oetallic Hinerals - - 442 2,585 312 805 25 34 779 3,424 5.5 17.6 Baic metals 20 684 490 2,650 1,088 268 204 282 1,802 3,883 12.4 18.4 i.ts1 Products - - (22) (63) 65 90 68 89 iII 115 1.0 0.8 .la,hinery and Eqsipant 68 - 43 (5') 361 475 155 237 627 655 2.9 1.8 Trsnport Equipnent 45 68 95 (27) 4,112 2,029 190 212 4,642 2,282 16.9 13.8 Other Hanufacturin,8 Industries - - (180) 1,239 76 37 69 73 (35) 1,349 (0.6) 18.8 TOTAL . 2,326 5,336 2,486 18,502 7,994 5,341 1,955 2,893 14,761 32,072 5.8 8.8 S l Chemical industr haere excludes petroleus and derivatives, and basac petracheseals.  TRANSFERS AND SUBSIDIES BY SECTOR, t975-t981 2481 218- 2t9- 178- L L t1-: F 1238 IL,W| -0 t e- 118- 7 p Eso a o 78- 48- to PRIMARY a FINAL DEMAND -Is ~ ~ ~ ~ ~ ~ ~~~v o  TRANSFERS AND SUBSIDIES TO INDUSTRY AS PERCENTAGE OF PUBLIC EXPENDITURE !6- Icz-~~ 10- El~~~~~~~~~~~~~~~~~~~~~~~~~PE '~~~~~~~~~~~~~~~~~~~~~~~~~~~C 52 n 11 g | g 11 j~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ S~~~ 0-~t =  MANUFACTURING INDUSTRY STRUCTURE OF TRANSFERS AND SUBSIDIES 1975 AND 1980 TAX I6 . S4 DIRECT 16 76X ~~~~~~~~~~~~~~~~~~~~~~~~F114ANIAL. 0. S2ZC TAX 54.l609 4II INCENTIVES PRIME FINANCIAL MS 24X DI 16. 64X 1975 1980 TOTAL VALUES Nm mILLONNS OF coSAr 07 PESOS 14,701 IN 1075 AND S2.072 IN IOSS  - 18 - 3.07 Table 3.3 compares transfers and subsidies to manufacturing subsectors in 1975 -and 1980. Several subsectors stand out aLs receiving significant transfers and subsidies inD 1980, both in absolute terms and as a percentage of that subsector's value added. Table 3.4 shows the four subsectors receiving the largest absolute value transfers and subsidies in 1975 and in 1980; there are significant changes in the composition, but in both years four subsectors jointly accounted for around 75% of total transfers and subsidies to manufacturing industry. Table 3.4: MANUFACTURING INDUSTRY: SUBSECTORS RECEIVING LARGEST TRA!NSFERS AND SUBSIDIES, 1975 AND 1980 Transfers Transfers and and Subsidies Subsidies (millions of as % of Subsector 1975 pesos) Total 1975 1. Transport equipment 4,442 30.1 2. Food products 3,475 23.5 3. Basic metals 1,802 12.2 4. Chemical Industry 1,206 8.2 10,925 74.0 1980 1. Chemical Industry 9,956 31.0 2. Food products 7,147 22.3 3. Basic metals 3,883 12.1 4. Non-metallic mLnerals 3,424 10.7 24,4:10 76.1 Source: Mission estimates based on data from SHCP 3.08 Seven subsectors received transfers and subsidies accounting for over 10% of the subsector's value added in 1980. These subsectors, ranked by decreasing ratio of T&S to value added, are: Chemical Industry, Other Manufacturing Industries, Basic Metals, Non-metallic Minerals, Paper and Cardboard, Transport Equipment, and Food Products. Analyzin.g the type of transfers and subsidies received by these subsectors, the first five received the bulk of their T&S in 1980 through input prices (from 61% to 92%). in the case of Transport Equipmaent, tax incentives accounted for 895% of T&S received; for Food Products, 48% of T&S were direct, 27% were through input prices, and 22% were finiancial.  - 19 - 3.09 In 1975, by comparison, only 2 subsectors received T&S exceeding 10% of value added: Transport Equipment and Basic Metals. Tax incentives accoLuted for most of their T&S, 93% and 60% respectively. Generalizing, tax incentives accounted for some 54% of T&S to manufacturing industry in 1975, declining to below 17% in 1980. Input prices, on the other hand, provided only 17% in 1975, but in 1980 almost 58% of T&S could be attributed to them. Figure 3.5 illustrates this shift in T&S structure between 1975 and 1980. 3.10 Thus, transfers and subsidies increased significantly in total, to industry, and to manufacturing industry during the period under review; T&S to manufacturing industry more than doubled in real terms between 1975 and 1980 and increased as a percentage of value added in manufacturing although the sector was not the principal beneficiary of the system. The type of transfers and subsidies changed significantly, with input prices becoming the main instrument in 1980, instead of tax incentives which predominated in 1975. At the same time, the system tended to discriminate against primary production. B. Characteristics of the Import Protection System 3.11 Prior to 1976, as noted earlier, import protection was essentially based on prior license requirements with relatively moderate tariff levels. The policy changes introduced beginning in 1977 had as their principal objec- tive a reduction in the relative importance of licensing, although providing a somewhat higher average level of tariff protection. Nevertheless compared to other developing countries that level continued to be relatively moderate. The number ofb items freed from liceTising was increased at various times in subsequent years. By the end of 1979, only about one-fourth of the 7,40C items in the customs nomenclature were subject to this regulation. While those items still accounted for some 60% of import value in 1979, it was also the case that many of these products were capital goods which had been the main category of imports given both the current stage of development of Mexican industry and the investment boom which began in 1978. 3.12 The structure of protection which emerged for 1979 has been analyzed by the Government's working group on this topic. Based on direct price comparisons for a large sample of goods, nominal protection for the economy as a whole is estimated at approximately 0, but for the manufacturing sector (comprising 48 subsectors and excluding petroleum refining) average nominal protection was about 3%. In general, dispersion was relatively small and 11 subsectors, mainly in the processed food category, but some basic petro-chemicals and fertilizers, were subject to negative protection. Of the remaining subsectors, 2C) had rates of protection 0 to 9%, ten from 10 to 24%, six from 25 to 49%, and one was above 50% (electronic equipment, essentially consumer items). (See Table 3.5.)  TOTAL TRANSFERS AND SUBSIDIES SECTOZAL DISTRIBUTION 1975 AND 1980 PRIMARY 12. 72X MANUFACTURIG6 17. On SECTOR MAWUFACIIJUNG 2 t. I I INDUSTRY FINAL 13.18X 10 ~~~DEMNAD PgaIARY 36.74D SECTOR _~ ~~~~~~~~~~~~~~~ _3 FM4/L DEMAt ' SERVICES 40.394X 1975 1980 TOTAL VALUES IN CONSTANT 197S PS 69,918 IN t87 AM 18t,96 aN tOS  - 21 - Table 3.5: DISTRIBUTION OF NOMINAL AND EFFECTIVE PROTECTION FOR MANUFACTURING ACTIVITIESa/ - 1979 Numbers of Product Categories In Rates (%) Nominal Effective Less than -50 1 3 between -49 and -25 2 4 between -24 and 0 8 7 between 1 and 25 31 13 between 26 and 50 5 7 between 51 and 100 1 9 between 101 and 200 0 2 over 201 0 3 TOTAL 48 48 a! Excludes 'petroleum refining. Source: Working Group for the Study of Effective Protection. 3.13 The average effective protection level for the sector was higher, of the order of 11%. There was, however, considerably more dispersion with rates of the order, or Ln excess, of 200% prevailing for a small group of advanced products, largely consumer durables, which includes household appliances, autos, electronic equipment and plastics. The distribution of both nominal and effective protection is compared in Table 3.5. The relatively high level of effective protection vis-a-vis nominal protection reflects the impact of subsidized or controlled prices for a number of inputs such as natural gas, basic petrochemicals and basic foodstuffs. Negative effective protection for a few final goods (e.g. fertilizers) resulted from controlled prices on those items which were not fully compensated for by subsidized inputs. 3.14 While the increased dispersion in effective protection relative to nominal protection reflected the impact of subsidies, it remained true that the structure of nominal largely determined the structure of effective protection. A rank correlation coefficient between the two measures at the disaggregated level is relatively high. 0 3.15 Largely as a consequence of the measures to liberalize imports, nominal protection in 1979 was lower than in 1970, as shown by a comparison of the results of the recent study with those of an analysis undertaken earlier by the same private research group using a similar methodologyl/. In 1970 nominal protection for the manufacturing sector was estimated at 20% as coimpared to 3% in 1979. However, the later results are affected by negative protection resulting from the pricing policies mentioned above. Adjusting the results of the two studies to provide 27 comparable subsectors 1/ See the study of ten Kate, op. cit.  - 22 - covering manufacturing output, 24 had levels of nominal protection in 1979 below those prevailing :Ln 1970 (See Tab:le 3.6). The reductioin in effective protection between the two years, from 56% in 1970 to 11% in 1979, was also substantial, but the dispersion was somewhat greater in 1979. 3.16 In spite of the decline in nominal protection between the two years, the structure of the system showed little change. A correlation coefficient based on the ranking of 27 comparable subgroups for the two years was quite high. A simiLar rank correlation test of effective protection between the two years yielded a somewhat: smaller coefficient, mainly due to the influence of the subsidies on the 1979 structure of protection. 3.17 Examination of the schedule oiE legal tariffs for 1979 indicates that there existed some degree of tariff redundancy. The average unweighted tariff for some 7,000 mtnufactured item; in the customs nomenclature was 25%. Calculations undertaken by the Government's study group for the 48 manufacturing sub-groups indicated that 40 of those groups hadi legal tariffs higher than nominal protection while only 8 had legal tariffs below the levels estimated for nominal protection. Among the latter were four (electric and non-electric machinery, electronic and transporl equipment) which were the main product lines subject to prior license requirements at the time. 3.18 The significanice of tariff redundancy is that prices of locally produced goods could rise relative to the prices of imports. A calculation for 1980 relative to 1979, based on data from the Office of the Economic Advisers to the Presidency, shows an increase in nominal protection between the two years of from 7%' to 10%, i.e. from 3% in 1979 to 10% to 13% in 1980. Moreover, ranking the sub-groups by the ratio of the legal tariff to 1979 nominal protection, and comparing this with price increases for those branches in 1980, yields a high positive! correlation. This means that many enterprises were able tc, compete with imports despite facing rising costs. C. Effective Subsidy 3.19 In the previous sections, the structure and historical development of commercial and subsidly policies have been described. The total impact of these policies is measured by the rate of effective subsidy; this measurement takes into account effective protection,, including net subsidies on inputs, and the remaining subsidLies which are not related to inputs. Calculations of these rates for 1979 are included in the Government's effective protection study. The main components for the prirLcipal groups of manufatcturing industries are summarized in Table 3.7. The table does not iinclude petroleum refining which, if included in the analysis, would distort the results. In accordance with the definition of effective protection, input subsidies are directly estimated in that. calculation since these affect costs of production. Fiscal, financial and direct subsidies, which do not affect production costs, but can affect prices, are shown separately.  - 23 - Table 3.6: COMPARISONS OF ESTIMATE OF NOMINAL AND EFFECTIVE PROTECTION, 1970 AND 1979 BY PRINCIPAL INDUSTRIAL BRANCHES Nominal Protection Effective Protection 1970 1979 1970 1979 Meat and dairy products 12 9 28 98 Flour and flour products 18 -16 44 -28 Other foodstuffs 9 - 7 26 10 Beverages 4 -21 2 107 Tobacco 0 -25 - 4 -27 Textiles (soft fibres) 30 1 45 0 Other textiles 2 2 - 2 0 Clothing, footwear and leather 28 17 38 39 Wood products 8 2 12 19 Paper and paper products 14 20 22 57 Printing and publishing 11 0 12 - 7 Petroleum refining and basic petrochemicals 6 -60 5 21 Basic chemicals 22 9 62 51 Ferti'Lizers 34 -17 1,183 -15 Synthetics, plastics 44 34 258 219 Pharmaceuticals 52 9 236 13 Soaps, detergents, cosmetics 31 4 65 11 Other chemicals 23 16 99 22 Rubber 15 12 15 33 Non-metallic minerals 6 3 6 10 Basic metals 9 3 23 24 Metal products 24 8 46 14 Non-electric machinery & equipment 41 19 67 36 Electric and electronic machinery and equipment 43 37 88 122 Autos and parts 31 28 113 120 Other transport equipment 45 17 91 25 Other manufactures 29 20 80 52 Sources: 1970 estimates from 'ten Kate and Wallace, op. cit. 1979 estimates from Working Group on the Study of Effective Protection! data from these sources have been' aggregated into comparable groups by the mission.  Table 3.7: STRUCTURE OF EFFECTIVE SUBSIDIES, 1979 Traditional Intermediate Modern Manufacturing Products Manufacturing Total A. Total transfers & subsidies (billion pesos) -20.7 38.4 78.4 96.1 1. Effective protection -29.6 23.2 67.2 60.8 a. Nominal protectiona/ -45.5 9.0 78.9 42.4 b. Energy & petroleum inputs subsidies 10.7 16.6 3 . 1 30.4 c. Agricultural inputs subsidies 9.5 0.1 0.1 9.7 d. Intra-Sectoral transfers 6.0 4.0 -8.9 1.1 2. Financial subsidies 2.4 0.9 0.9 4.2 3. Fiscal subsidies 1.8 3.1 5.1 10.0 4. Direct subsidies 4.7 11.2 5.2 21.1 B. Value added at external prices (billion pesos) 329.1 150.8 95. 575.1 C. Effective subsidy - (A/B)% - 6 25 79 17 D. Effective protection rate (AI/B)% - 9 15 69 11 E. Nominal protection (based on price comparisons)% - 5 2 26 3 Source: Working Group on the Study of Effective Protection a/ Other transfers not shown separately in this table include deductions for taxes paid on foreign trade and inter-sectoral transfers.  - 25 - 3.20 On the basis of the estimate of nominal protection, more than half of the gross value of manufacturing output was marketed in Mexico at prices below world market levels; these are the traditional industries which include foodstuffs and non-durable consumer goods. Prices for another one-fourth (intermediates) were only slightly above world levels. On the other hand, the remainder (less than one-fourth) were traded at prices some 25% above those in external markets; these are the so-called modern industries which inclulde electric, electronic and non-electric machinery, transport equipment and advanced metal and plastic products. 3.21 For the traditional industries, the value of these products at external prices exceeds the valuation at domestic prices by more than the amounit of transfers and subsidies it receives2/, as shown in Table 3.7. To maintain negative nominal protection, which was found mainly in the food industries, various administrative measures, including price and export controls, were used. While these branches accounted for more than half of the gross value of output, they received some 40% of the total of indirect input and other transfers and subsidies provided to the manufacturing sector. 3.22 In the case of the industries produci-ng intermediate products, their ability to sell locally at prices more or less equal to world market levels required substantial direct subsidies as well as the subsidized energy and petroleum based inpputs. Basic steel accounted for 80% of the group's direct budgetary subsidLes while the bulk of the remaining 20% went to fertilizers. Most of these branches received energy input subsidies, mainly for rLatural gas and coaL; these had particular significance for non-ferrous metals since energy accounts for a large proportion of production costs. The intra-sectoral transfers reflect the low prices for basic petro-chemicals which benefitted the secondary petrochemical industries (e.g., artificial and synthetic fibers). Overall this group accounted for almost 45% of the total transfers and subsidies received by manufacturing, compared to its share of about 25% of the gross value of output. 3.23 The high level of nominal protection for modern manufacturing associated with prevailing tariffs as well as import restrictions is mainly responsible for the substantial effective subsidy received by these indus- tries. As shown in Table 3.7, the total value of these products in domestic prices exceeded the value in external prices by 78.9 billion pesos (nominal protection, line A.1.a) due to protective measures. The value added in domestic prices exceeded value added in external prices (the basis for measuring effective protection) by 67.2 billion pesos (effective protection, line A.1) with the difference largely due to the high prices paid for some of the cutput of these branches which are inputs for other industries in the same group; energy and input subsidies are not important overall. The higher rate of effective subsidy (79%) as compared to effective protection (69%) results from the relatively substantial fiscal and direct subsidies. However, there were conesiderable variations among the branches involved. While the fiscal subsidies were relatively wide-spread, the bulk of the direct subsidies was received by the public sector's bus and truck assembly plants. Moreover, as noted in the earlier discussion, the dispersion of 2/ Allowance must alsc be made for some higher prices it paid for inputs from the two other manufacturing groups.  - 26 - effective protection within this group was particularly large with rates of over 200% for consumer durables such as autos, domestic appliances and elect:ronic equipment (mainly radios, television etc.). On the other hand, the capital goods producing industries, which were given high priority in sectoral policy, had moderate rates of effective subsidy. D. Resource Allocation 3.24 To what extent did the measures introduced after 1977 affect the flow of private sector investment? National accounts data provide information only on the total gross fixed capital formation with no breakdown for investment in industry by either the public or private sector. Some indications of areas of interest of private sector investors can be obtained from data provided by SHECP on the value of tax reimbursement certificate (CEPROFIs) issued for inavestment expenditures actually undertaken under the incentive law; these are shown in Table 3.8 for the period from the incep- tion of the system (197'9) through the end of 1981. In addition SEPAFIN provides data on the value of investment approved under the incentive law and the expected employment associated witln those projects. These approvals represent intended investment, which is being, or will be carried out. Data on approvals during the period 1979 through 1981 are shown in Table 3.9. Table 3.8: MANUFACTURING INDUSTRIES: CEPROFIS ISSUED FOR ACTUAL INVESTMENT - 1979-81 Amount (million pesos) % of Total Food, Beverages & Tobacco 387.1 3.5 Textiles, Shoes and Clothing 581.7 4.7 Wood Products 223.9 2.0 Paper & Printing 329.5 3.0 Chemical Products 1,204.4 11.0 Rubber Products 157.4 1.4 Petroleum & Coal Derivatives 65.6 0.6 Non-Metallic Minerals 4,399.9 40.3 Basic Metals 2,298.7 21.0 Metal Products 230.7 2.1 Non-Electric Machinery & Equipment 498.0 4.6 Electric & Electronic Machinery 216.6 2,0 Transport Equipment 381.7 3.5 Other Manufactures 16.7 0.2 TOTAL 10,929.0 100.0 Source: SHCP  Table 3.9: VALUE OF INVESTMENT APPROVED UNDER INVESTMENT INCENTIVE LAWS a/ BY PRIORITY CLASSIFICATION - 1979 - 1981 Expected Increases Investment Total Value of Investment in Employment Per Job Value in Current Prices b/ % of % of Thousands US$/ Million Pesos Million US$ - Total Numbers Total Job Category I 128,539.2 5,488.4 52.5 79,680 43.3 68.9 Agro-industries/ 42,657.0 1,821.4 17.4 34,195 18.6 53.3 Capital goods 35,759.0 1,526.9 14.6 39,238 21.3 38.9 Strategic inputs d/ 50,123.2 2,140.2 20.5 6,247 3.4 342.6 Category II 116,091.4 4,956.9 47.5 104,491 56.7 47.4 Non-durable consumer goods 5. 32,144.9 1,372.5 13.1 31,326 17.0 43.8 Durable consumer goods LI 32,705.9 1,396.5 13.4 58,245 31.6 24.0 Intermediates./ 51,240.6 2,187.9 20.9 8.1 146.6 1 Grand Total 244,630.5 10,445.4 100.0 184.171 100.0 56.7 Source: SEPAFIN a/ The classification is given in the decree of Maich 9, 1979;with minor modifications introduced subsequently, b/ Converted at 23.42 pesos to the US dollar, the unweighted average exchange rate for years 1979 through 1981. c/ Includes mainly processed foodstuffs,inputs for the agricultural sector and agricultural products for industrial uses. d/ Includes basic iron and steel and cement. e/ Mainly textiles, garments, shoes, and leather and related proaucts. f/ Includes mainly domestic appliances and automotive parts. /I Includes petrochemicals, chemicals, non-ferrous metals and construction materials,  - 28 - 3.25 The SHCP data on tax rebates for actual investment which is classified by 18 major industrial branches, show that the total value of such investment during the three years was equivalent to around US$2.4 billion (asstuming the average incentive was 17%). Over 70% of the investment was accounted for by three categories: chemicals, basic metals and non-metallic minerals. These are the branches producing intermediates which have been shown to be the principal beneficiarieis of energy and input subsidies; these are also among the most capital intensive.3/ 3.26 The SEPAFIN data are classif:Led according to the major categories of priorities under the incentive legislation which cuts across standard product category lines. The total valuie of the projects approved was equiva- lent to approximately US$10.4 billion, four times the value of the investment actually undertaken as measured by the SHCP data. The distribution of investment costs between the two main categories was almost equal, 52.5% for Category 1 and 47.5% for Category 2. Hlowever, the latter group is expected to generate more employment per unit of investment, 56.7% of the total new jobs as compared with 43.3% for Category 1. 3.27 More than 40% of investment approvals are accounted for by the intermediate "strategic" inputs in Category 1 (iron and steel and cement) and the group within Category 2 which includes petrochemicals, chemicals, non- ferrous metals and construction materials (metal products, glass and wood products). For the first group, investment costs per job created were expected to be of the order of US$340,0)00, while for the second group such costs are of the order of US$150,000; these can be compared to averages ranging from US$24,000 of US$53,000 among the other four categories. The average investment per job created for all approved projects is approximately US$56,700. This may be compared to the average cost per job for FONEI- financed projects which in 1980 was approximately US$40,000. 3.28 The Bank of Mexico has recently publiLshed estimates of gross fixed capital formation in aLl sectors of the! economyT excluding agriculture and education, based on a sample survey of existing enterprises, including parastatals. The significance of these data, however, is not clear due to uncertainty on coverage of new firms. What is important, however, is the relative importance of such expenditures by the textile and related industries, considerably higher than that emerging from the dlata for investment under the incentive laws which covered new plants and expansions. In addition, imports of specialized machinery for the textile industry represented in the period 1979 through 1981 one of the largest categories within industrial equipment totalling close to US$1 billion in the three years. Trade sources suggest that expansion in this sub-sector took place without use of the fiscal incentives as disadvantages associated with the administration of the law were considered to offset the advantages. At the same time, however, substantial capital expenditure was undertaken for improvement and modernization and this may well have reflected the need to become more competitive given the reduction in input growth. 3/ Data from the input-output matrix for 1975 was analyzed to determine if there was any correlation between capital-intensity and labor-intensity; the results were not very significant. In part this may reflect the fact that low prices were resulting in an inefficient use of energy.  - 29 - 3.29 While these various indicators are not completely comparable, they suggest a rather complex pattern of private sector investment. On the one hand, considerable investment in capital-intensive industries has been undertaken responding to the set of energy and input subsidies and fiscal incentives. On the other, industries facing competition from imports because of import liberalization undertook substantial expenditure to improve quality of output, with some expansion of capacity. Finally, in spite of the priority given to agro-industries, little interest was shown by the private sector in new investment in these activities, reflecting negative protection and in general the extent of price and other controls on both inputs and outputs. 3.30 The second element in investment policy is direct public sector investment in industrial facilities where decisions are not subject to the impact of incentives on relative profitability. Available data based on budgets or planned expenditure are summarized in Table 3.10. Excluding PEMEX, planned investment declined sharply after completion of the first stage of the Las Truchas steel mill in 1974-76. Planned investment rose again after 1978, due to an expansion of fertilizer plants, and the undertaking of substant:Lal expenditures for food industries, a large part of which was for the modernization and expansion of the sugar industry; public sector plants provide more than 60% of the output of sugar and related by-products. There is iLndication of a further substantial growth in planned expenditure in 1981 and at the beginning of 1982 with the initiation of a large number of projects which had been in the planning stages. These include the second stage of Las Truchas, a heavy foundry-forge facility and a steel pipe mill. 3.31 The investment: activity of PEMEX, which amounted to US$6.4 billion in 1980, is estimated to have represented some 15% of total gross fixed capital formation. The preliminary estimate for 1981 is for US$7.6 billion, about the same proportion of total investment. It is not possible to identify for all years the components representing petroleum refining and petrochemical production which fall within the manufacturing category. The PEMEX budgetary presentation for 1980 indicated that these two categories were expected to account: for 8.7% and 13.5% respectively of planned total investment during that year which in total would amount to US$1.6 billion of actual investment if these percentages planned are applied to actual investment. 3.32 More detailed analysis is required to evaluate both the appropriateness of this allocation of investment resources and the efficiency of the Mexican petrochemical development program. It is clear, however, that this investment is highly capital intensive and that it is linked to private sector investment programs in secondary petrochemicals which not only were also capital intensive but whose profitability depended upon input prices substantially below those prevailing in world markets.  - 30- Table 3.10: BUDGETED DIRECT PUBLIC SECTOR ][NVESTMENT IN MANUFACTURING 1970 - 1980 PEMEX-a Other Manufacturing Million Million MilliorL Million Pesos US',b/ Pesos US$b/ 1970 2,433 194.6 n.a. n.a. 1971 4,729 378.3 874 69.9 1972 4,723 377.8 1,327 106.2 1973 7, 040 563.2 2,1592 215.4 1974 9,221 737.7 5,935 474,8 1975 13,917 1,11.3.4 12,579 1,006.3 1976 24,048 1,558.9 12,534 813.9 1977 38,283. 1,696.0 9,1591 429.4 1978 69,533 3,054.1 8,339 366.2 1979 103,506 4,5-38.7 13,674 599.8 1980 147,836 6,441.4 25,885 1,127.9 Source: Mission estimates based orL Budget P'resentations.. a/ Data for PEMEX include all investment activities for exploration, oil and gas extraction, transportation fac:ilities, refining and petrochemical production. b/ Converted at an average exchange rate for the year. For 1976, the average was trade weighted, equal to 15.4 pesos to the dollar.  - 31 - IV. INDUSTRIAL PERFORMANCE AND STRUCTURE A. Introduction 4.01 In this chapter, it is proposed to examine the performance of in- dustry from 1977 through 1981 in order to determine the extent to which the changed economic environment affected the sector. The relationship between these developments and the policy framework are under the best of circum- stances difficult to quantify. Economic phenomena are multi-dimensional as well as interrelated. Although much progress has been made in advancing the ability to measure and to understand these developments, there are many aspec:ts which simply cannot be captured by the tools available at this time. Moreover, while Mexican statistics have improved considerably in recent years, there remain major gaps which restrict analysis or lead to the use of imperfect surrogates. The principal shortcomings are the inadequate data on industrial investment and capacity utilization, as well as the availability of only limited information on employment in the sector and on the volume of manufactured imports and exports. 4.02 Aside from these problems, there are questions of time in determin- ing the extent to which economic analysis can provide reasonable conclusions as to cause and effect. Firstly, one must ask whether the policies had been in effect long enough to have brought about changes in behavior which could be obiserved. Secondly, one must also consider whether enough time had lapsed so that any new investment induced by the new environment would have begun to bear fruit. While the answer to the first would probably be positive, it is precisely the lack of investment detail which limits the capability to measure effect at this time. The answer to the second, on the other hand, would appear to be negative but in this case, given the present situation, it may be extremely difficult in any future analysis to unravel the sequence of events. B. Overall Performance of Manufacturing, 1970-1980 4.03 During the decade of the 1970s, economic activity and in particular manufacturing productioni experienced rather sharp cyclical fluctuations with three relatively clearly defined periods of change (Tables 4.1 and 4.2). In the years 1970 through ]L974 manufacturing output grew at about 7% p.a., ap- proximately equal to th! average rate of growth of the previous two decades, although slightly below the 8% p.a. level recorded during the 1960s. These four years represented a period of world-wide economic expansion which led to a relatively substantial rise in Mexican exports including manufactured pro- ducts while overall domestic demand continued its past pattern of growth, close to 7%. 4.04 At the end of 1974, however, there were major changes in both the external and domestic economic environments. T'he recession which affected the industrialized countries following the first oil price rise at the end of 19/3 led to a fall in Mexico's exports of manufactures as well as of other exportable goods. This loss of dynamism in the external sector, the inabil- ity to contain domestic inflation which exceeded international inflation, and policy measures which adversely affected private sector expectations  Table 4.1: GROSS DOMESTIC PRODUCT - BY EXPENDITURE 1970, 1974, 1977-1981 1970 1974 1977 1978 1979 1980 1981 In current prices (billion pesos) 444.3 899.7 1,849.3 2,337.4 3,067.5 4,276.5 5,858.2 Consumption of public administration 32.2 82.3 199.0 255.2 334.3 462.8 - Private consumption 319.5 628.3 1,226.1 1,543.8 1,975.9 2,651.5 - Changes in stocks 12.3 29.7 59.1 59.2 77.6 169.8 - Gross fixed capital formation 88.7 178.9 363.3 492.4 718.5 1,032.9 - Exports of Goods and Services 34.4 75.7 190.8 244.7 343.3 537.2 - Less imports of goods and services -42.9 -95.2 -189.0 -258.0 -382.0 -577.8 - As % of total 100.0 100.0 100.0 100.0 100.0 100.0 - Consumption of public administration 7.3 9.2 10.8 10.9 10.9 10.8 - Private consumption 71.9 69.8 66.3 66.0 64.4 62.0 - Changes in stocks 2.8 3.3 3.2 2.5 2.5 4.0 - Gross fixed capital fo- mation 20.0 19.9 19.6 21.1 23.4 24.1 - Exports of Goods and Services 7.7 8.4 10.3 10.5 11.2 12.6 - Less Imports of Goods and Services -9.7 -10.6 -10.2 -11.0 -12.4 -13.5 - Tn constant (1970) nrines (billion peso8) 444=3 577.6 657.7 711c9 777.2 841.9 1,053.5 Consumption of public administration 32.2 47.3 56.8 62.4 68.4 75.0 82.5 Private consumption 319.5 402.4 453.8 490.8 534.2 574.5 616.7 Changes in stocks 12.3 22.5 23.0 21.7 21.7 38.6 46.4 Gross fixed capital formation 88.7 121.1 124.0 142e8 171.7 197.4 226.4 Exports of Goods and Services 34.4 47.5 57.8 64.5 72.3 76.7 81.5 T -ssT --m-rt-s of Goods and Servi c -4.9 a. -63.3 -,7~ .7 - 7 0 .2. -7 -;2- -.7 Source:1. Sitm de