Document of Tlie World Bank 6479 v2 FOR OFFICIAL.USE ONLY CONFIDENTIAL Report No. 6479-IN • INDIA lNDUSTRIAL REGULATORY POLipY STUDY VOLUME II Subsector Reports December 9, 1986 . Industrial Strategy & Policy Division Industry Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its rontents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENT November 1986 US$1.00 = Rs 12.85 • December 1985 US$1.00 = Rs 12.50 December 1984 US$1.00 = Rs 12.36 December 1983 US$1.00 = Rs 10.49 • December 1~82 US$1.00 = Rs 9.63 December 1981 US$1.00 = Rs 9.10 June 1975 US$1.00 = Rs 7.81 ACRONYMS ACC Associated Cement Companies ACMA Automotivf.: Component Manufacturers' Association BTU British Thermal Unit CCI Chief Controller Imports CDOT Center for the Development of Telematics CIF Cost plus Insurance and Freight CKD Completely I~nocked Down CMA Cement Manufacturers Association CMC Computer Maintenance Corporation CNC Computer Numerical Control CV Commercial Vehicle DCPL Development Consultants Private Limited DGTD Directorate General of Technical Development DCA. Department of Company Affairs DOE Department of Electronics DOT Department of Telecommunications EGEAF Engineering Goods Exports Assistance Fund ETTDC Electronic Trade and Technology Development Corporation FAI Fertilizer Association of India FCI Fertilizer Corporation of India FERA Foreign Exchange Restrictions Act FICC Fertilizer Industry Coordination Communications FOB Free on Board GOI. Government of India • HCL Hindustan Cables Ltd. HMT Hindustan Machine Tools Ltd. HPC Hindustan Paper Corporation ' HTL Hindustan Teleprinters Ltd. ICICI Industrial Credit and Investment Corporation of India !CB International Competitive Bidding • ICP Indian Crossbar Project IISCO Indian Iron and Steel Company !TI Ind~an Telephone Industries Ltd. JPC Joint Plant Committee LD Linz Donawitz MES Minimum Efficient Scale MODVAT Modified Value Added Tax .- FOR OFFICIAL USE ONLY CONFIDENTIAL .. INDIA ,. INDUSTRIAL REGULATORY POLICY STUDY VOL. II Subsector Reports TABLE OF CONTENTS Page No. 1. PULP AND PAPER l ······················~·················· A. Background ••••••••••••••••••••••••••••••••••••••••• 1 B. The Regulatory Environment ••••••••••••••••••••••••• 3 l. Licensing Policies ••••••••••••••••••••••••••••• 3 2. Price and Production Controls •••••••••••••••••• 4 3. Tax Policies ••••••••••••••••••••••••••••••••••• 6 4. Infrastructural and Raw Material Constraints ••• 1 5. Trade Policy.••••••••••••••••••••••••••••••••••• 9 c. Impact of the Regulatory Policies •v•••••••••••••••• 10 l. Market Structure ••••••••••••••••••••••••••••••• 10 2. Economics of Vertical and Horizontal Integration •••••••••••••••••••••••••••••••••• 12 3. Firm Behavior •••••••••••••••••••••••••••••••••• 13 4. Impact of Policies on P~rformance •••••••••••••• 16 II. CEMENT • •••••••••••••••••••••••••••••••••••••••••••••••• 19 A. Background ••••••••••••••••••••••••••••••••••••••••• 1-9 B. The Regulatory Environment: 1942-82 ••••••••••••••• 21 1. Licensing Policies ••••••••••••••••••••••••••••• 21 2. Price Controls ••••••••••••••••••••••••••••••••• 21 3. Distribution Controls ••••••••••••••••••••••••;• 22 4. Infrastructural Constraints (Energy, Coal, Transport) •••••••••••••••••••• 22 5. Impact on Market Structure and Perfo~nce ••••• 23 c. Recent Policy Moves: 1982-85 •••••••••••••••••••••• 25 l~ Impact on Structure and Behavior ••••••••••••••• 26 2. Impact on Performance •••••••••••••••••••••••••• 29 This document has a resmcied distribution and may be used b~ recipients only in the perfori.nance of their official duues. Its con&ents may not otherwise be disclosed without World Bank authonzauon. TABLE OF CONTENTS (Continued) Page No. III. NITROGENOUS FERTILIZERS • ••••••••••••••••••••••••••••••• 31 A. Introduction ••••~•••••••••••••••••••••••••••••••••• 31 B. Regulatory Policy Framework •••••••••••••••••••••••• 34 • 1. Licensing•••••••••••••••••••••••••••••••••••••• 34 2. Pricing •••••••••••••••••••••••••••••••••••••••• 35 3. Exit Policy •••••••••••••••••••••••••••••••••••• 36 4. Import Policy •••••••••••••••••••••••••••••••••• 36 5. Regulation of Public Sector Firms •••••••••••••• 37 C. Impact of Regulatory Policies •••••••••••••••••••••• 38 1. Structure •••••••••••••••••••••••••••••••••••••• 38 2. Behavior ••••••••••••••••••••••••••••••••••••••• 40 3. Performance •••••••••••••••••••••••••••••••••••• 42 IV. STEEL AND~••••••••••••••••••••••••••••••••••••••••• 45 A. Background ••••••••••••••••••••••••••••••••••••••••• 45 a. Current Regulatory Environment ••••••••••••••••••••• 48 c. Industrial Structure and Sehavior •••••••••••••••••• 51 D. Impact on Industrial Performance ••••••••••••••••••• 52 1. Capacity Expansion ••••••••••••••••••••••••••••• 52 2. Capacity Utilization ••••••••••••••••••••••••••• 52 3. Technology ••••••••••••••••••••••••••••••••••••• 53 4. Efficiency of Input Use •••••••••••••••••••••••• 53 5. Labor Productivity ••••••••••••••••••••••••••••• 54 6. Competitiveness •••••••••••••••••••••••••••••••• 55 v. CAPITAL GOODS ............... ........................ . ~ 57 A. Ba~kground ••••••••••••••••••••••••••••••••••••••••• 57 B. Regulatory Pol~cies •••••••••••••••••••••••••••••••• 58 1. Licensing•••••••'•••••••••••••••••••••••••••••• 58 2. Pricing Policies ••••••••••••••••••••••••••••••• 60 3. Trade Policies and Incentives •••••••••••••••••• 60 4. Technology Policy •••••••••••••••••••••••••••••• 61 C. Structure of Industry ••••••••••••••••••••••••••••~• 62 1. Firm Size and Concentration •••••••••••••••••••• 62 2. Subcontracting••••••••••••••••••••••••••••••••• 63 3. Ownership Patterns ••••••••••••••••••••••••••••• 68 D. Impact of Policies on Behavior ••••••••••••••••••••• 68 E. Impact of Policy on Performance •••••••••••••••••••• 70 I 1. Growth in Output ••••••••••••••••••••••••••••••• 70 2. International Competitiveness •••••••••••••••••• 73 3. Profitability of Capital Goods ••••••••••••••••• 73 F. Recent Policy Changes •••••••••••••••••••••••••••••• 74 G. Conclusions ••••··~··••••••••••••••••••••••••••••••• 75 TABLE OF CONTENTS (Continued) Page No • • VI. MACHINE TOOLS • ••••••••••••••••••••••••••••••••••••••••• 77 A. Background ••••••••••••••••••••••••~•••••••••••••••• 77 B. Regulatory Policies •••••••••••••••••••••••••••••••• 80 c. Impact on Industrial Development ••••••••••••••••••• 82 1. Impact on Market Structure and Performance ••••• 82 2. Competitiveness and Economics of Scale ••••••••• 85 3. Technology and Product Development ••••••••••••• 87 VII. AUTOMOTIVE PRODUCTS • ••••••••••••••••••••••••••••••••••• 91 A. Background ••••••••••••••••••••••••••••••••••••••••• 91 B. The Regulatory Policy Environment •••••••••••••••••• 93 1. Capacity Licensing••••••••••••••••••••••••••••• 93 2. MRTP Act and FERA •••• ·• ••••••••••••••••••••••••• 96 3. Small-Scale Industry Reservation ••••••••••••••• 103 4. "Sick Industry" Policies ••••••••••••••••••••••• 104 c. The Impact of the Regulatory Environment on Performance •••••••••••••••t•••••••••••••••••••••• ios o. Recent Policy Changes and Modernization •••••••••••• 107 VIII. TELECOMMUNICATIONS EOUIPMENT • •••••••••••••••••••••••••• 111 A. Background ••••••••••••••••••••••••••••••••••••••••• 111 B. The Impact of Past Policies •••••••••••••••••••••••• 114 C. Recent Policy Moves •••••••••••••••••••••••••••••••• 120 D. Summacy • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 125 IX. ELECTRONICS • ••••••••••••••••••••••••••••••••••••••••••• 129 A. Background ••••••••••••••••••••••••••••••••••••••••• 129 B. Regulatory Policies Prior to Recent Changes •••••••• 130 C. Impact on Industrial Development ••••••••••••••••••• 132 1. Role of the Public Sector •••••••••••••••••••••• 132 2. Limited Use of Foreign Collaboration and Technology ••••••••••••••••••••••••••••••• 133 3. Fragmented Production Base and Limited Foreign \ Trade ••••••••••••••••• • • •• • • • • • • •• • • • • • • • • • • • 135 4. Economics of Scale and International Competitiveness •••••••••••••••••••••••••••••• 137 D. Recent Policy Changes and the Supply Response •••••• 140 TABLE OF CONTENTS (Continued) Page No. X. COTTON TEXTILES ••••••••••••••••••••••••••••••••••••••• 145 A. Introduction ••••••••••••••••••••••••••••••••••••••• 145 B. The Policy Framework ••••••••••••••••••••••••••••••• 147 1. Controls on Raw Materials •••••••••••••••••••••• 147 2. Controls over Spinning ••••••••••••••••••••••••• 148 3. Controls over Weaving Operations and Cloth ••••• 148 C. The Impact of the Regulatory Framework on Structure and Performance ••••· •••••••••••••••••••• 151 1. Spinning ••••••••••••••••••••••••••••••••••••••• 151 2. Weaving •••••••••••••••••••••••••••••••••••••••• 155 D. Policy Options: The New Textile Policy and What Needs to be Done •••••••••••••••••••••••••••• 160 XI • CLOTHING ••••••••••.• •••••••••••••••••••••••••••••••••••• 163 A. 'Background ••••••••••••••••••••••••••••••••••••••••• 163 1. The Market ••••••••••••••••••••••••••••••••••••• 163 2. Structure of the Subsector ••••••••••••••••••••• 167 3. Variability of Demand and Labor Policies ••••••• 167 B. Performance •••••••••••••••••••••••••••••••••••••••• 172 , INDIA - INDU~TRIAL REGULATORY POLICY STUDY ~ I. Subsector Review--Pulp and Paper A. BACKGROUND 1.01 India's pulp and paper industry was essentially self-sufficient by 1940, even before India's independen·~e was declared and paper was classified as a "core" industry by the government. As early as 1918, the industry had succeeded in establishing a bamboo-based paper mill, a~d by 1938 the use of bamboo pulp w~s more economical than the tariff-levied imported pulp. The 1940s witnessed the entry of the large Indian houses into the industry (Birla, Thapar and Bajoria). Under the protection of an import-substituting regime, the industry enjoyed rapid growth through the 1950s into the 1960s. 1.02 Beginning in the late sixties, the industry entered a difficult period. In addition to the imposition of price controls from 1961 to 1968, the industry faced rising energy costs and scare~ raw materials.I/ Expansion of capacity fell from an annual rate of 7.8 percent during 1960-68 to 5.6 percent during 1968-79, while output growth fell to half its previous level, from 8.1 percent to 4.3 percent (Table 1). The percentage of production devoted to printing paper also fell, as paper mills shifted to the production of other grades of paper for a better return. 1.03 These trends, coupled with a worldwide shortage of paper, contri- buted to the government's rising concern that paper supply constraints would curtail the country's educational and commercial development. The 1974 Production Control Order was implemented, requiring large mills to devote 57 percent of their production to six varieties of paper, which were subject to informal price controls. This order was modified by the Paper Control Order of 1979, which imposed statutory price controls. 1.04 Since 1979, policies regarding the paper industry have remained essentially unchanged but have become the focus of considerable contro- versy. Administered prices, particularly in the 1970s, were not frequently revised and have hovered as low as 50 percent of free market prices. The • combination of price controls, escalating costs of r~w materials, and infrastructural bottlenecks have led to a persistant decline in profits, which in turn have provided disincentives to further investment by larger firms. Moreover, by 1985, a number of firms were using scarce raw materials inefficiently and needed modernizing. Thus, the industry is caught in a classic vicious circle: outdated technology contributes to high operating costs and inefficient use of scarce resources, yet the high costs and controlled low prices do not allow firms to generate sufficient prof its to modernize. ]:_/ On a per capita basis, India has only one-tenth the forest area of the rest of the world. - 2 - 1.0S In a parallel development, the government has sought to promote additional investment through substantial fiscal incentives. These incentives, which have included exemption from the paper control order and relief from excise taxes, have been targeted at small mills and mills using unconventional raw materials.2/ The GOI believed these mills would require less capital investment, could begin operating sooner, and would be able to exploit better the relatively small quantities of unconventional materials available. 1.06 In response to these measures, the number of small and medium- sized mills with capacities of less than 20,000 tons per year (tpy) mushroomed between 1979 and 1985, rising from 30 percent to 50 percent of total installed capacity. Because of a number of factors, ho'l>'ever, these mills have not performed as well as expected. As a result, financial institutions have been reluctant to provide additional financing, and growth in capacity of the smaller mills slowed substantially in 1985-86. 1.07 To counter the lack of investment by large mills, the GOI entered the industry in the 1970s through the establishment of the Hindustan Paper Corporation (HPC). Until the entry of HPC, which is expected to achieve a 10 percent market share in the next several years, the sector was dominated almost entirely by the private sector. Although a projected supply shortfall of over a million tons annually by the year 2000 indicates that additional capacity is needed, the desirability of public sector expansion is unclear. 1.08 This paper examines the effect of government policies (price controls, taxation and licensing) on the industry's structure, behavior and performance. The focus is on writing and printing paper, industrial paper and paperboard. The analysis is complicated by the fact that raw material constraints have played a central role in the problems faced by the industry. 2/ Defined.as anything other than hardwood, softwood, bamboo, reeds.and rags. -· 3 - '!'able 1: CAPACITY AND PRODUCT!ON OF PAPER AND PAPERBOARD, 1960-85 Capacity Growth of Output Growth of Capacity Year ('000 tons) Capacity ('000 tons) Output Utilization (%) (%) (%) 1960 400 345 86.25 1961 410 2.50 364 5.51 88.78 1962 434 5.85 387 6.32 89.17 1963 502 15.67 464 19.90 92.43 1964 556 10.76 490 5.60 88.13 1965 644 15.83 335 -31.63 52.02 1966 644 o.oo 377 12.54 58.54 1967 701 8.85 609 61.54 86.88 1968 730 4.14 645 5.91 88.36 1969 768 5.21 706 9.46 91.93 1970 868 13.02 758 7.37 87.33 1971 901 3.80 781 3.03 86.68 1972 954 5.88 803 2.82 84.17 1973 962 0.84 797 -0.75 82.85 1974 1,069 11.12 837 5.02 78.30 1975 1,069 o.oo 829 -0.96 77.55 1976 1,040 -2. 71 880 6.15 84.62 1977 1,137 9.33 936 6.36 82.32 1978 1,265 11.26 1,005 7.37 79.45 1979 1,380 9.0Q 1,047 4.18 7 5.87 1980 1,538 11.45 1, 112 6.21 72.30 1981 1,656 7.67 1,235 11.06 74.58 1982 1,816 9.66 l,23f. 0.08 68.06 1983 1,915 5.45 1,168 -5.50 60.99 1984 2,195 14.62 1,350 15.58 61.50 1985 2,350 7.06 1,500 11.11 63.83 1960-70 8.06 8.19 1970-80 5.89 3.91 1930-85 8.85 6.17 Sour:e: Development Consultants Private Limited (DCPL), "Development of --- Paper Industry in India," 1985 • .. B. THE REGULATORY .ENVIRONMENT 1. Licensing Policies 1.09 Licensing policies have not been a deterrent to entry or expan- sion in paper production, although iter.'S produced from industrial paper and paperboard (bags, cartons, envelopes, etc.) have been reserved for small-scale industry. The GOI, through the DGTD, has traditionally been - 4 - liberal in granting letters of intent to pulp and paper mills, since many units fail to survive. 3/ Although the Monopolies and Restrictive Trade Practices Act (MRTP) applies, these companies do not seem to have any problems getting clearances for expansion. Despite the liberal licensing policies, however, entry and expansion into the industry (except for small units) have been limited because of other constraints, including price and production controls, high excise duties and infrastructural/raw material shortages. 2. Price and Production Controls 1.10 The paper industry has been subject to price and production controls on and off since 1959. Price controls were imposed between 1961 and 1968, and reintroduced under the Paper Control Order of 1974. A general shortage of paper in the early seventies led to steep price increases, and the government wished to ensure the availability of certain educational grades of paper. Under the order, informal price controls on white printing p~per were set at rupees (Rs) 2,750 per metric ton. Over 57 percent of total paper and paperboard production was also controlled: 30 percent of total production would be devoted to white printing paper, 16 percent to cream laid or wove paper and 11 percent to 5 other specified varieties. The 1979 Paper Control Order established a statutory price for white printing paper, to be delivered to specified consumers of educational paper at Rs 3,000 per metric ton; it also set the price for cream laid or wove paper at Rs 3,785 per metric ton. Sixty-three percent of total production by large and medium fit'lllS with licensed capacity of over 25 tons a day or 8,000 tons a year was controlled by the government, with at least 30 percent of production to be devoted to white printing paper and 20 percent to cream laid or wove paper. 1.11 The current policy, established iu April 1983, is a liberalized version of earlier policies. Large manufacturers, now defined as those with an installed capacity greater than 24,000, must devote 20 percent or 25 percent of their production to white printing paper. As of January 1986, the price for that product was set at Rs 7,200 per metric ton. From 1979 through 1982, units using unconventional raw materials as well as new units have been granted complete exemptions, but as of 1983, total exemp- tions have been granted to only small manufacturers and to units suffering financial ~ar~ship. New llllits get partial exemptions, but 5 percent of their production must be white printing paper in·the first year, 10 percent in the se:ond, etc., up to 25 percent in the fifth and subsequent years. 1.12 The Paper Control Orders have been the subject of constant criticism (particularly by the paper industry) since 1974. The primary complaint pertains to the administered prices, which have not kept pace with the rise in production costs. Even the most recent price revision, which raised the price from Rs 6,400 to 7,200 per ton, is reported to be Rs 1,000 to 1,200 less than the production costs of cream wove paper for '}__/ It has been estimated that for every 10 letters of intent issued, only 1 unit actually begins production. - 5 - most mills. The prices are in fact about Rs 2,000 per ton below the free market prices. Historically, the set prices have been revised infrequent- ly, remaining unchanged between 1975 and 1978 and again between 1981 and 1983. As a percentage of the average sales realized by large firms, the set price has varied between 50 percent and 89 percent.:t 1.13 Despite the imposition of price controls, prices of Indian paper products compare poorly to international prices for similar products. Tariff protection and lack of import competition have not encouraged producers to operate as ef f icie~tly as abroad. Yet at the current level of efficiency, price controls are set too low to generate profits which could be used to finance modernization. The solution to this dilemma is a strategy which combines free market pricing with reduced protection from import competition. While price liberalization should remove the disincen- tive to expand and T110dernize, import liberalization should provide the impetus to benefit from resulting higher profits to increase production efficiency. 1.14 A number of mills have reduced their supply quotas for the mandated paper by obtaining interim stay orders from the high courts; in some cases they have been given interim relief in the form of higher prices on the grounds that the controlled price was uneconomic. Consequently, the Paper Control Order actually applies to only 46 percent of installed capacity. A number of mills in that 46 percent have either stopped produc- tion or been granted exemptions from the allocation, a situation that has narrowed the figure to 25 percent. This segment, whose obligation is only 20-25 percent of production, supplies the entire quota of the educational sector. Apart from the exempt units, t~e order applies to all large units, including those not geared to the production of writing and printing paper •.~/ 1.15 The compulsory yroductic·n has led to an increase in costs as W'ell as to a substantial reduction in capacity utilization for large units. Overall, the order has weakened the financial health of these firms, whose profitability for the most part has declined over the last 10 years.6/ This decline has been noted at all levels and pertains to net operatir.g and total profits, as well as to returns on the funds of equity shareholders (Table 2). Moreover, the production of writing and printing paper as a proportion of total production in the paper industry declined from 56 percent in 1971 to 45 p.ercent in 1983. In addition to jeopardizing the !±./ Average sales are based on the average of the administered and free prices received by each firm. These figures are taken from che Ferguson study (1984), which surveyed over 50 percent of the large firms in the industry. 'if The Indian Paper Mills Associ~tion argued in its president's December 1985 address that this situation was inequitable. §.../ See A. F. Ferguson, "A Study Relating to the Financial Health of Paper Industry in India with Reference .to Large Integrated Paper Un~ts," 1969. - 6 - GOI's original objectives, the order also appears to ngve encouraged informal market activities. Although it is impossible to measure the extent to which this has occurred, a number of agents who obtain white printing paper for educational purposes at levy prices resell it at free market prices. Table 2: TRENDS IN PROFITABILITY --------Year------- Item 1973-74 1982-83 Contribution to Net Sales (%) 46.0 28.0 Operating Profit Before Depreciation, Interest and Taxes (OPDIT) to Net Sales (%) 23.0 8.0 Operating Prof it Before Interest and Taxes (OPBIT) to Net Sales (%) 17.0 2.0 Operating Profit Before Taxes (OPBT) to Net Sales (%) 14.0 -5.7 Profit After Taxes (PAT) to Net Sales (%) 9.1 -5.6 Return on Total Assets (%) 15.3 1.5 Source: A. F. Ferguson and company. 3. Tax Policies 1.16 Pulp and paper mills producing all varieties of paper (with the exception of newsprint) are charged a variety of excise duties. From 1977 to 1983, excise taxes for writing and printing paper were set at 25 percent (ad valorem) of sales. For paperboard and packing an~ wrapping paper, they ranged from 30 percent to 40 percent. 1.17 In 1983, excise duties were made uniform and changed to 10 percent ad valorem plus a lump sum of Rs 2,500 per metric ton. In 1984, the lump sum was reduced somewhat, but the present excise duties generally are high. The cumulative effect of excise duties, state and local taxes, - 7 - and other charges range from 35 percent to over 50 percent of the basic ex-mill price. 7; 1.18 Excise duties are not applied uniformly across all firms, and a number of units are granted large concessions. The small units receive an excise exemption between 50 percent and 75 percent and units using uncon- ventional raw materials are either partially or totally exempt. New units using conventional raw materials to produce printing and writing paper have an exemption of 50 percent. 1.19 Large producers have complained that the excise taxes are too high 8/ and that they must bear a substantial part of the tax burden, as demand is relatively price-elastic. Their contention is substantiated by the sluggish growth in consumption beginning in the 1970s (Table 3). Growth in consumption of paper and paperboard fell from 7.4 percent during the 1960s to 4.7 percent between 1970 and 1980. 1.20 During the Sixth Plan period (1980-85), the 3.1 percent growth in consumption was much lower than the annual target of 7 percent annually, but despite the low levels of capacity utilization, there was no shortage of paper (except for newsprint, which continued to be imported in large amounts). In 1985, enterprises were facing financial difficulties because of an accumulation of stocks that could not be sold. These factors point to an environment in which demand constraints play an important role. 4. Infrastructural and Raw Material Constraints 1.21 The challenges faced by the pulp and paper industry in the area of energy and raw materials are formidahle. As in other sectors of the Indian economy, firms must cope with uncertain supplies of coal and power. As paper production is a continuous stream process, unanticipated shutdowns are costly. The large units have responded to energy shutdowns by install- ing captive power generating plants for part of their requirements, but the small paper mills continue to entirely depend on outside power, a situation that contributes to their poor performance. 'f.he supply of coal is both inadequate and of inferior quality and leads to inefficient utilization of steam~raising equipment and higher costs of production. Sanjay Lall, in a study of the paper industry published in 1984, estimated that the shortages of coal and power were responsible for 10-20 percent of production losses. The shortage of rail wagons for coal transport has further aggravated the problem, since road transport is more expensive. 1.22 India is poorly endowed with traditional raw materials for pulp and paper processing. The per capita forest area is 0.11 hectare, versus a world average of 1.03. Currently, bamboo pulp constitutes 60 percent of the total wood-based raw materials used in production, with mixed hardwoods Development Consultants Private Limited (DCPL), Development of the Paper Industry in India, 1985, prepared for the World Bank. Excise duties have increased more rapidly in this sector than elsewhere, with pulp and paper firms accounting for 2.7 percent of the total excise collection by 1931, compared to 1.2 percent in 1975-76. - 8 - composing the remaining 40 percent. Bamboo, historically the primary raw material for paper in India, is, however, gradually being replaced by hard- woods because its supply is falling, competing demands are rising and large-scale plantations are quite costly.IO/ Thus, pulp mills geared to bamboo are now being forced to use hardwood, another factor contributing to poor performance. 1.23 In recognition of raw material constraints, the GO! has insti- tuted a number of measures to encourage the use of unconventional sources for pulp. The GOI has not, however, curtailed its liberal licensing polices, which was the government's response to coal shortages faced by the cement industry. 11/ In addition to the exemptions from excise duties and production controls mentioned earlier, units manufacturifig paper from unconventional raw materials have been delicensed, and both pulp and waste paper imports have been placed on the Open General License (OGL) list. 1.24 In two key areas, however, a change in policy is needed. Nearly all hardwood and bamboo forests are owned or controlled by the state governments, which issue short-term leases to mills for royalty fees vary- ing from 15-30 percent of the cost of the bamboo or wood. In some cases, the state governments have, however, terminated the long-term agreements, a move that has compelled paper mills to resort to open market purchases at higher prices. Private firms are not allowed to purchase forest lands directly, according to a law that forbids private ownership of over 12 hectares. These policies have not provided firms with the proper incen- tives to replant or preserve the environment, and forest lands in general are not well-managed. 1.25 The second area requiring reform is the promotion of bagasse- based paper mills. Bagasse, the residue from sugar production, is consi- dered one of the most promising sources of raw materials for paper produc- tion, and some mills already manufacture paper using it. However, bagasse is currently also used by the sugar mills as fuel, and any further develop- ment by the paper industry in this area is contingent on convincing the sugar mills that alternative fuel sources (such as coal) will be made available. One successful example of this approach is the Tamil Nadu News- print Project, financed by the World Bank. For this project, the paper manufacturer is covering the costs to convert sugarmill boilers from bagasse to coal-based while also providing the necessary coal from a centralized stock at the papermill. Yet until the GOI takes specific action in this area, it iR unlikely that the R&O efforts by the paper companies in bagasse-based paper production will evolve into a full-fledged effort. 1!!._/ Andrew Ewing, Energy Efficiency in the Pulp and Paper Industry with Emphasis on Developing Ceuntries, World Bank Technical Paper Number 34, World Bank, Washington, D.C., 1985 • ..!ll See Subsector Review--Cement, this volume. - 9 - Table 3: GROWTH IN CONSUMPTION OF PAPER AND PAPER PRODUCTS, 1960-85 • Growth of Consumption as % Year Consumption Consumption of Production 1960 371 107.54 1961 389 4.85 106.87 1962 410 S.40 105.94 1963 488 19.02 105.17 1964 512 4.92 104.49 1965 564 10.16 168.36 1966 599 6.21 158.89 1967 617 3.01 101.31 1968 637 3.24 98.76 1969 699 9.73 99.01 1970 755 8.01 99.60 1971 790 4.64 101.15 1972 820 3.80 102.12 1973 801 -2.32 100.so 1974 847 5.74 101.19 1975 843 -0.47 101.69 1976 893 5.93 101.48 1977 957 7.17 102.24 1978 1,025 7.11 101.99 1979 1,094 6.73 104.49 1980 1,200 9.69 107.91 1981 1,292 7.67 104.62 1982 1,259 -2.55 101.86 1983 1,181 -6.20 101.11 1984 1,361 15.24 100.81 1985 1,400 2.87 93.33 1960-70 7.36 1970-80 4.74 1980-85 3.13 Source: DCPL (1985). 5. Trade Poli£!. 1.26 Tariffs were imposed on most varieties of paper in 1925, and in this protected environment the industry experienced rapid growth until the beginning of independence. By 1947, the industry had so mastered the bamboo-based technology that Indian paper was half the CIF prices and pro- tection was temporarily discontinued. Protection was resumed in the sixties and early seventies, and almost all paper imports (with the excep- tion of newsprint and certain specialty items issued uncler Actual User (AU) - 10 - licenses) were banned. As production growth and capacity utilization fell, however, in 1981 the policy was changed to allow limited imports of print- ing and writing paper. These are brought in (canalized) through the State . Trading Corporation (STC) and distributed· to actual users. Since 1974, newsprint has also been canalized through the STC. In recognition of the scarcities of raw materials, waste paper and pulp were placed on the OGL list in 1981, a move the paper industry has applauded. 1.27 Although India did export some paper products in the early seventies, they were banned between 1973 and 1983 because of domestic shortages. In 1983, the ban on exports of printing and writing paper was removed, and exports of these products are now permitted up to 10,000 on a first-come first-served basis. 1.28 In conjunction with its promotion of indigenous paper production for the domestic market, the GOI also protects the manufacturers of components and equipment for pulp and paper mills. Imports of components are subject to duties of 40 percent, while complete equipment is subject to a duty of 63 percent. However, from 1976 to 1981, the import of second- hand machinery was permitted at concessional duties for small mills produc- ing less than 24,000 tpy. This policy was initially introduced as a short- term measure to raise paper production by lowering capital costs. However, the import of second-hand machinery proved to be detrimental to paper producers (para. 1.30). C. IMPACT OF THE REGULATORY :>OLICIES 1. Market Structure 1.29 The combination of the incentives provided by the exemption from production controls and excise taxes led to the rapid growth of small mills. Between 1981 and 1985, there was a 100 percent increase in the number of mills with capacities of less than 10,000 tpy, while their contribution to industry capacity doubled ·from 20 percent to 40 percent. (When medium-sized mills of less than 20,000 tpy capacity are included, the figure rises to 50 percent.) In contrast, the number of large units remained virtually the same, increasing from 22 to 23 during this period. 1.30 Consequently~ Indian pulp and paper mills are below the optimum size that permits full exploitation of scale economies. The capacity of the average mill in 1985 ~as 10,600 tpy in 1985, whereas mill sizes elsewhere range from an ave~age of 25,000 tpy in the Latin American countries to an average of over 100,000 tpy in the United States and Canada (Table 4). . -· 11 - Table 4: COMPARISON OF PLANT SIZES AND CAPACITY UTILIZATION, 1984 Capacity Number Capacity Average Utilization Country of Mills ('000 t.) Size (%) India 222 2, 195 9,887 64 China 1,600 7 ,810 4,881 97 Taiwan 149 2,300 15,436 84 Asia 2,982 41, 926 14,060 83 Latin American Countries 442 11, 136 25,195 78 North America 688 81,550 118,532 94 World 5,910 210,607 35,636 90 Source: Various trade journals. 1.31 The GOI's promotion of small-scale mills was based on their lower capital investment and gestation period, their abili~y to use unconvention- al raw materials that are renewable and require no investment to ensure availability, and their lower infrastructural requirements. In fact, these expected advantages have been offset by a number of factors. First, these mills have made no provision for captive power generation and so are parti- cularly vulnerGble to the power failures that afflict the economy. Second, the small units have no chemical recovery facilities, which are normally not incorporated in plants of less than 20,000 tpy annual capacity, another factor that raises their energy costs substantially while causing signif- icant pollution. Third, although small units were expected to meet their raw material needs from locally available (mostly unconventional) resour- ces, their availability on a sustained basis has been extremely difficult. Fourth, small mills have no facilities for effluent treatment, and recent pollution control measures requiring the installation of expensive control devices put the mills in a difficult position. Fifth, small mills of ten operate obsolete plant and machinery. The second-hand machinery that was imported at lower capital costs has been expensive to adapt to the Indian environment, and its tmreliability has led to rates of capacity utilization at 42 percent among small units. Finally, these small mills suffer from a lack of trained manpower and from inexperienced management. 1.32 Twenty-two of the appr.oximately 220 small mills have closed and have approached financial institutions for rehabilitation funds, while a number of others have stopped production. Some industry experts believe that the rest can only continue to operate because of the concessions granted by the GOI. Between January 1985 and early 1986, the pace of capacity augmentation by small mills slowed considerably because of their disappointing performance. Low capacity utilization, poor resource mobili- zation and inadequate financial returns have made Indian financial institu- tions reluctant to provide additional financing. - 12 - 1.33 The predicament of the small mills raises two serious issues: (i) the desirability of promoting units of this size through tax and production control concessions, and (ii) the necessity for a strategy for the rehabilitation or exit of financially strapped units. 2. Economies of Vertical and Horizontal Integration 1.34 There appear to be no particular benefits associated with expand- ing across product lines in the pulp and paper industry. Among the largest international pulp and paper producers, the percent of sales from pulp and paper can be as low as 22 percent or as high as 98 percent. The biggest firm in the Indian paper industry, Orient Paper and Industries Limited, only recently diversified away from paper production into cement. However, the second largest paper producer, Ballapur Industries, is much more diversified, with five chemical plants, a salt mill and a can manufacturing unit. Ballarpur Industries has exhibited much superior performance to Orient in terms of both high capacity utilization and profits. Given the financial difficulties of paper mills, Ballarpur has probably benefited from the diversification into 100re profitable endeavors. 1.35 A number of firms are also diversifying into other products as prospects in the paper industry remain poor. One interesting development has been the investment by some large firms such as Orient and Ballarpur in pulp and paper projects abroad. Orient, for example, has competed success- fully on the international market to obtain a contract to set up a paper plant in Nigeria and has participated in a joint venture to help Kenya establish the Pan African Paper Mills. In fact, Orient has displayed greater dynamism abroad than in India, where it lags behind Ballarpur in production, despite its greater installed capacity. Ballarpur has also been successful in exporting its technical capabilities through direct investment over2eas. 1.36 The scope for either upstream integration into large-scale plantations or d~~stream into specialty papers has been curtailed by the state leasing of forest lands llllder uncertain conditions and the GOI's small-scale reservation policy. Investment in large-scale plantations by pulp mills would be highly desirable, but without proper incentives, they will not do so. According to Ballarpur, the small-scale reservation policies that limit downstream production of paper products provide a dis- incentive to further investment. In addition, the items produced by the small-sc~le industry are generally of poor quality that in turn affects other sectors of the economy, ranging from agriculture (poor packing materials lead to a loss of produce) to consumer goods. 1.37 Another area where greater dynamism is necessary is marketing. In the past, supply shortages in the paper industry made it unnecessary for producers to engage in marketing. However, as capacity has increased in the face of slack demand, the necessity for more aggressive marketing techniques has become apparent. - 13 - 3. Firm Behavior 1.38 In conjunction with the elimination of shortages, the domestic environment has become more competitive. In contrast to many sectors of the Indian economy, entry into the pulp and paper industry is relatively easy. Licensing policies are quite liberal, and existing firms have no real advantages over new entrants, who can lower operating costs through use of modern equipment, and receive at least partial exemptions from excise taxes and price controls. Nevertheless, low expected profits have aeterred the entry of large firms. 1.39 The now fairly competitive domestic environment is indicated by the falling market shares of incumbent firms; the four-firm concentration ratio fell from 42.5 percent in 1976 to 31.1 percent in 1981, and is still falling in 1985. Table 5 shows the change in market shares for a selected sample of 13 firms over the 1979-1983 period. Most firms, both large and small, show a decline in their share of the market. This decline is due to the significant entry of small units into the industry during this period, as discussed earlier. 1.40 The combination of soft demand, rising input costs, high excise du~ies, price controls and an increasing number of firms has created an environment in which firms are apparently struggling to keep profit margins from falling below zero. One solution bas been to reduce production. The growth rate of paper production declined from over 11 percent annually in the 1950s to as lo~ as 1.4 percent between 1981 and 1983 (mirroring the fall in consumption), while capacity utilization fell from 84 percent in 1970 to 57 percent by 19a4. Given the desirability of raising low per capita consumption of paper, which has repercussions on educational and social objectives, however, it is imperative that the industry be motivated to increase, not decrease, production.· Freeing the prices of previously controlled paper products, combined with reduced excise taxes, could produce lower retail prices, and be used in conjunction with more aggres- sive marketing by firms.to stimulate demand and restore the financial health of the industry. - 14 - Table 5: MARKET SHARE OVER TIME BY FIRM (Ratio yearwise) Name of Firm 1979-80 1982-83 Ballarpur a/ 12.02 10.90 Orient 11.85 8.55 Bengal 3.40 2.70 Sirpur 4.80 4.33 Star 2.92 2.86 Titaghur 6.50 4.25 West Coast 3.30 4.92 Andhra Pradesh 5.70 4.37 Seshasayee 3.90 3.55 Punalur 1.00 0.94 Rohit 1.58 1.35 Bhadrachalam 3.75 Rayalaseema 0.57 2.82 a/ Values for 1981 and 1984. Sources: Data sheets of Annual Reports related to above Paper Mills. For Ballarpur, figures taken from CMIE, Market and Market Shares, 1986. 1.41 In contrast to the fairly competitive domestic market, the pulp and paper industry has been effectively insulated from the international markets for the last 25 years. At the time of independence, prices for Indian paper prices were half the CIF prices; and exporters were in a position to compete in the world markets. As Table 6 indicates, the current price for Indian paper products is now higher than international prices for a variety of pulp and paper products. These data partially reflect the impact of excise taxes on producer prices. Since the demand for paper products appears elastic, however, presumably much of the incidence of the tax (tax burden) is absorbed by producers. High prices could be taken as eithe4 an indication of high costs of production or high profit margins. Since profits in the industry are generally low and falling market shares indicate that the indust.ry is f~irly competitive, it is unlikely that prices are set significantly above marginal costs. Rather, the combination of high raw material costs and technological inefficiency appear to contribute to high costs, and consequently to high prices. I~ particular, operating costs of small mills operating at below minimum efficient scales contribute to higher prices. Although the entry of more units could in theory promote price competition, these mills are too inefficient to challenge existing firms. - 15 - Table 6: lN!'EmATIOW. PR!<£S ~S 1ll£S1IC PRIk.J..: CEMENT CAPACITY AND PRODUCTION, 1960-85 Capacity Production Fiscal Tons Growth rons Growth Capacity Year ( '000). Rate ( 1 000) Rate Utilization 1960-61 9.30 7 .83 84.19 1961-62 9.47 1.83 8.25 5.36 87.12 1962-63 10.00 5.60 8.59 4.12 85.90 1963-64 10.50 s.oo 9.36 8.96 89.14 1964-65 11.24 7.05 9.69 3.53 86.21 1965-66 12.00 6.76 10.58 9.18 88.17 1966-67 12.56 4.67 11.06 4.54 88.06 1967-68 13. 78 9.71 11.30 2.17 82.00 1968-69 14.98 8.71 12.24 8.32 81.71 1969-70 15.98 6.68 13.80 12.75 86.36 1970-71 17.61 10.20 14.35 3.99 81.49 1971-72 19.56 11.01 15.07 5.02 77.04 1972-73 19.76 1.02 15.55 3.19 78.69 1973-74 19.76 o.oo 14.66 -5.72 74.19 1974-75 20.06 1.52 14.80 0.95 73.78 1975-76 21.16 5.48 17 .29 16.82 81.71 1976-77 21.46 1.42 18.84 8.96 87.79 1977-78 21.91 2.10 19.38 2.87 88.45 1978-79 22.55 2.92 19.42 0.21 86.12 1979-80 24.29 7.72 17.69 -8.91 72.83 1980-81 27.92 14.94 18.66 5.48 66.83 1981-82 29.26 4.80 21.01 12.59 71.80 1982-83 34.39 17.53 23.32 10.99 67.81 1983-84 37.04 7.30 27 .07 15.78 73.17 1984-85 41.24 15.72 30.20 11.85 (70.73) Source: Cement Production and Dispatches, various years, and World Bank (1985 India Cement Project). 2.05 India is an economic producer of cement by 1ntt~rnatio11al standards and could be cc.npetitive on the world market.. De~pite its comparative advantage, however, India imports small quantities of cement to help alleviate the continuing shortages. The planned expansion in produc- tion capacity is expected to make India self-sufficient in cement manufac- turing within the next sever.al years. - 21 - B. THE REGULATORY ENVIRONMENT: 1942-82 1. Licensing Policies .2.06 Licensing for cement manufac•uring has been mandatory since 1951, . except in 1966-70. In the latter years, the GOI barred large companies from building additional capacity on the grounds that new entrepreneurs should be encouraged to invest in this highly concentrated i11dustry. It withdrew the ban two years later because there were no further applications for new plants. To address that situation, in February 1973 the GOI added cement to the list of Appendix I industries, a move that again permitted and encouraged large firms to invest. To provide a balance to the large firms and allow exploitation of the smaller limestone deposits located in areas with potential locational economies, in 1979 the GOI introduced incentives to promote mini-plants (installed capacity not exceed 200 tons a day or 66,000 tons a year). In 1980, licensing was made even less cumber- some; even MRTP firms were liberally defined--as those with over a 30 percent market share. (By 1983, even the largest private sector firm, Associated Cement Companies Limited, or ACC, accounted for only 26 percent of the market.) 2. Price Controls 2.07 The cement industry has been subject to price and distribution controls since 1942, when the government introduced cost-plus pricing for individual plants under the Defense of India rules •. Between 1946 and 1952, uniform prices were set on the basis of the production costs incurred by ACC, the largest cement company. From 1952 to 1966, a weighted average retention price was imposed, with a special additional price for high-cost units. As noted, the government decontrolled prices from 1966 to 1967, although they were still set informally by the industry through the Cement Allocation and Coordination Organization with the guidance of the govern- ment. In 1967, a tmiform retention price was re-established at the request of the industry. 2.08 A l&ck of new investment in the 1970s led the GOI to revise the retention prices for new plants; the formula devised then allowed an expected rate of return .on capital of 12 percent post-tax. At the same time, a committee was established to review the p1ices for all other plants. The committee's price recommendations were adopted in 1979 and retained until 1982. During this period, the GOI adopted a three-tier retention price structure with provisions for a~nual increases. Units were classified as low-, medium- and high-cost and were assigned retention prices of Rs 185, 205, and 220 a ton, respectively. New plants. continued to receive preferential price of Rs 296 a ton. 2.09 Retention prices failed to keep up with general inflation, and revisions were made only infrequently. As a result, the pric~ of cement - 22 - (excluding that sold to new units) relative to non-primary commodities declined 18 percent between 1966 and 1980 (World Bank, 1980). 3. Distribution Controls 2.10 As in other sectors of the Indian economy (fertilizer, paper and steel), price controls have been accompanied by cont~ol over distribution. From 1968 to 1982, the Cement Controller in the Ministry of Industry regulated all cement distribution, which was channeled through the State Trading Corporation according to a priority system--first the central and state governments, second the key industries and third the general public, under the so-called "free sale" category. Industry-level priorities were determined by a variety of agencies using criteria that were not always well-defined. During the 1968-79 period, dispatches under the free sale category fell from 59 percent to 38 percent, a reflection of the rising proportion of cement going to the government (32 percent to 46 percent) and to priority industrial users. 2.11 In the face of chronic shortages, black markets developed in the late seventies where prices were at least 25 percent higher than.the official ones. The distribution agencies were accused of demanding graft that offset the differential between the controlled retention prices and the market prices to the consumer. In addition, the priority sectors were also able to leak cement to the black market. According to some observers, private construction flourished despite a lack of cement allocations, whenever a public sector project was under construction nearby. According to various estimates, between 2 percent and 10 percent of cement supplies were sold through the black market. 2.12 In addition to control over di~tribution, the GOI has also regulated the prices for freight since 1956 through the freight equaliza- tion system. The basic objective is to have a uniform destination price throughout the country. As such, cement transport costs from a given plant to a certain location are subsidized or taxed so that the actual freight charges per ton exceed or fall short of the average freight per ton of cement transported from all plants. Consequently, plants are located near deposits which are easily accessible in terms of costs, but are often far from product markets. 4. Infrastructural Constraints (Energy, Coal, Transport) 2.13 Historically, energy and transport shortages have contributed to a low utilization of capacity in the cement industry (Table 2). Power interruptions and failures as well as fluctuating voltages have increased energy consumption and caused unsteady operations and mechanical break- downs. The power problems were particularly chronic in the seventies, accounting for between 15-30 percent of the losses in cement production. In response, many large plants installed captive pow.er generating facili- ties. While these measures reduced their dependency on uncertain energy supplies, they also contributed to rising operating costs, since this source of energy is more expensive. The insufficiency of coal supplies accounted for as much a~ 30 percent in lost production in 1974 but only 10-15 percent b~tween 1976 and 1979. - 23 - 2.14 In addition, the ash content of the coal delivered to plants has varied substantially. Coal with a high ash content upset kiln condi- tions, harmed the quality of clinker and led to the production of poorer quality cement. An inadequate supply of wagons for both the transport of coal to the plants and of cement from the plants, has also affected performance. This problem has on the one hand resulted in temporary shortages in certain markets, and on the other to the accumulation of clinker and cement stocks beyond storage capacity and hence to the curtail- ment of production. Another constraint has been India's railway system, which consists of lines with different gauges. Where the transport of cement has involved one or more transshipments, delays have increased the transport costs and production losses. 5. Impact on Market Structure and Performance 2.15 Market Structure. In 1978, which can be taken as a sample year for assessing the pre-1982 period, the average plant size for cement manufacturing in India was nearly 300,000 tpy. This figure was much lower than the size of modern cement plants in industrialized countries, which now have capacities of 900,000 tons per year (tpy). For the 75 plants in operation in India, only 4 had capacities below 600 tpd or 200,000 tpy, an indication of a fairly narrow size distribution. 2.16 In terms of product types, the 1970s witnessed a shift in produc- tion from ordinary Portland cement (OPC), accounting for 90 percent of production in 1970, to Portland Pozzolana cement (PPC) and Portland blast furnace slag (PSC), accounting for 47 percent of production in 1978 (Table 3).2/ This shift may be traced to the pricing policy, which sets prices which do not adequately reflect differences in production costs for the three types of cement. Since the costs of PPC and PSC were lower, the share of the more expensive cement OPC, declined. In addition, the intro- duction of revised norms (which just establish minimum strength required) permitting the production of lower quality cement contributed to a decline in the quality of all types of cement. 2.17 Entry and expansion was vigorous during the 1960s (Table 1). Licensing deregulation during the 1966-70 period led to a surge in capacity growth between 1966 and 1972. The annual increase in capacity was as high as 11 percent, a figure which was not reached' again until pricing policies were liberalized in the early 1980s. Since the gestation period for a greenfield cement plant is three to four years, the impact of deregulation in the cement industry continued to be reflected in capacity increases after 1970. Between 1972 and 1979, capacity growth declined to an average rate of only 1.8 percent a year. This drop was the result of both licensing restrictions and price controls. Since the cement industry is highly capital-intensive and requires large investments, the ban on invest- ment by large houses during 1970-73 also essentially brought investment to a halt. Between 1972 and 1975, capacity grew at particularly negligible '!:_/ These three types of cement account for 99 percent of all cement produced in India. - 24 - rates (Table 1), given that no investments were commissionel between 1970 and 1973 and that the gestation period for a greenfield cement plant is three to four years. Capacity continued to g=ow on an average at only about 2 percent annually during 1976 through 1979, as price controls led to unrenumerative profits that discouraged growth. When the three-tiered price retention system was established in 1979, growth ~n capacity accelerated. 2.18 Industry Performance. In conjunction with lagging growth in capacity, growth in production also fell during the 1970s. The former went from an annual average of 6.5 perceni between 1960 and 1970 to 4.3 percent between 1970 and 1982, while the latter declined even faster, from 6.6 percent to 3.3 percent. The falling rate of growth in production parallels the decline in capacity utilization, which averap,ed 86 percent during the sixties and 78 percent between 1970 and 1982. The fall-off occurred despite chronic shortages and excess demand, estimated by the World Bank to have varied between 10-25 percent of annual consumption. In 1979, the shortages ~ere estimated to have been at least 15 percent of available cement. With prices set below the market-clearing level, t.mofficial markets, as noted, emerged whe~e cement was sold from 25-100 percent above the official prices. 2.19 In response to the scarce domestic supplies, exports of cement were banned in 1977 (except for those to Nepal, which had a special agree- ment with the Indian government), and beginning in 1978, imports were canalized through the STC and distributed by the Cement Controller's Office. 2.20 According to a 1980 World Bank report, the unit costs of indivi- dual plants in the cement industry were highly variable, a reflection of the wide diversity in technologies, vintages and production costs. Half the plants were over 20 years old and in various states of disrepair. Only 35 percent of kiln capacity was based on the dry or semi-dry processes introduced in the 1950s, which are significantly more efficient than the wet process technology employed by other Indian plants. The t.mit operating costs of plants operating with wet processes were estimated to be 6 percent higher than plants using dry process technology. The smaller plants were up to 20 percent less efficient than larger plants. More recent studies have estimated that plants operating under dry process technology can save up to 40 percent on energy costs (World Bank, 1985). 2.21 This diversity in performance was at least partially attributable to several aspects of the policy environment. The three-tier pricing system accommodated high-cost plants; price and production controls discouraged investment in modernizing equipment (with repercussions on the cement equipment industry); and the promotion of mini-plants beginning in 1979 encouraged production at inefficient scales. The low capacity utili- zation at units suffering from infrastructural constraints or machinery breakdowns also affected the costs of production: the Bank report estimated that for every 10 percent fall in capacity utilization, there was a 4 percent increase in unit operating costs. In addition, on average, the input/output coefficients for labor were much higher than internationally: the average plant needed 6.7 man-hours to produce a ton of cement compared - 25 - to 0.6 in developed countries. The report attributed the over-employment to traditional socio-political conditions in India and not eo the teclmolo- gies used. 2.22 Despite these problems, the estimated long-run marginal costs were extremely competitive compared to international prices: Rs 265 versus Rs 479 per ton of bagged cement in 1978-79. India's comparative advantage in cement production could be traced to lower labor as well as energy (coal) costs. These cost advantages, however, were not reflected in the market price of bagged cement, which was nearly as high as the landed cost of cement imports (Rs 437) because of the imposition of various indirect taxes and freight charges. C. RECENT POLICY MOVES: 1982-85 2.23 In February 1982, the government discarded the three-tier system of price controls and introduced a system that required firms to sell 66.6 percent of their installed capacity to the government at controlled prices and allowed them to sell the rest at prices determined by the market. To promote entry, units that entered production after January 1982 would be granted levy quotas equal to 50 percent of installed capacity three years and even less during the first three years of operation. Special arrange- ments were also made for "sick" firms, which would be given new-firm status for a period of three years. 2.24 In June 1985, the levy quota was set at 60 percent for old llllits and 40 percent for new or sick ones, and the basis for the quota was changed from licensed capacity to actual production. The distribution of cement on the free market was liberalized and is now the responsibility of individual producers. Mini-cement plants continue to be exempt from all pricing and distribution controls and, as before, they are also exempt from the licensing requirement. However, they must register with the Directorate General Technical Development (DGTD). 2.25 Although the GOI has assured the World Bank that it intends to liberalize the cement industry completely by 1990, there may be some resistance to these policies, particularly from the mini-plants, "sick" units and firms that face locational disecono1~ies. The mini-plants benefit from the right to sell exclusively on the free :tJarket, where prices are still higher than levy prices (although they have fallen considerably since the 1982 policy change). Complete liberalization will probably lead to the closure of some of these plants, which have higher Ui.'it costs of production than do the larger plants. The "sick" units that have new firm status and that are permitted to sell a greater portion of their output on the free market also benefit from higher prices. Finally, firms that are at great distances from consumers and have benefited. from subsic.ized transport costs will be at a disadvantage when this support is eliminated. 2.26 The response to the 1982 measures, as noted, has been impres- sive. Capacity expanded from 29.3 to 41.2 million tons, while production rose from 21.0 to 30.2 million tons between 1981-82 and 1984-85. The .· - 26 - equivalent average annual growth rates are 12 percent for capacity and 13 percent for production. The government has granted letters of intent and/or licenses for an additional 50 million tons, which would more than double current installed capacity. However, the.Ministry of Industry is not consinering any applications for greenfield plants now for a number of reasons, including uncertainty regarding market size in certain areas, the lack of coal mining capacity and bottlenecks in rail transportation. Obtaining a license is thus also dependent on gaining clearance from the railway authorities. (The applications are not actually bet-g rejected, but capital financing has been cut off f lr all new plants.) By the end of the Seventh Plan period, the Ministry estimates that 62 million tons of installed capacity will be in operation; at 85 percent capacity utiliza- tion, the coal requirement is 13.2 million tons a year. For the Eighth Plan period, 33.3 million tons of additional capacity are already under consideration, requiring an additional 7 million tons of coal. Since a supply of only 3.2 million additional tons of coal is foreseen, the supply gap w-111 be 3.8 million tons. 2.27 According to the Cement Manufacturers Association (CMA), there is scope for investment if capital were not restricted. The CMA also noted that only one-fourth to one-half of all letters of intent actually make it to the licensing stage. Since 30 million of the additional 50 million tons of capacity to be added during the Seventh and Eighth Plan periods are at the letter of intent stage, it appears that there may in fact be some scope for granting additional capacity. The Office of the Development Commis- sioner for Cement has suggested that it might be desirable to permit those states that have the infrastructure and coal resources to determine whether to license new cement capacity. Since the central government finances 80 percent of all projects, however, a system for coordinating capacity licensing and financing would be necessary. 1. Impact on Structure and Behavior 2.28 The construction of many larger units led to an increase in the average size of cement plants in India from 300,000 tpy in 1978 to 535,000 tpy in 1985 (excluding mini-plants). The capacity of new plants now being built is 1 million tpy, which is about the average size for modern cement plants in industrialized countries. Indeed, average plant sizes compare favorably with those in industrialized countries such as Argentina) which has average plant sizes of 600,000 tpy in cement. However, the size distribution of cement plants has become more dispersed over the last eight years, primarily because of the increase in very small and very large plants. Government support has encouraged the continued entry by small units. While they are expected to account for only 5 percent of total capacity, they will make up 45 percent of the total number of plants by 1990.3/ When the mini-plants are taken into account, average plant size drops-to 4so,oon tpy. Although plants with million ton capacities have risen from one in 1978 to eight in 1985, only one-third of cement plants have capacities greater than a half million tpy. ll World Bank, 1985. - 27 - 2.29 The trend away from the production of ordinary Portland cement toward production of less expensive (and lower quality) varieties has continued (Table 2), and OPC now accounts for only 25 percent of produc- tion. Despite changes introduced in 1982, the fixed price for OPC is only Rs 15 per ton higher than that for PPC and PSC. This difference is too small to cover the variation in production costs, and consequently manufac- turers have increasingly concentrated on producing the cheaper varieties. Some of the lower quality types of cement, such as PPL, can be used for applications where high strength is not required. However, only high quality cement is sold on the free market, which indicates that consumers prefer the higher priced, higher quality product. 2.30 Opportunities for forward integration in cement were basically non-existent before 1982, when all distribution and allocation of cement were controlled by the GO! and state governments. Since the distribution for free sale cement is now the responsibility of firms, however, some have set up their own distribution facilities and sales staff. The lack of facilities to accommodate the unprecedented supply of free sale cement has in fact created distribution and marketing problems. 2.31 These problems will not be easily resolved. During the years of cement shortages under the cement controller, output was sold before shipping, and there was no need to establish distribution facilities or maintain cement stocks away from plants. Now a much higher volume is being shipped to the markets without prior commitment by an end-user and the lack of inventory capacity is creating marketing problems. In addition, distribution is cumbersome b~cause all shipping is done in bags instead of in bulk, the standard method of shi~ping in both industrialized and industrializing countries such as the Republic of Korea. The bags are made of jute, which leaks, ·and all handling is manual, a time-consuming and costly process. A13 liberalization proceeds and the volume of cement for distribution increases, these problems will become even more severe unless action is taken. 2.32 The cement industry has developed almost entirely in the private sector. In 1936, 10 major producers joined to become the ACC, which conse- quently controlled 90 percent of all production. The lack of monopolistic behavior on the part of ACC is, however, clear from the steady decline in its share of production--to 56 percent in 1951, 32 percent in 1978 and 25 percent in 1985. Nevertheless, in 1978 the four largest producers ac- counted for at least 60 percent of the capacity in the industry, and by 1985 the six largest producers accounted for 60 percent. These six firms (which include the public sector Cement Corporation of India, or CCI) will continue to dominate the market,. accounting for 55 percent of total produc- tion in 1990. 2.33 Although market shares are often used as a proxy to measure market power or non-competitive behavior, recent theory (particularly in the area of contestable markets) suggests that market shares may be an inaccurate indicator of market power and that profit margins provide a more appropriate measure. Given the fact t~at non-levy prices have been falling while costs have. been rising since the decontrol was inst.ituted in 1982, it appears that these large firms have not been exercising such market power. - 28 - Table 2: LOSS IN CEMENT PRODUCTION BY CAUSE, SELECTED YEARS (%) Cause 1974 1978 1982 1983 Power Cuts/Failures 18.07 23.55 48.00 40.00 Mechanical Trouble 7.50 43.76 29.00 26.00 Short age of Coal 33.29 2. 78 o.oo 3.00 Labor Troubles 3.36 8.03 1.00 4.00 Other 37.79 21.88 22.00 27.00 Total 100.00 100.00 100.00 100.00 Table 3: RELATIVE SHARES OF TYPES OF CEMENT 1970 1977 1978 1983 OPC 89.80 71.30 52.30 25.02 PPC 0.60 10.20 31.10 56.27 PSC 9.20 17.20 15.80 17 ,11 White Cement 0.30 0.40 0.40 0.32 Other Types 0.10 0.90 0.40 1.28 Total 100.00 100.00 100.00 100.00 Source: Cement Production and Dispatches (1983), and World Bank, India Cement Sector Report (1980). In addition, the GO! has sought to promote a competitive market environment through a more liberalized import policy and, as noted, the promotion of mini-plants. 2.34 The GOI instituted more liberal import policy in 1982 to help stabilize prices and alleviate the shortages. ·As the process of decontrol continues, the GOI is expected to put increasing emphasis on imports as a means of stabilizing prices. However, this policy can only be effective if the STC can improve the current distribution system, which is neither capable of handling large quantities of imported cement nor providing transportation to inland areas. - 29 - 2.35 The 1979 policy to·encourage mini plants was partially introduced to provide competition to large producers. However, the operating costs of these small units are much higher and they probably could not compete with the larger plants in the absence of price controls (see para 11). In addition, large units have been aided by the policy of freight equaliza- tion, which permits them to bypass the normal locational economies of an industry such as cement, whose transport costs are high. This advantage will en.i with complete price decontrol, since there is no freight equaliza- tion on non-levy cement. 2. Impact on Performance 2.36 The retention prices (FOB plant) for free sale cement have declined from Rs 800-850 a ton when prices were decontrolled in 1982 to Rs 600-650 a ton today. The current price compares favorably with the price of imported cement, which sells for Rs 878 a ton net of taxes and duties. At the same time, the government has all~~ed the price of levy cement to rise from Rs 335 a ton in 1982 to Rs 379 a ton. 2.37 Although the gap between levy and non-levy cement has narrowed considerably, the CMA claims that the levy price is about Rs 85 a ton belcw what it should be according to the pricing formula, which provides for a 12 percent return on net worth. The GOI, however, prefers to limit price increases to minimize inflation. As a result, the GO! has responded by expanding the percentage of cement which can be sold on the free market. Regardless of what is the appropriate policy, it is clear that moderniza- tion of old plants requires large investments that cannot be achieved on the basis of the return on levy sales.4/. 2.38 In terms of prices, domestic cement appears quite competitive with imports, although the differences in quality can be great. Ninety- five percent of all applicants for import licenses to the STC (the only authorized import agent) cite low quality as their primary reason for importing cement. 2.39 India continues to be an economic producer of cement, despite increasing capital and fuel/power costs. In 1985, the cost of a million tpy capacity plant was estimated to be US$90-120 a ton, compared to US$170-200 internationally. The lower investment costs are the result of the high degree of good domestic capabilities in cement equipment manu- facture and low labor costs for the installation and engineerillg. However, the average operating costs (including depreciation) for an Indian dry process plant, currently US$30 per ton of unbagged cement, are higher than those at plants in Spain and Greece and may soon be higher than those in other European countries. In the past, India has benefited from lower costs for raw materials (coal) and wages, but this advantage is disappear-· ing because of declining produc.tivity in the older plants. Although the proportion of wet process plants has fallen, these plants still accounted !!_I The World Bank is currently funding a project to lll0dernize·s1x large cement units. - .:. 30 - for 42 percent of .capacity in 1985, as compared with 10 percent in Turkey (1978), 21 percent in Tunisia (1981) and less than 5 percent in developed countries such as the Feder.~l Republic of Germany. Since energy consump- tion i.s directly correlated with the type of process used, energy use in India per ton of cement is high compared to that in other countries. 2.40 The cement industry has exhibited impressive increases in produc- tion, although quality has generally remained poor. The sector also continues to be characterized by low levels of capacity utilization, which ranged between 69 percent and 74 percent during the 1980-85 period, desEite excess demand. This low capacity utilization is the result of many of the same factors that affected plants during the 1970s, including power and coal problems and the non-availability of wagons for transport. Mechanical troubles are also responsible for production shortfalls, particularly at old plants. In 1983, producers blamed power cuts for 40 percent of the loss in production, while mechanical troubles at old plants were respon- sible for 26 percent.Sf There is a significant variance in capacity utili- zation between the large cement units and the mini-plants: the latter operated at 48 percent, the former at 71 percent. In part, the discrepancy can be explained by the lack of captive power generating facilities at the mini-plants and properly trained operating personnel. It is also interest- ing to note that the only public sector company, CCI, has performed rela- tively well (in comparison with the private sector), with an average capa- city utilization of 75 percent in 1983. 5/ Office of the Cement Controller, 1983. - 31 - INDIA - INDUSTRIAL REGULATORY POLICY STUDY III. Subsector Review--Nitrogenous Fertilizers A. INTRODUCTION 3.01 The first modern nitrogenous fertilizer plant in India was established in 1933 by Tata Iron and Steel (TISCO), a private company that produced ammonium sulphate as a by-product of its steel manufactures. The chemical fertilizer industry, however, did not begin to exhibit rapid growth until a decade after India's independence. In 1957, the public sector Fertilizer Corporation of India (FCI) began to manufacture urea, which currently constitutes 85 percent of the nitrogenous f~rtilizer production in India. During the next several years, the r.overnment of India (GOI) invested heavily to expand the public sector, and production grew at over 20 percent annually between 1960 and 1970. Farmgate and . ex-factory prices were generally controlled by the GOI over this period, with the former directly linked to the latter until June 1974. In that year, the differential impact of oil price increases on producers us~ng different f eedstocks led to a revision of the pricing formula. The system of retention price was introduced in 1977, establishing specific prices for each plant in the industry and has been only slightly modified since 1977. 3.02 For three decades, the GOI has actively promoted the fertilizer industry, which is a key component of its agricultural policy. The limited scope for expanding agricultural land has increased the importance of modern inputs such as fertilizers as a means of expanding agricultural production •.!./ In 1982, the Worl~ Bank esti~dted that 30 percent of crop production could be attributed to the use of chemical fertilizers, and this share is expected to increase to over 50 percent by the year 2000. Both through direct !~vestment (public sector enterprises account for over half the installed capacity in the industry) and the implementation of licens- ing, distribution and pricing policies, the GOI has sought to ensure that fertilizer production and consumption continue to grow. 3.03 As in many other sectors of the economy, the government regulates aspects of the industry ranging from the choice of raw materials (feed- stocks), technology and location for new plants to prices and distribu- tion. The pricing policy encourages both production and consumption through a formula that rewards producers with attractive returns to the extent they conform to reasonably efficient plant-specific production norms and provides farmers with fertilizer at low prices· through budgetary subsi- dies. It does not however provide strong incentives for overall economic efficiency. lf India Fertilizer Distribution Project: Project Brief, Industry Department, World Bank, Washington, D.C., April 1986, p.4. - 32 - 3.04 Fertilizer production grew at an annual rate of 10 percent during the 1970s. Over the years 1981-82 to 1985-86, however, annual growth has maintained itself to a 9.6 percent annual average (Table 1) only because of the commission of four new plants in 1985-86. Although this rate compares favorably with the world aqerage (4 percent annually for 1981-82 to 1983-84), accelerated growth in installed capacity in conjunction with increases in productivity will continue to be necessary to meet the ambitious 10 percent target for the Seventh Plan Period (1985-90). Table 1: NITROGENOUS FERTILIZER CAPACITY AND PROuUCTION Capacity a/ Growth Production Growth Capacity Year ('000 tons) Rate ('000 tons) Rate Utilization 1960-61 162.20 112.00 69.05 1961-62 246.30 51.85 154.30 37. 77 62.65 1962-63 246.30 o.oo 194.20 25.86 78.85 1963-64 326.70 32.64 219.10 12.82' 67.06 1964-65 324.00 -0.83 243.20 11.00 75.06 1965-66 324.00 o.oo 237.90 -2.18 73.43 1966-67 524.90 62.01 309.00 29.89 58.87 1967-68 632.40 20.48 402.60 30.29 63.66 1968-69 904.50 43.03 563.00 39.84 62.24 1969-70 1,136.40 25.64 730.60 29. 77 64.29 1970-71 1,349.30 18.73 832.50 13.95 61. 70 1971-72 1,514.70 12.26 949.20 14.02 62.67 1972-73 1,471.20 -2.87 1,054.50 11.09 71.68 1973-74 1,933.50 31.42 1,049.90 -0.44 54.30 1974-75 2,162.10 11.82 1,186.60 13.02 54.88 1975-76 2,625.'10 21.41 1,508.00 27.09 57.45 1976-77 3,024.10 15.20 1,862.40 23.50 61.59 1977-78 3,068.80 1.48 1,999.80 7.38 65.17 1918-79 3,295.80 7.40 2,173.00 8.66 65.93 1979-80 3,901.80 18. 39 2,224.40 2.37 57.01 1980-81 4,357.80 11.69 2,163.90 -2.72 49.66 1981-82 4,635.80 6.38 3,143.30 45.26 67.80 1982-83 5,178.50 6.38 3,429.70 9.11 66.23 1983-84 5,201.30 0.44 3,491.50 1.80 67.13 1984-85 5,201.30 o.oo 3,917.30 12.20 75.31 1985-86 6,689.60 28.60 4,310.10 10.03 64.43 Source: Fertilizer Statistics, 1984-85, Fertilizer Association of India, New Delhi; Industry Department, the World Bank. a/ ·Capacity based on Fertilizer Association of India figure for September 31 of the fiscal year. - 33 - Table 21 tNl>lA • COOIBllATlVI PUTlLIZ!R IHDUSTRY PllOJBCT C.UACtTY MQ) PRODUCTIOll or PRillClPAI. lllTROGBN PBllTtLlZllll PWTS 1984/85 • March 1985 Production C/ Capacity C<111panyIP lant Product !J Peed1tock b/ Capacity 1981/82 1982/83 1983/84 1985/86 UUUaatton • - ( '000 tpy) ( '000 tpy) m A. Public Sector \, rertUber atld Chuicala Travancon Ltd. (PM:T) Alvaye • Udyogmaftoration Ltd. (lli'Ct.) Bareuni - 'lihar Urea N 152.0 74.2 78.5 60.8 61 Ourgapur - Weat Bengal Urea N 152.0 60.0 43,2 69.4 10 Naairup - Aaeae Urea, AS NG 197.0 104.9 100.3 81.4 38 4. Madraa FMUliaer Ltd. (MIL) !t Manali - Taeil Nadu lire a, NP, NPK N 176.0 154.0 135.0 113.8 129.l 73 ~acional Fertiliser Led. (NFL) Bhatinda - Punjab Urea FO 235.0 133.7 150.7 162.0 143.6 168.2 72 Nangal - P.injab CAN, Urea FO 232.0 170.4 180.8 188.0 190.4 142.l bO Pantpat - llaryana Urea ro 235.0 182.1 168.9 161.0 157.1 199.1 8b 6. Neyveli Lignite Cot~. (NLC) NeyveU - Taetl ~adu Urea 10.0 45.2 45,3 ss.8 59.1 84 7• Raahtriya Ch81licala and Fercllller Ltd. (RCFL) Trotlbay - Maharashtra Urea, \IP, NIP HG, A 317.0 167.3 209.6 277.8 269.0 ~. Raahtriya Cheeicala and Fertiliser TllAL - Maharashtra Urea NG 683.0 376.5 bl 9, Steel Authority of India Ltd. (SAIL) 11.ourkela - Oriasa CAN COG, N 120.0 53.4 10.0 22.Cl 34.5 l9 8 aaiall plants AS coo 36.0 16.1 17 .o 14.9 18.7 5l 10. 111.L Haldia Urea FO 290.0 1.1 56 Total Public Sector 2,052.3 55 8. Private Sector l. Cor0111&ndel Fertiliaer t.td. (CFLJ Vtaa11.- Alldhra Pradesh Urea, llr, NPK N 81.8 2. F.tO - Parry (tndia) tnnore - Taeil Nadu APS N 16.0 9.6 60 3, Gujarat Nat'1118da Valley Fert111zer Ltd. (GNVF) BharJch • Gujarat Urea 273.o 2.4 178.2 212.9 212.s 270.8 99 4. Gujarat State Fertlltaer co. (GSFC) Baroda •Gujarat Urea, AS, OAP NG, N 236.0 200.8 184.4 205.9 211.6 252 .7 107 s. Hart Fertiltzera Ltd. (NCOH) Varanasi • Uttar Pradesh A/Cl COG 2.1 1.2 3,9 39 b. Indian Explo•ives Ltd, (IEL) Kanpur • Uttat Pradesh Urea N 310.0 202.0 262.9 237 .s 2116.) l!b 7. '1angalore Chemicals and Fertilizers Ltd. (MCFL) lla1111alore - Karnataka Urea N 156.0 IUl.8 104,2 I J4,0 110.11 11 8. Punjab National Ferttliltt'ra Naya Nangal • Punjab Urea FO 16.0 2.t 4l 9. Shrtraa Fe rt llizer Corp. ( SFC) Kota - Ra1asthan Urea N 152.0 121.1 143.4 136.7 143.7 lbJ, I 1117 10. Southern Petrocheaical tnduatriee Corp. (SPIC) Tutieortn ·Tamil Nadu Urea, NP, NPK N 29).0 200.2 278.t llJS 11. Tata Iron and Steel Co. (TtSCO) Jeaishedpur - Rihar AS 4,0 1.J 2.'~ n.a. n.a. 12. Zuarl Agro Cheetcala Ltd. (ZACL) San Cole - Goa Urea N 198.0 t.14.1 171.11 1~0.9 1411.t 191.J 91- 11. Paradeep Phosphate OAP A -11:Q. ~ 1 Total Private Sector 1,823.h l ,b67. 3 ~l c. Cooper at tve Sector I. Indian far1"1!n fent\t7.er i:ooperattve (tFfCO) . Kandla/Kalol - Gujarat Urea, NP, NPK 200.0 255.l i7 7.t 253.0 321.8 ~7l.ll 1•15 Phulpur • Uttar Pradesh Urea 228.0 181.2 lb6.!I 174.2 100.1. 191.0 S4 2, KR! BRCIJ, llaziza t & I l llrea -ill.& _122..:2. 19 Total Cooperative ~ector I I 156.0 51 TOTAL !NOIA a/ AS • Ammoniu., Sulfate; "1' • ~itro Phusphate; CAii • Calcium Nllmontum Nitrate, NPK • Nitro-Phospho-Potash; OAP • ,J!-Ammun!um Phusphat~. b1 FO • fuel oll; N • naphtha; C - roali MC - natttral ~as; A - ammonia; COc.i ... coke 011en .-tas. cl ReQott llll! pertods from April through March, it OWned Jo!ntlv with private •ector. ~: fertilizer As•odatton of tnd!a. Industry Department December I ~116 - 34 - 3.05 Achieving such a target would be helped by the implementation of a two-pronged strategy. First, raising the productivity of existing poorly performing plants would be highly desirable.. Accelerated growth in output • is constrained by the uneven performdnce among firms, several of which employ older technologies and consequently nave high production costs as well as low capacity utilization (Table 2). The public sector, in partic- ular, could raise its performance through a combination of modernization, improved planning and better mana~ement. 3.06 Second, the regulatory policies implemented in the past to encourage the expansion of thP. lndus~ry should be re-evaluated. Although the pricing policy has been effective in promoting continued production by plants with different costs of production, it has also supported high cost producers and minimized price competition in a seller's market. Since nitrogenous fertilizer pro1uction is now well-established, it may be appropriate to modify the retention price formula to more fully reflect economic pr.ices~ and to move as quickly as ?Ossible to a free market pricing system. 3.07 This review examines the regulatory framework in the nitrogenous fertilizer industry as part of a cross-sectoral study that seeks to eval- uate the impact of regulatory policies ~n industrial behavior and perform- ance. In Section B, the GOI's licensing pvlicies for the fertilizer sector are described, and the nature of entry, expansion and exit barriers is evaluated. The ~etention price formula is also reviewed and placed in historical perspective. Section C then ass~sses the impact of these policies on market structure, firm behavior and performance. B. REGULATORY POLICY FRAMEWORK 1. Licensing 3.08 The fertilizer industry follows the same licensing procedures required for all large industcial projects: firms seeking to expand or enter the industry must apply for a license. To make the process more efficient, however, all decisions for this sect~r are made by the Committee of Secretaries on Fertilizers, which oversees all licensing requirements, including those for entry, expansion, capital goods and foreign exchange clearances. The industry is also subject to the Monopolies an~ Restrictive Trade Practices (MRTP) Act, which requires private sector firms of a certain size or dominance to obtain additional clearance. Firms seeking to expand the c3pacity of existing plants by less than 25 percent need not, however, obtain a license; for expansions in excess of 25 percent, the procedure ls the same as for new plants. This provision facilitates some balancing and modernization but retains control over major expansions. 3.09 A limited number of licenses are issued for each Five-Year plan- ning period. The number is based on the GOI's estimation of: (i) the expected shortfall in supply, given estimated levels of demand and imports; and (ii) the expected availability of raw materials and infrastructure for ·'I - 35 - "1 naw plants. Before evaluating the applications, the GOI decides on the numbar and size of plants to be 'commissioned, their location, the feedstock .. (raw material) to be employed, the output produced and the plant techno- logy. Once these decisions have been made, the applications are evaluated ',on the basis of how closely they conform to the government's expansion plan. Such factors as past performance are also considered, as well as an appropriate public/private sector mix. 3.10 For the Seventh Plan period (1985-90), the GOI licensed 10 new ·plants with a capacity of 1,350 tons per day (tpd) ammonia. All the plants are to produce urea, the most efficient of the nitrogenous fertilizers. Under the GOt's current policy, natural gas is the preferred feedstock, followed by naphtha and coal. Consequently, all IO plants will utilize natural gas. Three of the plants are built by existing public sector firms, three by the cooperative sector and the remaining four by new private sector enterprises. All but 3 of these plants are completed or under construction. 2. Pricing 3.11 Nitrogenous fertilizers are covered by a complicated system of price controls that establishes individual ex-factory prices for producers and retail (farmgate) prices for farmers. Each unit is guaranteed a 25 percent pre-tax return on capital investment when the plant· operates at 80 percent capacity utilization and satisfies specified consumption norms set by the Fertilizer Industry Coordination Commission (PICC). Following the oil price increases in 1974, the government established a price pooling arrangement whereby f armgate prices were set on the basis of the weighted average costs of imported and locally produced fertilizers. However, a new pricing formula was introduced three years later in response to concerns that uniform prices did not take into account the significant variations in such critical factors as size, technology, feedstock and vintage. In light of the recommendations of the Marathe Committee (1976), in 1977 the current system of individual retention prices and a single f armgate price was introduced. The objectives of this pricing policy are to: (i) ensure the viability of fertilizer plants; (ii) motivate the industry to invest in new plants; (iii) give units an incentive to operate efficiently; and (iv) ensure that the s"bsidy to farmers does not become too large a burden. 3.12 The pricing formula allows the pre-tax return on investment to rise above or fall helow 25 percent as capacity utilization rises above or falls below 80 percent. In fact, the return on equity can be much higher than 25 percent because some firms have lower tax liabiliti~s due to investment credits and operate at capacity higher than 80 percent. This high "effective" rate of return has in fact encouraged applications for entry from firms in recent years; there have even been significantly more applications than licenses granted. 3.13 In contrast to other industries, where set prices often remained fixed for several years, the retention prices are appropriately subject to frequent revisions. In addition to revisions at the end of e&ch three-year pricing period, at which time the retention prices .for all plants are - 36 - reviewed, changes can he made annually (to reflect wage increases) and even on a shorter term (to reflect increases in feedstock prices) •.~/ Since the price of feedstocks varies significa~tly, there is a large difference iu • the retention prices paid to different units. The methodology for calcu- lating the input norms is not officially published by the FICC, so it is difficult to evaluate the procedure for determining individual retention prices. Periodically, the FICC estimates and revises the costs, based on plant-specific norms and actual performance with respect to cap~city utili- zation, working capital requirements and raw material costs. 3.14 Farmgate prices are set at the same level throughout India-- US$188 (at Rs 12.50/US$) per ton of urea in 1985/86. This price is 33 per- cent below the weighted retention price of urea (US$287 per ton, including the cost of freight and distribution) and 15 percent above the cost of imported urea (average US$164 per ton). Inasmuch as the farmgate prices have remained essentially unchanged in current terms since 1981, (in line with the government's policy of stimulating agricultqral production through increasing fertilizer application) they have actually fallen in real terms. 3. Exit Policy 3.15 In 1984-85, six plants in two public sector companies, HFC and FCI, were estimated by the Department of Fertilizers to have incurred losses which cost the government an amount greater than the net after-tax revenues the GOI received from all other public sector nitrogenous .fertil- izer plants. There have been discussions regarding closure or management take-over by private firms, but no plants that are making losses have been closed in the last five years, nor have any alternative management schemes been implemented. The rationale for this policy is only partially related to labor policies. The FAI has for example stated that it is cheaper to continue operating some plants at low levels of capacity utilization than to replace them with new plants, given current capital costs. 4. Import Policy 3.16 Imports of fertilizers are canalized through the government-owned Minerals and Metals Trading Corporation and distributed by the Central Fertilizer Pool through various institutional agencies. Imports of nitro- genous fertilizer have declined steadily, from an average of 70 percent of total consumption in 1960-69 to 37 percent in the 1970s and 26 percent between 1980 and 1985 (Table 3). There is no direct price competition between domestic and imported products, either at the producer or consumer level. The GOI determines domestic producer and.consumer prices separate- ly, and the farm-gate price is set at the same level for both imported and indigenous produ~ts. Nor is there any explicit policy linking domestic prices to international fertilizer prices, as there was prior to 1977. 2/ Anand P. Gupta, National Workshop on Fertilizer Consumer Prices, Bangalore, April 25-26, 1986, New ~elhi Office, World aank, New nelhi, P• 3. - 37 - Table 3: NITROGENOUS FERTILIZER CONSUMPTION Consumption Growth Imports Growth Imports as % Year ('000 tons) Rate ('000 tons) Rate of Consumption 1960-61 211. 70 399.00 188.47 1961-62 249.80 18.00 307.00 -23.06 122.90 1962-63 333.00 33.31 244.00 -20.52 73.27 1963-64 376.10 12.94 228.00 -6.56 60.62 1964-65 555~20 47.62 232.00 1.75 41. 79 1965-66 574 •.80 3.53 326.00 40.52 56.72 1966-67 737.80 28.36 632.00 93.87 85.66 1967-68 1,034.60 40.23 867.00 37 .18 83.80 1968-69 1,208.60 16.82 844.00 -2.65 69.83 1969-70 1,356.00 12.20 667.00 -20.97 49.19 1970-71 1,479.30 9.09 477.00 -28.49 32.24 1971-72 1,798.00 21.54 481.00 0.84 26.75 1972-73 1,839.00 2.28 665.00 38.25 36.16 1973-74 1,829.00 -0.54 659.00 -0.90 36.03 1974-75 1,705.70 -6.74 884.00 34.14 51.83 1975-76 2, 148_.60 25.97 996.00 12.67 46.36 1976-77 2,456.90 14.35 750.00 -24.70 30.53 1977-78 2,913.00 18.56 758.00 1.07 26.02 1978-79 3,419.50 17.39 1,228.00 62.01 35.91 1979-80 3,498.10 2.30 1,295.00 5.46 37.02 1980-81 3,678.10 5.15 1,510.00 16.60 41.05 198i-82 4,068.70 10.62 1,055.00 -30.13 25.93 1982-83 4,263.00 4.78 425.00 -59.72 9.97 1983-84 5,236.00 22.82 656.00 54.35 12.53 1984-85 5,486.00 4. 77 2,009.00 206.25 36.62 1985-86 5,812.00 5.94 n.a. Source: Fertilizer Statistics, 1984-85, Fertilizer Association of I~dia; Industry Department, The World Bank. 3.17 The GOI also regulates imports of fertilizer equipment. Until recently, the purchase of foreign equipment was discouraged through customs duties imposed on both imported capita1 goods and on equipment for mainte- nance. Recently, the duties on capital goods for fertilizer products were eliminated, but authorization is still needed for their import. Moreover,. the duties still apply to equipment for maintenance. S. Regulation of Public Sector Firms 3.18 Public and private sector firms follow the same licensing proce- dure, and both sectors come under the retention pricing scheme described above. Nonetheless, in the past there might have been a bias in the · licensing process towards public sector expansion. Despite greater ease of - 38 - licensing public sector enterprises were under greater pressure to employ domestic equipment and to experiment with new technology or feedstocks. In 1969, the FCI undertook to establish three coal-based plants, in the • interest of exploiting India's relatively abundant coal reserves, but in view of & number of difficulties, only two were built--FCI-Ramagundan and FCI-Talcher. One of the more critical problems in the industry has been the inability of these two small but high cost plants to overcome technical problems. Although public sector plants are under the same pricing policy as the private sector, capacity utilization at the two coal-based plants has been set at 60 percent of actual capacity at the plants for purposes of setting retention price. In 1985-86, the average capacity utilization of these two plants was 24 percent. 3.19 In past years, the boards of directors of public sector corpora- tions have been drawn primarily from the civil service, while private sector boards have been drawn from the business community. According to the FCI, there has been a broader social orientation to decisions at public sector enterprises (e.g., employment creation) than at private firms. The FCI gave as an example that there is work for only one-third of the 6,500 employees at the FCI-Sindri plant, yet the size of the workforce has been maintained. C. IMPACT OF REGULATORY POLICIES 1. Structure 3.20 In the nitrogenous fertilizer industry, entry and exit are essen- tially determined by regulatory policies. The industry is characterized by economies of scale; unit· costs continue to decline for ammonia plants up to a capacity at 1,350 tpd. In an unregulated environment, larger firms capable of installing plants of economic scale or expanding existing f acil- ities would have a strategic advantage over smaller firms. However, since the licensing procedure determines entry and exr.ansion, large firms have no advantage due to size. At the same time, the government has ensured that economies of scale will be exploited by imposing a size requirement for new plants of 1,350 tpd. This imposition eontrasts with policies pursued in other sectors, where promotion of small-scale units has led to the estab- lishment of high-cost plants at suboptimal economic scales. Nevertheless, entry is limited: for the Seventh Plan period, the Department of Fertil- izers estimated that there were two to three times as many applications as the number of plants to be built. Applications for licenses are already being accepted for the Eighth Plan period (1990-1995); whenever a new plant is considered, all firms now automatically apply. 3.21 Public sector firms have in the past accounted for most capacity and have continued to increase their share until recently. Between 1978 and 1985, the public sector increased its share of licensed capacity from 51 percent to 55 percent, while the share of the private sector dropped from 42 percent to 27 P.ercent. This is expected to change with the - 39 - completion of the new 10 gas-based plants. By 1992, the capacity shares of public sector firms will have receded to 43 percent. The predominance of publi~ sector enterprises is due to a number of interrelated factors. Prior to 1977, profits were not sufficiently high to attract many private investors. In addition, firms were obliged to borrow investment funds from the IDBI, which reserved the right to convert loans into equity. This method of financing deterred private investors, and consequently the public and joint sectors flourished. Although attractive rates of return in the fertilizer industry have stimulated entry by increasing numbers of private firms were until recently still favored in the licensing process, although currently this seems to be no longer true. 3.22 During the Seventh Plan, it appears that the licensing committee has sought to promote entry by new private sector firms but to consolidate public sector expansion within existing firms. Four new private sector firms have been licensed for greenfield plants, but no existing private firms have been grant~d licenses for building new plants •.~/ This licens- ing decision serves to reinforce a historical trend towards multi-plant public sector firms and single-plant private firms. While all private firms are single-plant firms, public sector enterprises manage an average of 2.2 plants (Table 2). The GOI's reluctance to permit private firms to expand to multi-plant enterprises may be short-sighted, if the private sector has benefited from learning by doing and could incorporate this knowledge in new·projects. Moreover, given the poorer performance of public sector units, there has been some controversy over the licensing committee's decision to award 3 of the 10 new units to the public sector. It appears that these units could have been assigned to private or cooperative sector firms. 3.23 Licenses have also been difficult to obtain for expansion of existing plants in both the public and the private sector. Limiting expan- sion seems inefficient, since (1) the industry is characterized by scale economies and (2) capital costs for greenfield plants are high. Production capacity could be increased at less cost if existing plants (in both public and private sectors) below minimum efficient scales were permitted to expand more easily. 3.24 In late 1986, 25 major firms were producing nitrogenous ferti- lizers. These firms operated 46 plants, with a total capacity of 6.7 mil- lion tons per year (tpy) of fertilizer; thus the average plant size was about 145,000 tpy or 441 tons per day (tpd).4/ No single firm or group of firms dominated the market (the four-firm co-;centration index was 43 per- cent). The distribution of firm sizes was quite broad, ranging from new plants with capacities of 1,350 tpd to plants of 100 tpd. Although the newer plants are being commissioned at sizes that fully exploit economies of scale, the same is not true for some of the older plants. ]} Note that this discussion refers to the nitrogenous fertilizer sector only. GSFC has been granted a license to build a DAP plant. 4/ Daily production capacity is based on 330 working ~ays per year. - 40 - 3.25 The smaller or technologically obsolete firms continue to operate because of exit barriers and the price retention formula, two interrelated factors. The retention prices make it possible for some inefficient firms· to operate profitably, since their prices are adjusted to account for vintage, technology and size. Other plants that make losses despite the pricing structure have not been allowed to close. Although the retention price formula may provide the correct signals regarding eff iciency--plants operating at low levels of capacity utilization are in fact making heavy losses--exit for these plants has not been possible. For plants which can- not be successfully renovated, closure or takeover by private sector management may be desirable. The government argues that continued opera- tion of some older, less efficient plants is economical since capital costs for new plants are high. The actual exit decision should be on the basis of long-run economic viability of the plant. 2. Behavior 3.26 Since the market is characterized by an increasing number of firms, in an unregulated environment these firms would be expected to engage in price competition. Even without many domestic f~rms, the high volume of imports could provide a measure of competition. Nevertheless, in the present regulatory environment, producers have faced very limited competition from either imports or each other. The producer prices for each plant are set individually, with no effort to link the prices of output from similar plants. The variance in the individual retention prices is large: in 1985-86, they range from rupees (Rs) 1,704 per ton to Rs 4,664 (Table 4). 3.27 The pricing formula does provide an incentive to raise capacity utilization and reduce operating ~osts. Since higher levels of capacity utilization reduce per unit costs, the inducement to operate above 80 percent capacity utilization encourages more efficient production. How- ever, the inducement to cut costs is not as strong, and high cost firms can continue to operate profitably within their consumption norms. The formula also fails to provide incentives for minimizing construction costs. In fact, producers have an incentive to show that relative to others, their costs of production are high and their production at 80 percent capacity utilization is low. On the other side of the market, consumers offer no constraint on changes in producer prices, since subsidies shield them from the price effects. Nor are there any market forces--either on the producer or consumer side--to motivate producers to seek minimum cost-price output combinations, and prices do not serve as a signal regarding the efficiency of firms. Consequently, the burden placed on the retention price formula to ensure behavior that approximates market forces is substantial. 3.28 One reason for the disparity in prices is the difference in feedstock prices payable by the plants. For example, in 1984-85 GSFC- Baroda paid only Rs 320 per 1,000 M3 of gas, while RCF-Bombay paid as much as Rs 1,790 per 1,000 M3.~/ The average price of natural gas in Indla is relatively high, at $3.20 per million BTU, compared to between $2 and $2.50 2../ A. Gupta, op. cit. - 41 - . . in Europe, less than $1 in Pakistan and Indonesia, and about $2 in the U.S. However, these comparisons must be approached with caution, since local natural gas prices vary tremendously across geographic regions. Table 4: PLANT RETENTION PRICES, OCTOBER 1986 Retention Price Ownership Vintage (Rs/ton) 1. HFC, NAMRUP Public 1969 1, 661 2. SFC, KOTA Private 1969 2,602 3. GSFC, BARODA Joint 1969 2,642 4. IEL, KANPUR Private 1970 3,246 s. MFL, MANALI Public 1971 2,781 6. FACT, COCHIN Public 1973 3, ll6 7. ZACL, GOA Private 1973 2,800 8. RFC, DURGAPUR Public 1974 3,291 9. SPIC, TUTICORIN Joint 1975 2,681 10. IFFCO, KALOL Cooper~tive 1975 2,615 11. MCFL, MANGALORE Joint 1976 2,715 12. HFC, BARAUNI Public 1976 3,583 13. RCFL, TROMBAY 1 Public 1976 4,267 14. FCI, GOARKHPUR Public 1976 3, 774 15. NFL, NANGAL Public 1978 3,134 16. NLC, NEYVELI Public 1979 3,688 17. FCI, SINDRI Public 1979 3,384 18. NFL, PANIPAT Public 1979 3,628 19. NFL, BHATINDA Public 1979 3,706 20. FCl, TALCHER Public 1980 4,630 21. FCI, RAMANGUNDAM Public 1980 4,605 22. IFFCO, PHULPUR Cooperative 1981 3,496 23. GNFC, BHARUCH Private 1982 3,651 24. RCFL, TROMBAY V Public 1982 3,801 25. RCFL, THAL-VAISHET Public 1985 3,593 26. KRIBHCO, HAZIRA Cooperative 1986 3,530 27. NFL, VIJAYPUR Public 1988 4,477 28. IFFCO, AONLA Cooperative 1988 ·4, 552 29. INDO-GULF, SULTANPUR Joint 1989 4,700 30. CAPARO, SHAHJAHANPUR Private 1991 5,020 • 31. ARAVALI, SAWAI MADHOPUR Private 1991 5,192 32. TATA, BABRALA Private 1992 5,373 Source: Industry Department, World Rank, Fertilizer Restructuring Study. Note: Plants have been grouped on the basis of increasing order of vintage. - 42 - J.29 Another factor affecting price variability is the conflicting nature of the incentives embodied in the retention price foI'T!lula. The system was designed to ensure that operational inefficiencies do not raise a plant's retention price. Producers are expected to recover their costs and earn a return on investment only if their plants operate at acceptable input consumption norms and levels of capacity utilization. A 1985 study by the Indian Bureau of Public Enterprises, which surveyed 11 plants using a variety of feedstocks and technology, found that 7 of the 11 reported that their breakeven levels at FICC norms were between 43 and 56 percent of capacity.6/ If management displays a "satisfying" (as opposed to profit maximizing) pattern of bahavior, the pricing formula would leave substan- tial room for x-(or managerial) inefficiencies. 3.30 The pricing system has been designed to penalize producers for capital cost increases suffered during construction as a result of delays beyond the original planned completion dates or because of increases ln pre-operating, interest and project management expenses. It has been estimated that the capital costs of the ~rivate sector plants currently being built in India may be higher than those of internationally competi- tive plants by about 10 percent. Several factors may be contributing to these higher capital costs. First, the government required that most (all but two) of the 10 gas-based fertilizer plants use the same ammonia techn0- logy and main contractor. Although the technology is considered reliabl~ and competitive, the absence of competition has signified that India i~ not benefitting adequately from a potentially favorable international contract- ing conditions. Second, the pricing system does not appear to give project owners an incentive to build a low-cost plant. Instead, they look to designs that favor more secure, sustained production capdbilities. Since prices are based in part on the value of investment, there is an incentive to over design plants. 3. Performance 3.31 Growth in Output and Capacity Utilization. The rate of growth in output for nitrogenous fertilizers averaged over 20 percent annually between 1960-61 and 1970-71, hovered at 10 percent annually between 1970-71 and 1980-81 and has held at 7.5 percent annually since 1981 (Tabl~ I). Capacity grew at an average of 24 percent annually in the 1960s, 23 percent in the 1970s and·9.6 percent since 1981. Despite high (although declining) levels of imports, capacity utilization has been relatively low, (ranging between 53 and 75 percent) although it has been rising in recent years. 3.32 Immediately following the introduction of the retention price formula in 1977, overall capacity utitization increased, but the trend was not sustained, and it is difficult to conclude that the formula has had a positive impact on capacity utilization. The levels of capacity utiliza- tion were low in the late '70s into the '80s, predominantly because of poor performance by the public sector. In 1984-85, that sector operated at an average rate of only 60 percent (nevertheless, it is a substantial increase over the 41 percent of 19e0-81), while the private and cooperative sectors Interfirm Comparison of Fertilizer P\ants, Bureau of Public Enterprises, Ministry of Finance, Government of India, September 1985, P• 65. - 43 - attained rates of 81 percent and 89 percent, respectively. The low rates of capacity utilization for public enterprises were not confined just to the older plants, but applied to those of different sizes and utilizing different feedstocks. 3.33 A survey conducted by the FAI indicated that 79 percent of the losses in production across all sectors in 1984-85 were attributable to either equipment breakdowns, operational problems or plant shutdowns. A recent study by the Ministry of Finance comparing different fertilizer plants also found that equipment failures predominated in those plants that had performed the worst. The study noted that these plants had been built with too great a reliance on unproven technology and design. In the public sector, poor design and faulty equipment are combined with ineffi- cient management, which tends to act slowly in identifying equipment prob- lems. This lack of speed in identifying and correcting equipment failures is particularly crucial when a plant first goes into operation and can lead to chronic inefficiencies if not rectified immediately. 3.34 Price Comparisons. Until 1982, the average ex-factory retention price compared favorably with the CIF prices for imported urea. However, as early as 1982, because of a world-wide downward trend in fertilizer prices, the CIF pri~e of imported urea had fallen below the domestic price by 20 percent. In 1984-85, domestic producer prices and imported urea were equivalent according to the above estimates. The market has fluctuated significantly since that period, and in early 1986 GOT purchased urea as low as US$70 FOB, or $100 CIF. However, the World Bank has estimated that fertilizer prices will return to their long-run equilibrium levels of US$182 FOB by 1990, which is approximately $215 CIF, and (assuming a constant nominal exchange rate) Rs 3,1~0 per ton. 3.35 The retention price will continue to be comparable with interna- tional prices if prices grow at a nominal rate of not more than 1.2 percent annually. Given past trends in price increases and the higher capital costs of new plants beginning operation during 1985-90 (translating into relatively high retention prices), such minimal price increases will be difficult to achieve. Consequently, continued efforts to reduce ineffi- ciency and renovate older units are highly desirable. It is important to recall that many producer prices are much higher than the average: some producers are receiving as much as US$380 per ton, which is 60 percent higher than the long-run equilibrium world price (CIF) calculated by the World Bank. 3.36 Although the low international prices of early 1986 indicate a zero subsidy for imports, the retail prices are at least 20 percent lower than the average ex-factory prices. As a result, fertilizer subsidies in India are substantial. 3.37 Fertilizer subsidies are responsible for two-fifths of the total outlay of central government subsidies and have grown at a nominal rate of over 20 percent annually since 1979; in 1984-85, they amounted to about US$1 billion and reached US$1.6 billion in 1985-86. The government has recognized the implications of maintaining the retail subsidies and raised fertilizer retail prices by IO. percent in February 1986. - 44 - Table S. PRICES FOR UREA, RS PER TON, '1979-86 Retail (Farmer) Average Producer Year Price Retention Price 1979-80 1,450 1,848 1980-81 2,000 2,006 1981-82 2,350 2,538 1982-83 2,350 2,913 1983-84 2,150 2,773 1984-85 2,150 2,940 Feb. 1986 2,350 Sources: 1979-83: Fertilizer Association of India; The World Bank. 3.38 The GOI has expressed concern that sudden price increases could discourage fertilizer use, which is still one of the lowest in the world. However, it is also important to realize that nitrogenous fertilizer application is heavily concentrated in the irrigated areas (which account for 85 percent of fertilizer use). The spread of fertilizer consumption is not a simple question of keeping consumer prices low but is also a question of extending its distribution to rain-fed areas, which requires extension and credit services, as well as agricultural research and other efforts not ex,licitly connected with pricing. 3.39 Examining unit input coefficients related to production avoids the pitfalls of price comparisons and allows a more direct measure of ef f i- ciency. Fertilizer plants in India were characterized in 1984-85 by the use of relatively high levels of feedstock, consuming on average between 10 (natural gas) and 18 (coal) million kilocalories per ton of annnonia, compared to 7.3 million kcal per metric ton of ammonia produced at modern international plants.7/ Although natural gas is the most efficient feed- stock in terms of kcal consumed per ton of ammonia produced, only recently has it been possible for India to begin exploiting its natural gas reserves for use in fertilizer production. In 1985, 44 percent of total plant capacity used natural gas as a feedstock, compared to 73 percent globally. With the implementation of additional gas-based plants, this figure is expected to rise in the future. 7/ Ibid., p.7 - 45 - INDIA - INDUSTRIAL REGULATORY POLICY STUDY IV. Subsector Review--Steel and Iron A. BACKGROUND 4.01 Indian steel production started in 1912, with TISCO, a private sector company having the first integrated steel plant in India. Its original capacity was 100,000 tons, reaching 1 million tons by 1939 and 2.0 million by 1959, a level it maintained until 1983. TISCO planned an expansion program in the early seventies, but it never materialized. The government opposed the program, fearing that it might lead to excessive capacity, given projected demand. TISCO in turn did not make great efforts to persuade the government to accept the program because (i) it realized that the expected returns on investment were low, given the price controls, and (ii} the condition under which the government would finance the program, namely, a convertibility provision whereby debt could be converted ·into equity, led TISCO's management to fear it might lose control of the company. IISCO (Indian Iron and Steel Company, another private firm) started producing steel in 1918, reaching 1 million tons by 1962, and in 1978 merged into SAIL, a public sector holding company. 4.02 The public sector entered the steel industry in the 1960s with three integrated steel plants. At this time the Government decided that new integrated steel plants would be a public sector activity thus introd~cing a major barrier to entry and expansion for private firms. In the roid-70s, another SAIL plant had been built, increasing the holding company's total capacity to 9.4 million tons. Since then, the rated capacity of the integrated plants has remained the same. On the other hand, mini-steel plants have increased their capacity rapidly from slightly over 1 million tons in 1974 to 4 million tons in 1984. Thus, ln 1984 the rated capacity of steel production facilities was 9.4 million tons for SAIL, 2.0 million tons for TISCO and 4 million tons for the mini-plants (Table 1). 4.03 The Indian steel industry was internationally competitive in both price and quality at an early stage in its development. By the second World War, India produced and exported armor-plated steel for tanks and was ranked ·as one of the top five most efficient steel producers in the world. Since the mid-1960s, however, the steel industry has progressively lost its international competitiveness. Exports of steel products, although they continued throughout the 1960s and 1970s, declined to zero in net terms by 1981, after having reached 1 million tons of pig iron and 1.4 million tons of steel in 1976. Imports of mild steel rose from 0.4 in 1976 to 1.0 million tons in 1981. Since the mid-seventies, net domestic d~mand has persistently exceeded supply. - 46 - Table 1: STRUCTURE OF CAPACITY (millions of tons/year) Late 1970s/ Ownership Late 1940s Mid-1960s Present Early 1990s Private Sector TIS CO 1.1 2.0 2.0 3.5 IISCO o.s 1.0 Minis/Midis 0.1 o.s 3.5 &.& (est.) Subtotal 2.3 3.5 5.5 9.5 Public Sector (SAIL) IISCO 0 0 1.0 1.0 Rourke la 0 1.0 1.8 2.0 Bhilai 0 1.0 2.s 4.0 Durgapur 0 1.0 1.6 2.0 Bokaro 0 0 2.5 4.0 Vi zag 0 0 2.s Subtotal 0 3.0 9.4 15.5 Total 2.3 6.5 14.9 25.0 % Public 0 46 63 62 Source: World Bank Steel Sector Report; mission estimates. Table 2: CONSUMPTION OF STEEL, 1965/66-1985/86 (mil lion tons) Apparent Year Production Imports Exports Consumption 1965/66 4.4 0.1 0.1 5.0 1970/71 4.5 0.6 o.s 4.6 1975/76 5.8 o.s Q.8 s.s 1980/81 6.8 1.0 0 7.8 ' '" 1982/83 s.1 1.3 0 9.4 1983/84 8.4 0.6 0 9.0 1984/85 8.7 1.7 0.2 10.2 1985/86 9.0 2.0 Q.3 10.7 Sources: To 1983/84, World Bank reconnaissance report; and 1984/85 and· 1985/86, Steel Sector mission estimates. - 47 - 4.04 Despite the generally deteriorating performance of the integrated producers, there has been great diversity between TISCO and the individual plants under SAIL. TISCO has performed much better than the average for SAIL, while Bhilai and Bokaro have consistently achieved better results than the other SAIL plants (Tables 3 and 4). There has been a significant • decline in capacity utilization and in productivity in SAIL's IISCO plant (Table 4). India's original sorrces of competitive adnantage in steel-making--low wages and cheap material inputs--have been neutralized by poor productivity and input quality. 4.05 Five key factors have been identified as having had important adverse effects on the structure and performance of the steel industry. (1) Price controls posed a severe constraint on modernization investments during the 1970s because the returns on capital were too low to yield investible surpluses and therefore to stimulate investments. (2) Do~estic regulations on investment and capacity expansion restrained producer flexibility and internal competition. (3) Since the mid-1970s, import policies that have shielded domestic producers from foreign competition have lessened the pressure to be internationally competitive. (4) Poor quality and supply of material inputs adversely affected productivity and capacity utilization. (5) Management of public sector units had insufficient autonomy and incentive to minimize costs and to change the composition and level of output and respond to market changes. There was widespread, but not universal, slackening of managerial and work standards, with political and bureaucratic interference in the staffing and running of the enterprises. Table 3: COMPARATIVE PRODUCTION COSTS, 1984/85 (US$ equivalent per ton) Production Costs FOB Price, Europe, Europe, Japan, Japan, Product Rourkela Bhilai Bokaro TIS CO Brazil Brazil Coke 87 96 77 67 90-100 Hot Metal 143 133 125 92 100-110 Ingot (OH) 282 210 209 169 150-175 Ingot (LD) 239 Slab (Rolled) 292 251 185 160-195 Slab (Cast) 150-180 150-170 Billet 250 197 165-200 170-180 HR Coil 370 286 226 190-240 220-2AO CR Sheet 595 395 240-300 275-300 Galvanized Sheet 718 290-350 328-360 Light Section 284 231 200-240 235 Sources: SAL:; World Steel Dynamics; CRU Metal Monitor; and others. - 48 ... Table 4: COMPARISON OF AVERAGE REVENUE COST PER TON OF SALABLE STEEL AND CAPACITY UTILIZATION SAIL IISCO TIS CO Avg. Avg. Cap. Avg. Avg. Cap. Avg. Avg. Cap. Rev. Cost Util. Rev. Cost Util. Rev. Cost Util. Year Index Index % . Index Index % Index Index % 1981/82 100 100 72 100 100 60 1982/83 112 120 72 105 122 62 1983/84 127 139 64 108 139 54 100 100 91 1984/85 151 161 68 111 143 44 119 109 95 Source: Annual Reports of SAIL, IISCO and TISCO. B. CURRENT REGULATORY ENVIRONMENT 4.06 In recent years, the GOI has taken measures to allow steel producers to invest in modernization and expansion. These measures include: (i) allowing TISCO and t-wo SAIL plants to implement modernization/expansion investments;.and (ii) allowing the privat~ mini-plants to expand by up to 50 percent of licensed capacity when 100 percent utilization is reached rather than the usual 25 percent over five years, and to diversify frQ.ely into the production ·of all grades of carbon and alloy steels. In 1984, the GOI decided to let the main producers determine the prices of steel products, which had been under government control since World War II. It should be noted that the products of the mini-steel plants and secondary reroller plants were never under the purview of price controls. 4.07 In response to these policy changes, steel producers have started to implement modernization progr8ms. TISCO is upgrading and expanding its integrated plant from 1.74 million tons to 2.10 million tons of saleable steel and in considerin~ major ex;>ansion. The largest mini·~steel plant, Mukand Iron and Steel Works, Ltd., is doubling its capacity to 270,000 tons. 4.08 Although reforms in the regulatory environment have been carried out, many controls remain. As with many other industries, the manufac- turers of primary steel products ar~ required to obtain industrial licenses from the gove~nment except when (i) the increase in output is marginal, i.e., less than 25 percent of existing licensed capacity, or (ii) expansion is for replacement/modernization/renovation of equipment. lf a marginal expansion involves financing from external sQurces or substantial imports of capital equipment, then appro,1al is required. In the case of Monopoly - 49 .• Restrictive Trade Practices Act (MRTP) companies, any expansion is subject to clearance if the additional capital outlay is more than 25 percent of the existing value of assets. Since the value of assets is based on historical costs, while capital costs have increased substantially, even expansion of less than 25 percent must go through MRTP clearance. Moreover, as incremental investments are generally not the most efficient way to initiate any major expansion in steel capacity, this provision is of lird.ted value. 4.09 The prices for all products of the main steel producers are now determined by the Joint Plant Committee (JPC) which is chaired by a government official. They are generally set to cover the costs of production, but the Indian Railways exert pressure to keep prices for its products down. The government previously set them on the basis of average costs plus return on capital. During the 1960s and early 1970s, the regulated prices were generally lower than market prices because India was an exp~cter and there was a time lag between the cost inquiry and the fixing of prices. This discrepancy squeezed the profit margins and limited the gen~ration of internal resources for modernization and expansion. 4 .10 In the early 1970s, a dual pr.ice structure was introduced whereby government purchasers and some priority users were granted lower prices, than other users. A problem with this system was the cross-subsidization in favor of the government sector versus the private sector. This system lasted U'ltil the early 1980s. 4.11 Steel producers are still required to contribute some funds, based on set rates per ton of steel, to the Steel Development Fund, Engineering Goods Exports Assistance Fund (EGEAF) and Freight Equalization Fund, in addition to paying the excise duty. As a result, the destination prices are higher than the retention (producer) prices by a 20-50 percent margin. That margin is broken down into: (i) the excise duty, at rates of Rs 365-935 per ton; (ii) Steel Development Fund (SDF), at rates of Rs 300-500 per ton, with the revenue used to finance the modernization and development of the steel industry at a subsidized interest rate; (iii) Engineering Goods Exports Assistance Fund, at Rs 100 per ton, with the revenue used to provide the export industry with steel at international prices; (iv) Freight Equalization Fund, at rates of Rs 320 per ton for pig iron and Rs 470 per ton for steel products with the revenue used, to supply products at uniform rail head prices throughout the country; and (v) JPC charges, at Rs 5 per ton, used to cover the administrative costs of the • JPC • 4.12 A shortcoming with the Freight Equalization Fund is .that it results in steel users located close to steel producing centers paying more than it costs to supply the steel, whereas steel users located far away pay le~s than it costs to deliver the steel to those locations. Hence, it has encouraged steel users to locate near their marketing centers rather than in the vicinity of the steel plants, even when this location may not be the most efficient one if they paid the full cost of steel inputs. The Engineering Goods Exports Assistance Fund is a type of drawback scheme for exports. It is designed to compensate the engineering goods ~xport - 50 - industry for the high steel prices it faces because of the various levies and taxes on steel products. 4.13 Imports of steel products into India are tightly controlled. While some items that are not produced locally can be imported without approval, most products must come in through a government agency (i.e., must be canalized). In this case, imports are permitted only to fill the gap between estimated demand and supply capacity and are available at relatively high prices compared with world levels. What imported steel products there are, are subject to heavy customs duties that vary a lot across items and for which there is a long number of user-specific exemp- tions. For example, the statutory customs duty (basic plus auxiliary) amounts to 65 percent for pig iron, 65-95 percent for mild steel products dnd 175-325 percent for stainless or speciality steel. Imports of sponge iron and scrap iron, which are used as raw materials for mini-steel plants, are canalized and subject to a 15 percent tariff duty. 4.14 For most steel products, the government projects demand and has a canalized government agency handle imports to meet the shortfalls. Leaving aside the accuracy of the projection, imports are in most cases not timely because the procedures for decision-making and implementation take time. Even when imports are well-managed, in the sense that supply meets demand overall, there tend to be shortages of some products because the items are too diversified in terms of kinds and qualities to be managed. 4.15 ·As with other industries, the foreign exchange requirements with regard to foreign collaboration, imports of know-how and technology, imports of capital goods and coking coal, ferro alloys and sponge iron have to be cleared through the government. Until recently, steel producers were forced to use domestic coking coal, which is of poor quality and lower price. This requirement, together with the limitations of domestic avail- ability, have restricted productivity increases. In recognition of this problem, recently the GOI decided to allow steel producers to import coking coal. 4.16 In adeition to the controls on expansion, pri~es and imports, government intervention in the management of SAIL, the dominant steel producer with a market share of 60 percent, is another important regulatory constraint. In fact, SAIL has lacked autonomy and has been subject to considerable intervention in almost every area of activity--pricing, choice of product mix, marketing, finance, investment and personnel. The govern- ment has stressed the importance of: employment generation in regional areas; reaching volume targets with little consideration for the costs of production; and basing the product mix on the need~ of other public sector users. Moreover, adequate attention has not been paid to maintenance and modernization. In day-to-day operations, SAIL has had answer a never- ending stream of questions from officials. Moreover, investment and production decisions take a long time for the bureaucracy to make, thus constraining SAIL's capacity to respond to market opportunities or needs. The modernization plan for the Durgapur mill, which has dragged on since 1978, is one example. - 51 - C. INDUSTRl.AL STRUCTURE- AND BEHAVIOR 4.17 The regulated industrial environment has given rise to a rela- tively fragmented structure. There are now five integrated steel plants (TISCO, IISCO, Rourkela, Bhilai and Bokaro) and many mini-mills. All integrated plants and all mini-mills are below the scale that would mini- mize unit production costs. The minimum efficient size for an integrated steel plant in the 1960s was 4 million tons a year, and technological changes since then have raised the norm to 6 million tons. This level is more than twice the size of the largest Indian plant (Table 1). This situation is greatly aggravated by the fact that each of the five integrated steel mills produces a wide range of products~ If they were more specialized, production costs could be reduced. The minimum efficient size for an electric furnace mini-mill is 0.4 million tons a year, whereas the largest in India is 0.27 . million tons.I/ - 4.18 Capacity licensing and MRTP controls have inhibited both entry into steel production and of expansion capacity. Moreover, when combined with the substantial protection against imports, they have given rise to the diversified range of items produced by individual mills. With a sheltered domestic market, producers have found it easier to acquire licenses to move horizontally into additional lines of production than to concentrate on producing a limited number of goods more efficiently. Licensing and MRTP controls have made it difficult for firms to expand capacity in an efficient, non-incremental fashion. 4.19 Price controls in the domestic market during the 1960s and 1970s are said t~ have suppressed returns and hence both the incentives to invest and the capacity of firms to generate an investible surplus from retained earnings. However, this situation no longer pertains, as domestic prices exceed the landed duty-free cost of imported steel by 30-60 percent. If there had been less controlled pricing, together with less protection against imports however, maintenance and modernization efforts might have been more diligent, and the Indian steel industry would have been more efficient and competitive in the mid-1980s. 4.20 Labor regulations hav1· also contributed to the relatively ineffi- cient structure of the steel industry by inhibiting internal rationaliza- tion and the closing of some of the least productive facilities. India's integrated steel plants, particularly those in SAIL are currently burdened with far more labor than they need for efficient operations. Union strength coupled with the Industrial Disputes Act which requires government approval prior to the retrenchment of workers for large employers have limited the flexibility of management to reduce employment. lf D. G. Tarr, "The Minimum Optimal Scale Steel Plant in the Mid-1970s," processed.; - 52 - D. IMPACT ON INDUSTRIAL PERFORMANCE 4.21 The ·restrictive policies on licensing, pricing and trade, as well as the intervention of the government in public sector operations, have resulted in supply shortages and, more importantly, in the deterioration of performance over the years. The integrated steel producers have continued to employ outmoded technologies and to invest little in maintenance and improvements. As a result, price competitiveness and productivity have deteriorated. 1. Capacity Expansion 4.22 During 1966-83, while consumption of steel products grew at an average of 4.8 percent a year, the growth in domestic production averaged 4.2 percent a y.ear and resulted in an increase in imports and decrease in exports. Except for 1968-70 and 1973-75, when India was a net exporter of steel products because of the recession, India was a net importer during the period. The level of imports has increased considerably since 1977. While domestic production rose by 18 percent during 1978-83, domestic consumption rose by 35 percent. What growth there was in domestic produc- tion during this ~eriod came primarily from the mini-steel plants: their share in total crude steel production jumped· from 11 percent in 1977 to 23 percent in 1983. This share is, however, relatively low compared to that in many developed countries where scrap is more readily available--30.4 percent in the US, 28.4 percent in Japan and 100 percent in Denmark. 2. Capacity Utilization 4.23 Apart from a slow rate of capacity expansion, a key reason for the supply shortages has been low capacity utilization (Table 4). While TISCO has had good utilization rates (86-98 percent), SAIL's have fluc- tuated, ranging from 37 percent to 88 percent over the past 14 years. As of 1983/84, SAIL's utilization was 64 percent, while TISCO's was 91 per- cent, with that for the industry as a whole about 70 percent. By contrast, capacity utilization in developed countries ranges between 50-65 percent, with 53 percent in West Germany in 1983, 63 percent for Japan in 1982 and 49 percent in the US in 1982. 4.24 The reasons for the low utilization rates in India and developed countries differ. In developed countries, the causes are often excess capacity and the oversupply in the world steel market. In India, which faces a supply shortage, low utilization is explained by infrastructural bottlenecks (such as power shortages and inadequate transport facilities), input supply problems, and technical inefficienciea (related to inadequate equipment for treatment of material inputs, imbalances in various sections of the plants and aging machinery). Further, price controls might also have contributed to low capacity ~tilization. If administered prices were not high enough to cover the costs of a certain plant and the plant could not be closed because of labor problems, the way to reduce the losses at that plant might be to reduce capacity utilization. That option may partially explain the improvement in capacity utilization of SAIL from - 53 - 67 percent in 1980 to 72 percent in 1983. In particular, the capacity utilization of the loss-making plant in SAIL, IISCO rose from 57 percent to 62 percent during the same period. 3. Technology 4.25 SAIL plants still largely employ the technologies of the 1950s, although there has been a recent trend toward improvement. Raw material blending facilities are inadequate, coke ovens and sinter plants are out- dated, a substantial amount of steel-making capacity involves open-hearth furnaces (OHF) instead of the more efficient basic oxygen (or LD) furnaces, continuou~ casting is little used and a large number of rolling mills are badly maintained. OHFs accounted for 53 percent of India's steel-making facilities in 1983, with the balance comprised of basic oxygen furnaces (24 p~rcent), electric arc furnaces (22 percent) and others (1 percent). 4.26 The state of India's technology is perhaps best illustrated by comparing the nature of its steel-making processes with those in other countries. India is among a few countries in which outmoded OHFs still represent a substantial share. The percentage shares of OHFs in other countries are: USSR (59 percent), Poland (48 percent), Yugoslavia (38 percent), Romania (36 percent), China (31 percent), Philippines (23 per- cent) and Argentina (22 percent). In the de··-~loped countries, the share of OHFs ranged from nil (Japan~ France and UK) to 8 percent (US). Use of small-size blast furnaces has also contributed to higher production costs in India. 4. Efficiency of Input Use 4.27 SAIL plants are marked by a poor state of repair. The leakage of the doors on some coke ovens is 20 percent, compared with 2 percent at well-operated plants. Open hearth furnaces require heat times of 11 hours, whereas the norm is 8. LD furnace linings have a life of 140 heats, where- as up to 2,000 are attained in Japan and other countries. The blast fur- naces at Durgapur produce about 0.7 tons of pig iron per cubic meter of furnace volume/day, whereas industry standards are 2.0 tons and more. Standards of housekeeping, operating health and safety, intra-plant trans- port and equipment condition are similarly poor. Leaks of hot blast air at one blast furnace were over 15 percent, whereas the rate should be zero; they caused energy losses and unbearable working conditions. Poor worker discipline and a lack of management incentives or supervisory follow-up both cause and contribute to these problems. 4.28 The poor quality of raw materials, has also contributed to high costs and poor productivity. While the iron ore reserves in India are adequate in quantity, their quality and consistency make it difficult for mills operating blast furnaces to produce a suitable grade of hot metal for steel-making. Coking coal in India is high in ash content. The steel plants were designed to run with about 17 percent ash in the coal, but because of deterioration and the lack of washing facilities, it is not unusual for them to use coal with up to 25 percent ash.• - 54 - 4.29 The energy consumption per ton of steel . ~ very high compared with that in develop~d countries. While the specific energy consumption per ton of salable steel ranges between 10-13 gega calories at Indian steel plants, this rate has been as low as 5.0 gega calories in a Japanese steel plant, 6.5 in the US and 5.5 in the Federal Republic of Germany (West Germany). Measures such as beneficiation of coal and iron ore, increased sinter usage, better sizing, bedding and blending facility of raw materials for quality control and the consistency or the blast furnace charge could offset the losses from the deterioration in the quality of run-of-the-~111 raw materials. The need for investment for raw material beneficiation or other facilities at an existing poor performance plant has not, however, been given adequate attention. s. Labor Productivity 4.30 During 1983-84, labor productivity ranged from about 34-63 tons of crude steel per man-year for SAIL plants (excluding IISCO) and averaged about 70 tons of crude steel per man-year for TISCO. These labor producti- vity rates are low when compared with the 200-300 tons/man-year in tradi- tional European or North American plants and over 400 tons/man-year in Japanese steel plants (Table 5). The significant advantage of low labor costs in India makes up for the low productivity, so that the prices Table 5: MEASURES OF LABOR PRODUCTIVITY, 1984-85 Ton Labor Cost per Shi ppea per Ton shipped Man-year a/ (US$) Japan b/ 434 58 us b/ - 211 187 Korea, Rep. of 594 15 Latin America 78 72 India (excluding construction & mines work force) a/ 34 57 Sources: P. Marcus, World Steel Dynamics; Karlis M. Kirsis, Financial Dynamics of 60 International Steelmakers, March 1986; SAIL; TISCO; and Statistics for Iron and Steel Industry in India. a/ Includes SAIL, IISCO and TISCO. b/ The most recent data--for early 1986--indicate significant changes in labor productivity, exchange rates and employment costs, with a marked reduction in the indicated differences in tons/man-year and labor cost/ton shipped for Japan versus the US.• - 55 - are comparable with those of the leading foreign steel-producing countries. However, overmanning has prevented India's steel industry from appropriating what should be its key comparative advantage--relatively low cost labor. 4.31 More important, however, is that productivity has been stagnating rather than improving. Labor productivity in steel is a function of the scale of operation and type of processes employed and employment practices. The scale of the operations at integrated steel plants stands between 1.0-2.5 million tons of crude steel production. For these plants, the average costs per ton of steel products continues to come down until capacity reaches 5 million tons. In this sense, SAIL plants and TISCO are below the optimal scale. The mini-steel plants operate at a much smaller scale than the optimal one. The largest mini-steel plant is, as noted earlier, implementing a program to expand its capacity up to 270,000 tons, whereas the viable scale of operations is estimated to be at least 500,000 tons. 4.32 Moreover, the types of processes used are very labor-intensive. For instance, SAIL employs 205,000 managers and workers. This level of staffing is far in excess of requirements, a reflection of past government policy for the steel industry to provide maximum employment. In South Korea, while total production capacity of steel products is roughly the same as India's, total employment in the steel industry, including mini- mills, amounts to only 65,000. 6. Competitiveness 4.33 International price comparisons for 13 products that account for about 85 percent of local production are given in Table 6, which gives an idea of international competitiveness. The international prices are spot Antwerp with freight and landing charges added (using FOB prices ex-Japan makes little difference). The table gives two estimates of the domestic producer price and hence the nominal rates reported, depending on how the Steel Development Fund is treated. There is a wide range in the price dif ferenc~s and nominal rates (domestic prices over world prices) for the various steel products, and they are sensitive to the treatment of the Fund. The production-weighted average nominal rate of protection is 17 percent when only the interest element of the Fund is included in the producer price, and 27 percent if the full Fund is included. These est4- mates give some idea of the nominal incentives to production in India and the marginal cost disadvantage of domestic production. The implicit assumption is that under the existing structure of prices there is no extra-budgetary support or subsidy to the steel industry. 4.34 In conclusion, the current competitiveness of Indian steel production is relatively poor. Producers can barely cover their variable costs if they have to compete against imports. Major investments to expand capacity would not yield satisfactory economic rates of return at present prices. - 56 - Table 6: PRICE a:>MPARISONS Producer Price (Ra/Ton) Nominal Rate (percent) CIF (Rs/Ton) Ex-Europe 8 / SDF Interest Full SDF b/ SDF Interest Full SDF Subsidy b/ Subsidy Slab 2,856 2,677 3,100 -6.0 9.0 Bill~t 2,889 2,837 3,260 -2.0 13.0 Round 3,407 3,584 3,960 5.0 16.0 Equal Angle 3,504 4,620 5,090 32.0 45.0 Joist 3,504 4,730 5,200 35.0 48.0 Wire Rod 3,602 3,834 4,210 6.0 17.0 HR Coil 3,569 4,645 5,115 30.0 43.0 HR Sheet 4,179 4,975 5,445 19.0 30.0 Light Plate 4,179 4,920 5,400 18.0 29.0 Heavy Plate 4,179 5,450 5,920 30.0 42.0 CR Coil 4, 1.53 6,240 6,710 50.0 62.0 CR Sheet 4,153 6,310 6,780 52.0 63.0 GP Sheet 4,873 9,161 9,490 88.0 95.0 Avg. Weighted by Output 17 .o 27.0 Source: London Metal Bulletin and SAIL. a/ FOB prices plus $40 per ton freight from Europe (spot Antwerp); also includes 4 percent service charge and 100 Rs/ton landing and port charges. Exchange rate used is 12.45 Rs = US$1. All prices as of December 1985. b/ In the first case, a small amollllt (Rs 30 in the case of joist) has been added to the retention price in calculating the return to producers, whereas in the full SDF case, the entire SDF levy has been added to get a figure for producer returns. Note: SDF means Steel Development Fund. - 57 - INDIA - INDUSTFtAL REGULATORY POLICY STUDY v. Subsector Review--Capital Goods A. BACKGROUND 5.01 This report examines the impact of regulatory policies on the structure, behavior and performance of the capital goods industry.I/ Since the performance of this industry is closely linked with developments in both the process industries and i~put markets, these are explicitly considered. 5.02 The capital goods industry is heterogeneous, combining metal products and electrical and non~electrical machinery and apoaratus, as well as transpo~t equipment (except passenger cars). Special emphasis is given here to non-electrical machinery and equipment (NEM), an area of the capital goods market that is well-represented by both public and private sector firms. 5.03 The capital goods·industry has played a crucial role in India's post-war industrialization. Shaped by a strategy focused on the develop- ment of heavy industry through import substitution and strongly supported by government incentives, the sector has grown and become quite diversi- fied, compared to that in other developing countries. Indian policy- makers have perceived of this industry as an important tool in fulfilling the goal of industrial and technological self-reliance.2/ One of the key factors responsible for the development of the capital goods sector is that unlike other industries where protection from imports combined with regula- tory policies limited domestic competition, it has been subject to few controls in terms of capacity creation, product mix, level of output, prices and distribution. 5.04 Development of the capital goods sector has proceeded at a rapid pace in comparison with other sectors, and production capabilities have been established in most of its segments. The industry is now at an important juncture. The slowdown in demand, particularly because of the fall in investment for large public sector projects, indicates that the industry will have to expand domestic sales and marketing efforts. Although import-substituting policies permitted development of some indigenous capabilities in capital goods, absence of competition meant !/ For a more detailed discussion of the capital goods sector, in particular its non-electric machinery segment, see World Bank Report No. 5095-IN of August 1984. '!:_./ In the early 1950s, India depended on imports for about 70 percent for its machinery and equipment requirements; in contrast, by 1981-82, nearly 87 percent of the overall domestic demand for capital goods was met through indigenous production. - 58 - that in the past, producers were not given sufficiently strong incentives to absorb new technologies and engage in product development. Increased competition from foreign producers, in both import and export markets, should lead to an improvement in the technological base. B. REGULATORY POLICIES 5.05 The Government of India (GOI) has actively sponsored private sector participation in the capital goods industries. Import substitution policies have raised profitability, thus offering relatively large incen- tives for investment. At the same time, fewer direct controls on capacity, production and output prices have created a degree of domestic competition absent in many other branches. 1. Licensing 5.06 Licensing procedures have not been a major constraint on entry into and expansion of the sector. Industrial licenses for the engineering industries have been issued at a higher rate than in other branches. The sector's share r0~e from close to 50 percent of all licenses issued in the early 1970s to over 65 percent in 1983/84 (Table 1). Fewer licenses were issued in non-electrical machinery (NEM) in 1983/84 than in 1972/73, 'but the reason was probably the ongoing process of gradual delicensing of the subsector rather than a decrease in capacity creation. 5.07 In 1974, industrial machinery manufacturers were allowed to diversify their production within their overall licensed capacities for machinery production with the approval of the Department of Heavy Indus- try. Thus, the rigidity in product mix attributable to capacity licensing was relaxed, and manufacturers were given wide scope to adjust their output to changing demand. In addition, entry into the sector by relatively small firms has been delicensed since late 1975, when industrial machinery was among 21 industries exempted from the licensing provisions of the IDRA Act (1951). However, the exemption was not applied to large industrial houses and MRTP companies, so that entry and expansion by all enterprises of significant size have been subject to licensing constraints. 5.08 Although private sector firms have been encouraged to enter and expand in the indu~try, India has concurrently followed a policy of exten- sive investments in the public sector to build up the capital goods indus- try. 3/ Early planners laid primary emphasis on ensuring the availability of investment goods at reasonably low prices to help the growth of "key sectors" of the economy, and consequently public sector capital goods 11· See K.R. Paramesvar, "The Capital Goods Industry in the Public Sector," prepared for the Indo-French Seminar on Role and Management of the Public Sector, January 1986, processed. 59 - Table 1: ISSUED INllJSTRIAL LICENSES. FOR ENGINEERIM; GOODS Sl. No. Indwitry 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1. Metallurgical 62 68 285 201 79 55 24 47 56 66 61 406 388 2 Boilers and steam- • generating plants 1 2 1 3 1 Prime 1110vers (other than elec- ttic generators) 3 1 6 3· 2 1 3 . 2 4 5 4. Electrical equipment 65 60 94 92 79 90 58 52 70 55 75 101 102 5. Telecommmtcations 15 18 8 6 2 4 3 4 2 6 5 18 31 6. Transportation 43 37 31 32 15 17 13 20 11 15 13 21 38 1. Industrial machinery 42 40 55 69 53 46 29 35 20 21 31 43 31 8. Machine tools 12 16 10 8 14 12 9 5 8 10 3 12 3 9. Agricultural machinery 12 13 12 9 1 3 1 1 10 Earth"1!Dving machinery 2 5 5 3 1 2 3 1 2 11 Misc. mechanical and engineeting industries 17 21 14 11 31 11 7 11 14 8 2 12 9 12 Commercial, office and houseb>ld equipment 6 9 1 2 4 5 5 4 5 8 Medic.al and surgical 1 4 3 4. 2 1 14 Industrial ins tl"Ulllents 2 3 8 5 9 6 3 11 7 5 15 9 is. Scientific instruments 1 1 . - 3 1 1 2 16. Mathematical, surveying and drawing instruments Total eqgineeri!IR industries 281 284 531 453 294 253. 154 168 213 200 209 645 623 Total all industries 563 596 1,099 1,027 662 518 348 365 475 475 432 1,075 905 ShAL~ of licenses issued to engin- eering industries 49.9 47.7 48.3 44.1 44.4 48.8 44.3 46.0 44.8 47.0 48.4 60.0 88.8 ~: Secretariat for Industrial Approvals, Ministry of Industry, Gove~nt of India; and Eastern Economist. - 60 - units (BHEL, HEC, HMT and MAMC) were set up to cater to the infrastructure sector and major industries such as fertilizers, steel· and mining.~/ In addition, a number of failing enterprises have been taken over by the public sector in recent years. 2. Pricing Policies 5.09 The capital goods sector has .not been subject to direct price controls, although price controls on some user industries, particularly cement, sugar and paper, have affected the subsector indirectly. In these industries, controls have led to low profitability and reduced investment and consequently have depressed the demand for equipment and machinery. Government policies with regard to the pricing strategies followed by public sector enterprises have alsc been important. In principle, the prices of public sector enterprises are set to cover not only the cost of production but also to generate reasonable profits at a satisfactory level of capacity utilization. This approach is reinforced, in practice, by the fact that public sector enterprises receive a 15 percent price preference for capital goods procured by the government. This pricing policy has caused some competitive friction between public and private sector firms. 3. Trade Policie~ and Incentives s.10 The level of import protection for capital goods is below the average for industrial goods, but protection has become increasingly differentiated. The tariff on machinery and equipment for fertilizer plants is zero and for power plant equipment 25 percent. Project-related investment goods were charged a rate of 45 percent until early 1986, when the tariff was raised to 55 percent as a result of concerns of the domestic capital goods industry. Other imported machinery and equipment competing with locally made products are subject to several non-tariff barriers. 5.11 Despite the tariff and non-tariff barriers, many goods do enter the country in the context of new projects, and the domestic industry has felt the pressure of international competition to an increasing extent. Bilateral aid contracts often involve the import of equipment that is available domestically but cannot meet the specifications demanded by project authorities in the context of updated technologies. This situation is perceived by some capital goods suppliers as a preference for imported capita1. goods. In addition, recent developments :in world markets have put pressure on capital goods prices, and dumping of international competitors has mounted. India does not currently possess the institutional framework to neutralize dumping practices.within the context of a more liberalized import regime. 5.12 Imported inputs are subject to higher import duties than final capital goods products. Duties on inputs range from 50 percent to 90 percent, and some go as high as 300 percent. Among the crucial steel inputs, 9 of the 15 types are either restricted or banned and require import licenses. Pig iron imports are canalized. Combined with sporadic 4/ Ibid., P• 11. 61 - shortages of domestic supplies, canalization has created delays and high costs from inventory accumulation. 5.13 The higher tariff protection on inputs creates a net disincentive effect for the sector. That effect has often been greater than what is implied by the tariff structure, because domestic competition has kept the sector's output prices at about the level of international prices, while intermediate inputs have been priced up to the protection provided by customs duties. The net disincentive to producers can be measured by effective protection coefficients (EPC), estimated by comparing dom~stic and international prices. The EPCs calculated for the various capital goods segments for the 1980-82 period were all below one, indicative of negative incentive effect (see Table 2). 4. Technology Policy S.14 Several government committees have made recommendations to ioprove technology policies. The Hussain Committee has suggested putting technology imports under OGL in case foreign equity participation is not necessary, and allowing selective imports (to be regulated by FIB) where foreign equity participation is unavoidable. The Committee has also suggested combining these measures with a commitment by firms to increase internally generated R&D for indigenous technological development. Simi- larly, the Ramanathan Committee has proposed greater cooperation between users and manufacturers by drawing up perspective plans for technology development. The first step is the creation of a venture capital fund, which will provide equity capital to pilot plants attempting commercial application of indigenously developed technology. S.15 Although technical cooperation agreements in the capital goods sector have increased, there are still numerous obstacles. Domestic manu- facturers claim that it is easier to import products thar. the corresponding product or process technology. In addition, present procedures for approv- al of foreign tie-ups are complex and time-consuming, particularly because of the income tax rates and RBI approval of remittances of lump sum and royalty payments. Generally, technology imports are studied on an indivi- dual basis, and a high degree of discretion is left to the government agency to decide whether to approve the technology collaboration and terms of the contract. Not only are these procedures not very expedient, but they bring a substantial degree of uncertainty to investment decision- making. - 62 - Table 2: EPCs AND DRCs OF SELECTED MACHINERY AND EQUIPMENT MANUFACTURED_ BY THE SUBSECTOR, 1980-82 EPC on Value Added DRC Cement-Making Machinery Cement mill .34-.41 .42-.45 Crusher .99 .73 Complete plant (1,000 TPD) a/ .49-.55 .34-.38 Complete mini-plant (200 T!'D) !_/ .45-.57 .39-.44 Pulp and Paper Manufacturing Machinery Rotary spherical digester .60 .57 Bleach wash filter .48 .60 Horizontal pope reel .62 .60 Chip washer .54 .53 Sugar Manufacturing Machinery Clarifier .58-.78 .59-.81 Centrifuge .40 .61 4 milling tandems .40 .45 Complete plant (1,000 TCD) a/ .40-.65 .29-.56 Chemical Industry Machinery Vacuum concentrator .39 2.12 Stainless steel splitting column .57 .60 Aluminium absorption tower .16 2.08 Stainless steel absorption tower .72 .91 Separator vessel .81 .91 Horizontal ammonia let-down vessel .30 2 .14 Boilers 3 x 210 MW power boilers .59 .84 Source: Bank staff estimates., 1984 .. a/ Excluding purchased components. C. STRUCTURE OF THE INDUS'!'RY 1. Firm Size and Distribution 5.16 The capital goods sector is characterized 9Y a heterogeneous firm size distribution, with the presence of a relatively large number of small- and medium-size enterprises. Judg:tng from the number of parti~ipants in each of its segments, the capital goods industry would seem to reflect a - 63 - competitive structure, with most markets supplied by between 15 and 100 firms. In the early 1980s, there were over 160 firms producing textile machinery, 90 firms producing chemical equipment, over 25 producing sugar and paper machinery and 13 producing cement equipment (Table 3) • • 5.17 Firm growth has nonetheless been constrained by the size of the domes~ic market and by the fact that negative effective protection has discour~ged the development of export markets in the capital good8 sector. A corollary of slow firm growth is high firm conce~tration. Although regu- latory policies are more liberal in this sector than elsewher~, licensing ~'.lolicies limiting entry at&.J expansion by large units have also contributed to high concentration. In several subsectors, the four largest producers ·coatrol over 60 percent of total output. For printing machinery, the mar~~t share of the four largest firms is 84 percent; for boilers, 88 percent; for cranes, 77 percent; and for cement equipment, 73 percent. In addition, a few major producers of non-electrical machinery dominate production and sales in about a dozen subsectors of the capital goods industry, where 15 large multi-product companies account for over 90 percent of the output of most non-electrical machinery (Table 3). 2. Subcontracting 5.18 The large number of smaller units present in many capital goods segments has led to an excessive degree of fragmentation. Larger units could: subcontract to smaller firms for parts and components, but this type of vectical disintegration has been limited. In industrtalized countries, subcontracting is essential to exploit horizontal economies of scope, which are economically more advantageous than vertical linkages. In South Korea a~~ Japan, most small firms are parts suppliers rather than independent ma~hine-builders, and the share of all bought-out items (sub-supplied and sabcontracted) in a complete unit or plant accounts for as much as 60 percent of the total value of equipment and machinery. 5.19 Fr'Jtll this perspective, the 15 percent share in some Indian capital goods plants is too low. In India, the share of subcontracted pa~ts in the value of final sales amounts in many plants to only about 15 percent, and small firms are in many cases independent machine-builders. Note that subcontracting has not been sufficiently developed despite the reservation of 232 engineering products for production by small-scale firms. The subcontracting network in India's capital goods sector has been ~hther limited in the past primarily because it is unreliable, costly and irregular in delivery. In addition, the limited economic size of the . ~omestic market is a disincentive for vertical disintegration. As a conse- quence, the industry must make peripheral components in small batches, and the scope for specializat1on gets diluted, while domestic production becomes costly. - 64 - Table 3: STRUCTURE OF OWNERSHIP AND MARKET SHARE IN NON-ELECTRIC MACHINERY SECTOR, 1983-84 (1) (2) (3) (4) Ownership Sales During 1983-84 Subsector (Number of Companies) Value Percentage (Rs mil.) Share 1. Earth-Moving Public Sector 2,706.6 66.8 Equipment (2 cos.) 10 Companies 4,049.8 100.00 2. Cement Machinery Private MRTP 696.1 77 .3 (7 cos.) Private Sector 203.9 22.7 (7 other units) 14 Companies 900.0 100.0 3. Chemical & Pharma- Public Se~tor 306.1 19.1 ceutical Machinery (2 cos.) Private FERA 66.5 3.9 (2 cos.) Total for Subsector 1,600.0 100.0 4. Water Treatment Private FERA 136.4 93.2 Plant (2 cos.) Private MRTP s.o 6.8 ( 3 cos.) Total 5 Cos. 141.4 100.0 s. Dairy Machinery Public Sector 9.9 9.2 (1 co.) Private FERA 64.9 60.4 (1 co.) Private MRTP 27.6 30.4 (4 cos.) Total for 6 Cos. 107.4 100.0 - 65 - . .. Table 3 (Continued) • (1) (2) (3) (4) Ownership Sales During 1983-84 Subsector (Number of Companies) Value Percentage (Rs mil.) Share 6. Drilling Equipment Private FERA 13.3 100.0 (2 cos.) 7. Drilling Equipment Public Sector 494.9 92.9 (1 co.) Private FERA 15.6 2.6 (2·COSe) Private MR.TP 22.1 4.5 (3 cos.) Total for 6 Cos. 532.6 100.0 8. Mining Machinery Public Sector 142.9 28.6 (2 cos.) Private FERA 78.0 27.0 (2 cos.) Private METP 248.2 18.2 (7 COb) Private 130.9 26.2 (Other cos.). Total for Subsector soo.o 100.0 9. Packaging Machinery Private FERA 1.0 11.8 (3 cos.) Private MRTP 9.7 88.2 • ( 3 cos.) Total for 6 Cos. 11.0 100.0 10. Printing Machinery Public Sector 49.3 35.8 (1 co.) Private MRTP 88.5 64.2 (4 cos.) Total for 5 Cos. 137.8 100.0 - 66 - Table 3 (Continued) (1) (2) (3) (4) Ownership Sales During 1983-84 • Subsector (Number of Companies) Value Percentage (Rs mil.) Share 11. Pulp and Paper- Public Sector 67.8 34.9 Making Machinery (2 cos.) Private FERA 0.1 o.4 (1 co.) Private MRT 22.6 10.3 (2 cos.) Private 108.9 44.6 (Other cos.) Total for the Subsector 200.0 100.00 12. Textile Machinery Public Sector 26.7 0.9 (1 co.) Private FERA 11.0 0.4 (1 co.) Private MRTP 2,267.5 81.0 Private 494.8 17.7 (Other cos.) Total for Subsector 2,800.00 100.0 13. Rubber Machinery Private FERA 25.9 8.6 (1 co.) Private MRTP 114.7 38.3 ( 3 cos.) Private 159.4 53.1 (Other cos.) Total for Subsector 300.00 100.0 - 67 - Table ~ (Continued) (1) (2) (3) (4) Ownership Sales During 1983-84 Subsector (Number of Companies) Value Percentage (Rs mil.) Share 14. Road Construction Private FERA 0.9 1.0 Machinery (1 co.) Private MR.TP 91.5 99.0 (4 cos.) Total for Subsector 92.4 100.0 15. Boilers Public Sector 3,603.0 75.1 (1 co.) Private MR.TP 642.9 13.4 (4 cos.) Private 554.1 11.5 (Other cos.) 16. Tea Machinery Public Sector 5.9 100.0 (1 co.) 17. Weighing Machinery Private FERA 159.9 100.0 (1 co.) 18. Industrial Filtra- Private FERA 91.8 92.4 tion Equipment ( 3 cos.) Total for Subsector 9.4 100.0 Source: Economic Intelligence Service, Markets & Market Shares, Bombay, 1986 • • - 68 - 3. Ownership Patterns 5.20 The ownership pattern in the capital goods sector is character- ized by private firms prevailing in such branches as cement, sugar and most other non-electrical machinery, while public sector firms dominate in heavy engineering, machine tools and boiler manufacturing. For the three engi- neering groups, i.e., non-electrical machinery, electrical machinery and transport equipment, the public sector accounts on average for 34 percent of the total number of employees, 29 percent of total investment and 23 percent of the total value of gross output. The heavy involvement of the public sector in this industry is reflected in the fact that out of 53 central government public sector ~irms, 46 were in the capital goods indus- try in 1983-84. The GOI's promotion of public enterprises in the capital goods sector has been, in this sense, highly effective. 5.21 FERA companies have emerged as dominant undertakings in selected product lines, which are usually technology-intensive and patent-protected (usually for 10-15 years), so that there has been a high degree of market concentration. They account for 93.2 percent of the market in water treat- ment plants, 60.4 percent in dairy machinery, 100 percent in food process- ing machinery, 100 percent in weighing machinery, 92.4 percent in indus- trial filtration equipment and 27 percent in mining machinery (Table 3). 5.22 Foreign investment in the capital goods sector has shifted over time from foreign subsidiaries to joint ventures with minority equity holdings by foreigners, and further to license agreements for technology transfer without equity investment. The pattern o~ foreign companies' operations in the Indian capital goods production has shifted from equity investment and managerial control to the sale of technology and marketing services as a means of directly controlling returns on corporate assets. For the 2,404 collaboration agreements approved during the period of 1981 to 1984 in Indian industries, foreign equity participation was found in only 450, or 18.7 percent, of the cases. In addition, the import of drawings and designs for particular products and processes without regular licensing agreements has been increasing in recent years. D. IMPACT OF POLICIES ON BEHAVIOR S.23 Although the capital goods industry was protected from inter- national competition until the late 1970s, liberal policies have ensured a more intense level of competition than in many other sectors of the econo- ~y. Even for sectors dominated by few suppliers, the possibility that axcessive profits might stimulate entry may·have encouraged a more competi- tive behavior by existing firms. The possibility or the threat of entry was made credible by more relaxed licensing regulations than in other sectors. 5.24 One indication of the level of competitiveness of the industry are estimates of domestic resource costs (DRCs). DRCs estimated for a World Bank study covering the 1980-82 period indicated .that long-estab- lished manufacturers of cement machinery, sugar machinery, pulp and paper - 69 - manufacturing machinery and boilers were efficient with ranges of DRCs significantly below 1.0 (~ee Table 2). The study concluded that the prices of Indian capital goods during this period were internationally competi- tive. The disadvantages of high input costs (resulting from both high tariffs on imported raw materials and monopsonistic markets in domestic inputs like steel) were offset by low unit labor costs. 5.25 Since the 1980-82 period, however, India's favorable position appears to have deteriorated. In particular, the sector has not been as successful as desired in developing high quality products and incorporating advanced technology. Although only 13 percent of domestic requirements for capital goods are imported, it covers vital high technology equipment for fertilizer plants, oil field equipment, steel plant equipment, etc. Some of the weaknesses of the technological base include a limited ability to absorb and adapt contemporary technologies, improve or develop new products. 5.26 The insufficient technological progress in the capital goods sector can be explained as the outcome of a number of interrelated factors, including public sector procurement policies, protection from international competition, and insufficient pressure imposed by do~estic user industries on manufacturers in the area of product development. Marketing as a management function has not received much attention over the last 25 years. Large purchases by the public sector, combined with a captive domestic market and low export volumes, have reduced the importance of marketing efforts. Since domestic user industries were developing concur- rently with the capital goods sector, the demand for high technology items was not as great as in the highly developed international market. In industries such as cement and paper, where price controls in the late '60s and '70s reduced profitability, purchasers of capital goods ·equipment were probably highly price-elastic and emphasized low prices instead of high quality, high technology products. 5.27 In the area of public sector procurement, the preferential 15 percent pricing granted to public sector firms has constrained competition between public and private firms. As in other sectors, several public firms have exhibited behavior that can be characterized more appropriately as bureaucratic than entrepreneurial.Sf HEC (Heavy Engineering Corpora- tion), for example, has suffered from-high turnover of its chief executive officers, exit by engineers and key specialized personnel, institutional barriers to adaptation and exit, an inability to shift its emphasis to industries other than steel, and continued dependence on the GOI's influence and resources. 5.28 The government's fiscal constraints have often led to sudden and drastic changes in the priorities and time schedule of public investment. Such changes affect. the producers of capital goods, which sell primarily custom-made equipment to state enterprises. Since the government is a major buyer of capital goods, fluctuations in its demand have had a nega- tive effect on capacity utilization and consequently on profitability. S/ · Ibid. , P• 22. • - 70 - This situation has an adverse effect on (i) the entry of new firms, and on (ii) technological d~velopment. 5.29. The relatively weak base in design erlineering, product develop- ment and production engineering is also a cons~quence of the high protec- tion against international producers. In fact, recent measures toward import liberalization have led to increased upgrading of technology in many sectors. Indian firms, both in public and private sectors, are attempting to improve their competitiveness and leadership in complex capital goods through foreign collaboration (Table 4). In response, foreign-controlled firms are striving to expand their market share by diversifying into a higher-priced and more sophisticated range of products. Table 4: NUMBER OF FOREIGN COLLABORATIONS IN SELECTED CAPITAL GOODS ITEMS Items 1981 1982 1983 1984 1985 Jan.-Mar. 1. Boilers 7 2 4 1 2. Prime Movers (Other ·than Electrical Generators) 2 4 2 6 1 3. Transportation 26 28 39 63 27 4. Industrial Machinery 130 107 115 138 25 s. Machine Tools 22 29 44 34 12 6. Earth-Moving Machinery 11 9 8 4 3 7. Agricultural Machinery 3 2 2 Source: E. IMPACT OF POLICY ON PERFORMANCE 1. Growth in Output 5.30 Since the early 1950s, when industry relied on imports for 70 percent of its machinery and equipment requirements, the growth and diver- sification of the capital goods sector have been substantial. Table 5 shows the index of production of different segments of the capital goods sector and the manufacturing sector. Production growth for non-electrical machinery far exceeded the growth for electrical machinery and transpo~t equipment, and for the manufacturing sector as a whole. Growth rates were particularly high betwee.n 1955 and 1965, when capital goods production - 71 - Table 5: INDIA, INDEX OF INDUSTRIAL PRODUCTION, CAPITAL OOODS (Base 1970 = 100) Weight (%) 1951 1955 1960 1965 1970 1975 1980 Industcy Non-electrical Machinery 5.55 a/ 5.4 8.7 24.5 78.6 100.0 159.6 218.3 Electrical Machinery 5.32 7.1 13.2 21.1 56.4 100.0 120.3 168.1 Transport Machinery 7.39 14.7 74.3 74.9 153.8 100.0 106.3 130.4 Total Manufacturing 81.08 b/ 30.5 41.3 55.9 86.1 100.0 116.0 145.8 Source: World Bank, "Economic Situation and Prospects of India," April 17, 1982. a/ rus industry group includes machinery and equipnent for process industries as well as diesel engines, machine tools, ball bearings, office machines, consumer durables, etc. Industrial machinery and equipment for process industries have a weight of about 2 percent. b/ The other industry groups accounting for 18.92 percent of the ~ight are mining and quarrying (9.69 percent) and electricity generation (9.23 percent). increased by annual rates in excess of 20 percent. With a significantly enlarged base, that pace of expansion co·Jld not be sustained in the second decade, however, when it averaged about , percent annually. 5.31 The growth path of the capital goods sector is characterized by a typical cyclical pattern, which can be explained in terms of fluctuations in the rate of fixed capital formation in general and public investment in parti~ular. In addition, regulatory policies have played a fairly impor- tant role. The decline in output growth of machinery and equipment in the decade 1965-75, for instance, can be traced to the restrictions placed by the government on some of the process industries by way of licensing and price and distribution controls, which not only inhibited the creation of new capacity but also served to slow modernization and plant rehabili- tation. On the other hand, the stable growth in public investment, with I clearly defined priorities produced dividends in the period 1955--65 in terms of a higher growth rate or capital goods production. 5.32 Recently, the expansion of output has been disappointing, with the engineering industry growing at 4.7 percent rather than the 10 percent targeted in the 6th Plan. The growth of NEM has been somewhat faster--at 5.3 percent--but still has not been much above the total sector growth of 4.4 percent. However, within that group there have been several industries with outstanding records, registering output gains well ahead _of 12 percent; they include boilers and cement machinery, as well as chemical and pharmaceutical machinery. 72 - ~I lllllTMUJ> PIQJUCl'llll AllD ~ACI'l'I UTlt.11.ATIOI IN IOf-lllJICl'tlCAI. ~ACI'l'I WlTAL 00008 SllS£Cl'ClR, 1979-80 TO 1986'-85 11111.t 1979-80 81>-81 81-82 82-83 83-84 84"-85 %Growth rate (l) (2) (3) (4) m (6) (7) (8) (9) l. Boiler b• ailliOG c 2~ 3,240 4,000 S,000 4,700 4,500 12.l p 1,594 3,064 3,467 4,359 4,624 5,376 15.7 u 102.l ss.1 86.7 87.2 102.6 119.S 2. 1>1-1 Bnglne th. Noe. c 311 317 317 313 313 295 (•I) 6.9 p 144.6 173.9 174.S 1'6.5 1'6.l 170.9 3.4 u 45.6 54.9 55.0 so.o so.o 57.9 3. Bart~.. Equipqent Noe. c 1,540 3,060 3,060 2,100 2,300 2,995 14.2 p l,285 996 1,792 2,212 1,1114 1,627 4.8 u 83.4 48.3 86.9 105.3 78.9 54.3 4. Cranes th. Noe. c 64 64 64 65 65 115 0.3 p 16.7 19.6 22.1 22.1 20.1 20.1 4.l u 26.1 30.6 35.5 34.9 35.4 31.8 s. r - r Drl"'"' Puaips th. b . c 400 420 420 420 420 749 l'l.4 p 346.8 430.6 373.0 460.6 491.6 482.8 6.9 u 86.7 102.S 88.8 109.7 111.0 64.s 6. Ur/gas Ccapreaaora Th· 11oe. c 11.2 lS.3 15.3 15.3 IS.3 17.4 5.6 p 13.3 16.4 16.6 29.4 2'J,2 30.a U!.3 u 91.1 JOtl.8 108.S 138.0 159.7 111.0 7. Bell end Roller Beeriage Ktllion Noe. c 35 4S 45 50 60 59,5 11.2 p 28.7 34.27 35,4 37.J6 40,77 47,93 11.1 u 80.8 76.2 78.7 74,4 68.0 ao.1 8. Agriculture Trectora Tb. Moe. c 70 n 100 90 90 91 s.s p 62.4 11.0 84.1 62.9 75.8 85.o 6.4 u 89.1 92.2 84.1 10.0 84.2 93.4 9. load Rollera th. Noe. c 1,800 1,800 1,800 1,800 3,060 11.2 p 823 948 1,065 926 440 309 c->62.s u 45.7 52.7 59.2 ~1.4 28.11 10.1 10. !flchine Tools 118. aillion , c 1,900 1,652 2,100 2,900 2,900 1,962 2,499 .2,699 2,697 3,400 3,407 3,028 12.4 12.9 u 86.9 93.4 86.2 93.1 llO.t 88.9 11. Air CondiUomra th. Noe. c 44.78 44.78 44.78 48.ll 48.8 49.3 2.0 p 26.52 28.19 32.23 27,23 28.63 29.11 1.9 u S9.2 62.9 71,9 ss.s sa.o 59.0 12. Sugar 'Ychinary b. llilliOll c 518 518 520 575 600 748 7 .6 p 316 242 263 415 498 437 u 61.0 46.7 S0.6 12.2 1!3.0 58.4 '"' 13 •. Hl.ning Kach1nary Ra. ail Uon c 420 450 450 500 500 554 5.6 p 303 373 437 Sl9 480 479 9,7 u 12.1 1!2.9 97.1 103.3 96.0 86.5 14. !fetelllltJllnal 'lachtnery 118. "11Uon c 800 800 800 800 800 1,226 8.9 p 422 SSb "2 656 622 630 3.4 u S2.3 69.S 80.3 1!2.0 85.8 Sl.9 . 1S. o-J.cal .aad Phenoac ..... ticel Machinery c p 118. odlliOll l,120 1,120 1,350 1,500 1,500 2,UOO 12.3 796 J,066 1,228 l,429 1,560 1,550 14.2 u 11.2 9S.2 90.9 95.) 102.8 77.5 16. C-nt l!achinary Ka. odl11on c 423 450 S60 680 750 800 tl.6 p 253 336 437 453 443 538 16.l u 59.8 74,7 78.0 66.1 59.7 67.2 ~: A.IEt, !!ES1neer1!!ll Production Profile, July 1985. Motee: C • Cepacity tnatalled. P • Production. U • C..pecity utiUzatlon in percentage. - 73 - 5.33 The higher rates of capacity utilization in the period 1979-80 to 1984-85 in subsectors such as boilers, air compressors, cement machinery, agriculture tractor, mining machinery, and chemical and pharmaceutical machinery owe much to investment growth in the user industries during the period (Table 6). Generally, however, capacity utilization has closely followed changes in output. Capacity utilization, 75 percent during 1961-66, went down to 60 percent in 1967-76 and has not changed signifi- cantly since. No major changes in capacity utilization have occurred until recently. 2. International Competitiveness 5.34 The share of capital goods exports in total exports reached 13.0 percent in 1980/81 and 13.4 percent in 1981/82, while its amount increased from Rs 51.6 million in 1956/57 to Rs 10,469.9 million in 1981/82. Howe- ver, considering that Indian engineering exports were only O.l percent of world engineering exports in 1979, India does not seem fully to exploit its capital goods export potential. 5.35 The subsector's exports have become significant only in the last decade. For boiler machinery, India is the most important exporter among the newly industrializing countries (NICs), accounting for 2 percent of world exports. In addition, exports of machinery for the cement and paper industries have grown rapidly in the early 1980s. However, there is a growing concern about losing that competitive edge, particularly since exports declined for engineering goods during the last year. S.36 India's lack of an export orientation is the result of several factors: insufficient efforts in export marketing, delays in receiving drawback duties, 'insufficient financing for project exports and inadequate infrastructure and high input costs. 5.37 Since the prices of Indian capital goods are below international prices, protection for the industry may be redundant.6/ Currently, many capital goods imports are subject to a 55 percent tariff rate and other non-tariff barriers such as the "indigenous angle clearance" procedure. What are the justifications for protection when domestic products can compete with imported products? One may be that even though prices are competitive, quality and reliability are inferior, and thus users tend to put a high premium on imported machinery that outweighs the price advan- tages. Another reason may be the strong pressure exerted by long- established industries. 3. Profitability of Capital Goods 5.38 The rates of return in the capital goods industries show a cycli- cal pattern, similar to the fluctuations in output growth, with a marked decline in the late sixties and strong recovery thereafter. With the 6/ Note that the measurements of price competitiveness were taken in 1q82/83. tit seems that the subsector's competitiveness. has deferfor.a ed since. - 74 - Table 7: FINANCIAL PERFORMANCE OF MEDIUM- AND LARGE-SIZE PRIVATE CAPITAL GOODS FIRMS (in %) Gross Prof its Gross Prof its Net Profits Year Net Sales Total Capital Net Worth 1960-66 12.8 11.4 12.a 1966-70 10.6 8.1 7.3 1971-75 11.8 11.8 12.2 1976-80 12.8 13.S 14.9 1981-83 12.1 13.0 17.7 Source: RBI, Financial Statistics of Joint Stock Companies in India; and RBI, Bulletin, various issues. issuing of fewer licenses during the mid-'70s and better capacity utilization, profitability increased. In the early 1980s, profit margins stabilized, although profit rates rose further (see'Table 7) •. The recent compression in profitability suggested in discussions with industry officials seem to be linked to changes in demand and increased import competition rather than to licensing policies and procedures. S.39 The balance sheets and income statements of some major public and private companies in the early 1980s show th~t on average profitability has been reasonable. Several problem companies i11 the public sector have lowered the average return on capital employed for that group to 4-5 per- centage points less than that of private companies (9.1 vs. 13.6 percent between 1980/82, and 8.5 vs. 12.7 percent between 1983/84). In part, the reason is that the public sector took over a niimber of loss-making private enterprises that have not been returned to profitability and are in urgent need of restructuring. In addition, it reflects differences in the style of management, some of which are not easily changed because of the more stringent constraints the public manager faces in the social sphere of his firm. However, there still seems to be ample room for improvement even within the more narrowly defined constraints, and the government is now ' considering steps to improve management decision-making at the finn level. F. RECENT POLICY CHANGES 5.40 Recent measures have further decreased the extent of domestic regulation. Within the 25 categories that were delice~sed in March 1985, seven groups belong to the engineering goods subsector. The industrial machin~ry group, as well as equipment for the rtibber, printing, footwear, - 15 - meat and poultry industries, can now be produced without prior licensing, provided the MRTP and FERA Acts, as well as the small-scale industry reservation restrictions, are satisfied. Some relaxation was also extended to MRTP and FERA companies in December 1985, when delicensing was extended for 22 out of the 27 industry groups, with those companies being exempted from sections 21 and 22 of the MRTP Act, provided that their plants were located in backward areas, as defined by the Government. The engineering industries were predominant in these industrial groups, including machinery for chemical industry, machine tools, and industrial machinery for drilling and production of natural oil and gas. 5.41 Capital goods (particularly machinery and equipment) were even more prevalent in the list of industries qualifying for broadbanding, which allows changes in product mix without prior licensing. Broad-banding encompasses agricultural as well as metallurgical machinery and equipment for the chemical, pharmaceuticals, petrochemical and fertilizer indus- tries. Further changes for MRTP and FERA companies introduced in late 1985 allow these compani~s to set up plants and equipment in all major capital goods branches, provided the products are not reserved for the SM! sector. If the major share of production is exported, all industries qualify for investment. Finally, capacity re-endorsement was facilitated for all those companies that had achieved at least 80 percent or more of licensed capacity during any of the five years preceding 1985. 5.42 These liberalization measures are expected to stimulate competi- tion by expanding a firm's choice of product mix, easing entry of MRTP/FERA companies and encouraging small-scale units t~ grow into medium-scale firms. While it is too early to assess definitively the response of the capital goods sector at this point, nearly 200 registrations had taken place in the 25 industry groups delicensed between March and December 1985. Such capital goods as industrial and metallurgical machinery were prominent among them, representing about 10 percent of the proposed invest- ment. Finally, Table 8 suggests, from a number of indicators, a positive investment climate that in all probability will affect the capital goods industry positively in the near future. G. CONCLUSIONS 5.43 There does not seem to be a uniform, one-way relationship between industrial licensing and other control measures and the competitive structure and performance of the capital goods sector in general and the NEM branches in particular. Greater freedom of entry and some import competition have kept prices more competitive and quality somewhat above standards in other industries. While it seems that some of the branches could have been further exposed to international competition in the early 1980s, recent developments have led to cutbacks in exports and slower than expected gcowth of domestic sales. S.44 Two major sectoral issues need to be addressed. First, the prices of most important inputs from domestic industry are still far above - 76 - international prices, and they have increased even further in recent years and lowered the capital goods competitiveness at a time when world industry is engaged in t:ut-throat r:ompetition. Second, technology in many branches of the Indian capital goods industry has no·c kept pace with the rapid developments abro~d. !t is obvious that f.n both cases, government policies will have a decisive role to play in the future in terms of disciplining the mostly state-owned steel industry and stimulating both the adoption of and indigenous attet:.pts at nt1:w tec..hn·~logy through an improved technological infrastructure, better incentives and strearolined pro~edures. Table 8: INDICA10RS OF INVESTMENT RESPONSE 1982 1983 1984 1985 1. Letters of Intent Granted 1,044 1,055 1,064 1,457 2. Foreign Collaborations Approved 590 673 752 1,024 3. Schemes Registered with DGTD, 1,538 2,024 1,854 1,416 of Which Schemes in Backward Areas 704 1,155 1,106 788 4. Cases Approved by Capit~l Goods Imports (Main Committee) (Rs crores) 502 607 713 747 s. Consent Giv~n by the Controller of Capital Issues:a/ i) Number 472 459 712 713 ii) Amount (Rs crores) 893 1,023 2,00J 2,104 6. Tern Loans Sanctioned by Financial Instttutions a/ (Rs crores) - 3, 184 4,044 5,624 4,510 7. Term Loans Disbursed by Financial Institutions a/ (Rs crores) 2,358 2,893 3,408 3, 116 Source: GOI, Economic Survev 1985-·86. a/ These data relate to Aprll-March of each consecutive year, except for 1985, which relate to the period AprU-December 1985. 1, \, ' " . - 77 J "' INDIA - INDUSTRIAL REGULATORY POLICY STUDY VI. Subsector Review--Machine Tools '\ 6.01 This paper looks first at the regulatory policies affecting the \ . machine tools industry, including recent steps toward liberalization, such as the introduction of broad-banding licensing and the reduction of import controls. The impact of these regulations on industrial structure, behav- I~ ior and performance is then analyzed. \ \. A. BACKGROUND 6.02 Presently, 137 firms produce machine tools in India, with an installed capacity worth US$283 million. Total employment in the sector is about 125,000. India ranks eighteenth in terms of production among 35 machine-building countries and sixteenth among 25 leading machine tool- consuming countries. Indian industry produces general purpose machine tools, heavy-duty machine tools such as horizontal and vertical boring machines, jig mills, hydraulic and pneumatic presses, universal cylindrical and other types of grinders, automatic lathes and plano millers. Numerical control (NC) and computer numerical control (CN~) machines are also produced in small numbers, with their output totaling 42 units in 1984.1/ In other countries, in contrast, NC/CNC have been replacing traditional- machine tools very rapidly since the mid 1970s. In seven 2/ OECD countries, the share of CNC lathes in total lathe production went from about 28 percent in 1975 to more than 50 percent by the beginfiing of the 1980s, while the share of CNC lathes in total investment in lathes went from 20 percent in 1975 to 50 percent in 1981.3/ CNC lathes are also being produced in newly industrializing countries (NICs), albeit more slowly. In the case of Taiwan, Province of China, for example, the share of CNC lathes in total outlays for lathes increased from about 7 percent in 1978 to around 20 percent in 1981. As a result, the market for traditional engine lathes has shrunk not only in relative but also in absolute terms. 6.03 This change in demand has been observed in India as well, ~ut, as noted, producers there have responded siowly, lagging behind in the ~ncor­ poration of CNC technology and electronic componentry int~ machine tools. 1f See "Machine Tools: Sophistication - Need of the Hour," Commerce (New Delhi}, February 1986. 2/ US, UK, Canada, Japan, France, the Federal Republic of Germany and Italy. 3/ See Staffan Jacobsson, "Entr·· into the Production· of Electronic-Based Machine Tools," June 1983, processed. - 78 - As a result, the growth rates of domestic production have been slower, and there has been a surge in imports of sophisticated machine tools since the early 19808. As compared to India's total production of CNC lathes in 1981 of 4 units, apparent demand was 33 units.4/ 6.04 In a capital goods industry such as machine tools, flexibility in production to meet changing demand is very important. Moreover, the users of machines are industrialists who in most countries can purchase the machines that produce quality commodities at viable prices. This tendency is particularly visible in open economies. India has been moving toward a more cOlilpetitive economy recently, a significant departure from its heavy protection of domestic production. This change in policies has acted as a catalyst to speed up the replacement by the user industry of machinery based on old technology. However, the machine tools industry has been slow to respond to this fast technical change. It still employs outmoded designs and equipment, an outcome of the many years of catering to a captive internal market under conditions of excess demand. As Table 1 shows, the machinery installed in the machine tool sector is fairly old~ most is 10-20 years old, with a significant proportion particularly of imported machinery more than 20 years ol~. 6.05 The difficulties facing the Indian machine tools industry since the early 1980s, such as slow growth and low utilization, can also be attributed in part to the failure to employ electronics technology: world wide, machine tools increasingly use computer-related electronics componentry. The atill incipient development of the Indian electronics industry, and the 1ifficulty to import electronic components, have constrained the growth of a modern machine tools industry. 6.06 Another feature of the machine tools industry in general is that demand is cyclical, since its output is an investment good. This charac- teristic has been particularly true for India. In general, protected ~co­ nomies such as India's have a greater tendency toward synchronous business fluctuations than do open ones, and the Indian market has been limited essentially to domestic buyers. In an open economy, assuming that all countries in the world economy do not have synchronous business cycles, sales c8n be redirected tc the countries experiencing demand growth. More important, most developing economies constitute only a small fraction of the world economy, so that the domestic industry can minimize the adverse impact of domestic business fluctuations by exporting. From this point of view, an open economy, providing better access to imported raw materials and components, and a larger share of exports in total output, can contribute significantly to more stable and sustained growth in the machine • tools industry. 6.07 The growth rate of the Indian machine tools industry has fluctuated over the past 25 years--with 3-5 years of high growth followed by 1-2 years of slow and negative growth. In the case of Japan, by comparison, the production of CNC lathes increased at a fairly stable rate: 18.1 percent a year during 1971-75, 54.4 percent a year during 1976-78 and 34.5 percent a year durifig 1979-81. 4/ Ibid. Table 1: VIN.rN:IB CF M\.QIINERY INSTALLED IN 1HE K\ClIINE 'lOOl..S INllJS'lRl 0- 5 Years 6 - 10 Years 11 - 15 :ears 16 - 20 Years Above 20 Years Total Dan. Inp. Dan. Inp. Dem. Inp. Dem. Tup. Dem. Inp. Dem. Inp. Metal-cutt~ 832 247 1,122 217 1,371 309 1,448 697 1,929 2,559 6,702 4,829 Machines (12.41) (6.13) (16.74) (S.38) (20.45) (7.66) (21.60) (17.?.9) (28.78) (63.51) (100.00) (100.00) Metal-Fonning 70 4 67 9 93 11 131 20 159 89 520 133 Machines a/ (13.46) (3.00) (12.88) (6.76) (17.88) (8.27) (25.19) (15.03) (30.57) (66.92) (100.00) (100.00) Jo~ ani Assembly 61 1 129 3 167 2 56 15 78 57 492 78 Machines b/ (12.60) (1.28) (26.22) (3.85) (33.94) (2 •.56) (11.36) (19.23) (15.85) (73.08) (100.00) (100.00) Other '&:{uiJIDellt c/ 198 61 177 76 179 62 167 100 65 220 786 528 (25.19) (11.55) (22.52) (14.39) (22.77) (11.74) (2la25) (20.64) (8.27) (41.67) (100.00} (l.00) l'C Machines 15 60 0 2 1 1 0 1 0 0 16 64 (93.75) (93.75) (0) (3~13) (6.25) (1.56) (0) (1.56) (0) (0) (100.00) (100.00} ~ Total 1,179 375 1,496 309 1,813 387 1,804 845 2,233 2,937 8,525 4,843 '° (13.83) (7.74) (17.55) (6.38) (21.27) (7.99) (21.16) (17.45) (26.19) (60.44) (100.00) (100.00) Source: Central Machi.re Tools Institute, Machine Tools Census, Bangalore, 1986. a/ TIE da~ are for mediun-siz.e ani large tmits only. b/ Includes welding, brazi.rg ani soldering, netallizirg ani rivett~ equipient ani assembly equiptent. c/ Includes plastic process~ lMChinexy ani ioundxy nnldirg machineiy. Note: Parentheses indicate tre percentage share of tre 1ll!IChinexy in each age group. Dan.: Donestic; imp.: imported. - 80 - B. REGULATORY POLICIES 6.08 Until recently, the Indian machine tools industry, like other industries, was subject to various regulations. New establishments and the expansion of existing firms had to be licensed by the government, while Monopoly and Restrictive Trade Practices Act (MRTP) and FERA companies were closely regulated for fear of possible domination by large or foreign-owned firms. Imports of machine tools were allowed selectively, while custom duties were at a level that effectively prohibited imports. Foreign equity participation was discouraged, although technology licensing was generally permitted. 6.09 This policy has changed significantly since 1980, when the government allowed automatic growth of production in five years up to 25 percent above licensed capacity. In 1982, the government expanded that liberalization measure, such that when a firm reached a certain level of capacity utilization, it was automatically allowed to. expand further. Then in 1983, the government introduced "broad-banding," a measure that allows companies to obtain licenses for the manufacture of machine tools as a category, instead of for individual items. Items reserved exclusively for the small-scale sector are, however, excluded. Thus, broad-banding gives producers the flexibility to alter their product mix to suit market demand. In the same year, the manufacture of machine tools was opened to MRTP and FERA companies, again with the exception of those items reserved for small-scale firms. In late 1985, the government delicensed the manufacture of machine tools, except for nine items that are still reserved for small-scale companies: (1) bench grinders up to 300 mm in diameter; (2) power hacksaws (mechanical 300 mm blade size); (3) turning, boring and threading "addas" (small, single or multiple operation devices); (4) machine tool accessories; (5) hydraulic jacks up to 30 tons; (6) hand threading tap holders; (7) 12 mm drilling machines up to 1,500 rpm, bench and pedestal type; (8) screw presses (manually operated); and (9) hand presses (manually operated), bench and pedestal types. 6.10 The import of machine tools also has been liberalized consider- ably. For instance, the new open general license (OGL) list for machine tools applicable under the import policy for 1985-88 includes 69 new items, which brings the total.number of items under OGL to 142. There is no list of banned machine tools and only two items are on the list of restricted capital goods. Actual users are able to import under OGL a number of machine tool components, small tools and electric items. Second-hand machine tools on OGL are allowed as long as they are not older than 9even years and do not have less than five years expected residual life. Beginning in lc~e 1985, a new scheme--the import-export passbook--was introduced. This scheme, once effective, will grant manufacturers/ exporters duty-free imports for the production of exports. The customs duty on complete machine tools has been reduced from 65 percent to 40 percent, although there has been no parallel readjustment of the custom duties on components for raachine tools. As a 'result, the machine tools industry currently is subject to negative protection. As Table 2 shows, the.duties on complete machinery are much lower than those on individual components. - 81 - 6.11 To catch up with international developments in technology, the • GO! has been generous in allowing technical collaboration with foreign partners, particularly in recent years (Table 3). However, until recently, foreign equity participation was rarely allowed. During the period 1960-80, the instances of foreign firms taking equity positions in the industry amounted to only 2 in 1966, 1 in 1968 and 2 in 1974. In contrast, during the period of 1981-85, five leading machine tool manufacturers were allowed to enter into collaborations with foreign firms for the manufacture of machining and drilling centers and NC/CNC machines. Companies with collaboration agreements are required to submit to the government a phased manufacturing program for indigenizing thr sourcing of components (currently imported) within three to five years. This program has been enforced flexibly in recent years, 'ilthough it still poses a major constraint on foreign collaboration. Table 2: IMPORT DUTY ON ITEMS FOR CNC MACHINE TOOLS Im}lort Duties Counter- Total Multi- Basic Auxiliary vailing plying Factor Excise Item Duty Duty Duty on CIF Duty (%) CNC System a/ 40 25 10 1.82 10 b/ DC Motor 100 35 10 2.59 10 b/ DC Motor Controller 60 35 10 2.15 10 b/ Ball Screw 4(J 25 10 1.82 10 b/ Measuring Systems 60 35 10 2.15 10 b/ Complete m/c Imported 40 25 10 1.82 Nil Indigenous m/c 10 Source: "The Indian Machine Tools Industry, A Farspective Plan, 1983-93," July ~~ n. !}_/ A higher rate of duty has sometimes been charged by classifying CNC systems as computers. EJ Paid on the finished machine. - ~ - Table 3: FOREIGN COLLABORATION IN THE INDIAN MACHINE TOOLS INDUSTRY, 1980-85 Number of Foreign Collaborations Number of Financial Design & Year Companies Technical (Equity} Drawings Total 1980 10 9 Nil 2 11 1981 9 4 1 4 9 1982 s 3 1 1 5 1983 12 13 1 3 17 1984 14 15 0 4 19 1985 21 23 2 6 31 Source: India Investment Centre, New Delhi. C. IMPACT ON INDUSTRIAL DEVELOPMENT 6.12 This section reviews the impact of government policy on market structure, including the number of firms and changes in capacity, as well as the effects of the market structure on market shares. It also looks at general indicators of performance (capacity utilization and growth). The international competitiveness of the Indian machine tools industry is dis- cussed, with the focus on economies of scale, marketing strategy, pricing and quality control. Finally, product development and technology are reviewed. 1. Impact on Market Structure and Performance 6.13 The total production capacity of the Indian machine tools industry has incrPased steadily (Table 4). Installed capacity in terms of millions of US$ jumped from 106 in 1977 to 283 in 1984. The number of firms rose from 115 in 1974 to 137 in 1983/84. It is noteworthy that in 1984, the number of firms fell by six compared to the previous year, while capacity showed a steady expansion. Installed capacity grew rapidly after 1980, from 9.2 percent a year during 1976-1980 to 19.8 percent a year during 1981-1984. One reason for this jump ~s that capacity utilization for 1980-82 exceeded 100 percent. The implication is that overutilization of capacity attracted newcomers and motivated existing firms to expand. More important, the large increase in capacity in the early eighties coin- cided with the liberalization measures begun at that time. Particularly in 1983, whea the government significantly 100sened the licensing policy, production capacity increased by 36 percent. It would appear that th~ government has successfully provided new momentum for the growth of the ·industry. ;.. 83 Table 4: ENTRY, CAPACI'lY Ul'ILIZATION AND TRADE; 1977-84 Installed Share of Number Capacity Capacity Indigenous Exports Imports of Firms (US$ Million) Utilization Production (US$ Million) (US$ Million) 1977 115 106.0 69.8 80.0 9.3 29.6 1978 132 138.0 75.2 82.1 24.4 48.8 1979 n.a. 121.0 95.1 83.0 20.1 46.1 1980 134 137.5 103.S 76.S 25.4 76.3 1981 n.a. 163.0 104.3 80.9 23.1 69.4 1982 143 192.0 113.0 73.1 19.5 88.1 1983 143 261.0 88.9 63.4 23.0 148 .. S 1984 137 283.0 78.8 63.4 20.s 132.3 Source: Ki.shore Jethanandani, "Entrepreneurship, Market Structure and Teclmology Development in the Indian Machine Tools Industry," processed , New Delhi, April 1986. Note: n.a. means not available. 6.14 As a large· number of Rewcomers has entered the machine tools industry in the 1980s, there has been a significant change in market struc- ture. The share of the top eight companies had increased steadily to 95.7 percent in 1978/79, but since that year, it declined to 76.3 percent in 1983/84 (Table 5). In the case of Hindustan Machine Tools Ltd. (HMT), the largest company in the industry and a well-managed public enterprise, its market share, after reaching 51.8 percent in 1978/79, declined to 44.1 percent in 1983/84. Thus, the new entrants have been competitive enough to penetrate established markets. This trend is important, given that the entry of new small-scale firms is still restricted, and large firms have several advantages. They are better able to provide marketing and after- sale services, which smaller firms cannot match. Moreover, a large public sector firm su~h as HMI' has ready access to public sector users and in addition has a 10 percent price preference. (Government users will purchase from HMI' if machinery of equivalent specifications is available from HMT at a price within 10 percent of the price of a •. rivate company's machines.) Large firms can also compete more effectively for bulk or 84 project purchases where competition is intense, as they sell their products at a discount. The successful penetration of small firms has been the result of the lack of significant scale economies in the industry, which allowed small new entrants to produce machine tools at comparable prices in many product lines. This teature is reinforced by the fact that the most competitive area is special purpose machine tools, for which economies of scale are less important than for general purpose machine tools. Further. small firms are often better able to cater to users' needs. For instance, in the market for surface grinding machine tools, it has been found that Alex Machine _Tools, a small-scale firm, is competing with Praga Tools, a large manufacturer, by selling at lower prices; it has even increased its market share substantially in this product. Table 5: MARKET SHARES OF LARGE FIRMS IN THE INDIAN MAOUNE TOOLS INDUSTRY (%) . 1964 1978/79 1980/81 1981/82 1982/83 1983/84 HMr Ltd. 43.59 51.84 36.61 45.16 41.56 44.06 Mysore Kirloskar 16.93 21.63 17.44 16.48 11.46 11.83 Cooper Eng. 4.39 6.87 7.41 6.29 6.20 Bharat Fritz Werner 3.75 2.84 2.28 2.30 Invest Machine Tools 3.49 Heavy Eng. 'Arp. 0 8.52 9.26 13.24 4.97 6.53 Praga Tools Ltd. 0 2.44 2.02 1.64 2.00 2.96 Kirloskar Brothers 0 1.59 2.74 2.ss 2.35 2.33 Total (for Large Firms) 12.15 95.73 75.48 85.36 64.62 76.31 Total (for Industry) 100.00 100.00 100.00 100.00 100.00 100.00 Source: Kishore Jethanandani, "Entrepreneurship, Market Structure and Technology Developrent in the Indian Machine Tools Industry," processed, New Delhi, April 1986. - 85 - 6.15 A second aspect of the market structure worth emphasizing is that MRTP regulations and licensing have been ineffective in preventing concen- tration. As mentioned earlier, the market share of the top eight companies increased until 1978/79. At that time, MRTP regulations and licensing severely restricted new entry and.expansion, although that result was contrary to what was intended. As a result, in the early 1980s, the government eased the MRTP regulations and licensing, and the market share of the dominant firms decli~ed steadily. 6.16 As to capacity utilization, it peaked at 113.0 percent in 1981/82 and then fell to 78.8 percent in 1983/84. In part the reason was the sharp increase in newly created capacity, compared to the increase in demand: capacity doubled during 1980/81-1983/84, while demand increased by 76 percent. However, given that utilization was at a high of 103.S percent in 1979/80, the marginal discrepancy between supply capacity and the increase in demand does not fulJy explain the drop in uti.lization to such a low level as 78 percent in 1983/84. Moreover, the growth rates for produc- tion at constant prices dropped f~om 5.7 percent in 1978-81 to 3.1 percent for 1981-84, despite the large capacity creation in 1981-84. The reason for the decline in capacity utilization and growth in proaJction during the early 1980s was largely a surge in imports: they went from US$46.l million in 1979 to US$132.3 million in 1984, while the ratio of imports over apparent consumption increased from 17.0 percent to 36.4 percent. 6.17 There seem to be two reasons for this development. First, as pointed out earlier, the Indian machine tools industry is suffering from negative protection for its portion of value-added. As such, imports of comple~e machine tools are less costly than domestic assembly with imported components. Given this situation, users prefer imported to domestically assembled machine tools. Second, the domestic machine tools industry does not meet users' requirements. The automotive industry, for example, a recent survey pointed out that the bulk of the machine tools it requires are high precision and NC/CNC machines, which the Indian machine tools industry does not have the technology to produce. As regards general purpose machine tools, the automotive industry has problems with the prices, quality, delivery and after-sales service of domestic firms. Finally, some machine tools do not have proper tooling and accessories, limitations that adversely affect the productivlty of the automotive industry.~/ 2. Competitiveness and Economics of Scale 6.18 Since 1977, exports of Indian machine tools have stagnated at around US$20 million: while imports have increased very rapidly. The ratio of ·exports to domestic production dropped from 23.0 percent in 1977 to 10.6 percent in 1984. HMT's plan to sell 25 percent of its total production abroad has not materialized, and most exports involve government-to- government transactions. The implication is that the Indian machine tools industry is not competitive by international standards. 5/ Op. cit., "Machine Tools •• ,." - 86 - 6.19 In the case of manually operated general purpose machines, a number of products are not competitive with their counterparts from other newly industrializing countries in terms of price (Table 6). Further, there has been a change in the demand for ~achine tools in the world market. That for conventional general purpose machines is declining, and only a limited market will remain. The technology of these m~chines is becoming obsolete, and they are being replaced by more productive and sophisticated types of machine tools. Instead, CNC technology is fast becoming a requirement for penetrating the international market, and doing so requires goods of acceptable quality. 6.20 Scale economies are less important with machine tools and other segments of the capital goods industry than with other mass production sectors. There are two main reasons. First, the share of capital in total production costs is generally low (S-7 percent is often mentioned). Thus, a source of scale economies~the high cost of setting automated machines-- is reduced. Second, the capital goods in this industry are highly divisi- ble. On the other hand, scale economies become important with exports because of the expense of world-wide marketing and the after-sales net- work. If scale with respect to overhead, marketing and labor organization in the CNC lathe industry is taken into account, the economies approach 40 percent as output increases from 100 units to 800-1,000 units a year. How- ever, given its low labor costs, the breakeven point in India should be far below 800-1,000 units of ·CNC lathes.6/ 6.21 Japan is a case of a country that has taken maximum advantage of the benefits of scale economies. Whereas in the past what mattered was performance rather than price, Japan has reversed this concept success- fully. The key factor in Japan's strategy has been the design of smaller and lower performance and cost CNC machines through standardizing the spe- cifications of the product as well as reaping the benefits of scale econo- mies. In Indi.a, in contrast, the production base is fragmented, and the number of units produced by individual firms is very small. For instance, in the case of CNC lathes, the largest Japanese firm produced 236 units in 1970 and 2,438 units in 1981, while all firms in India produced only .4 CNC lathes in 1981. At this scale of operations, an Indian machine tools firm cannot be competitive in the international market. 6.22 Other factors co~tributing to the uncompetitiveness of the Indian machine tools industry are overmanning and a lack of product specializa- tion. Machine tools in Taiwan (Province of China), and the Republic of Korea are manufactured in small/medium-sized firms, each of which employs around 10-80 personnel. Of the 200 machine tool factories operating in Taiwan, only about 25 employ more than 80 people. Moreover, they specialize mostly in the manufacture of one or two types of general purpose machine tools, such as lathes, drilling machines, milling machines, grinders, etc. This specialization in a few products has given them the competitive edge in product pricing. In the organized sector of Indian machine tool industry, in contrast, product variety .is the dominant characteristic, while too much labor is a typical Indian problem, with a large proportion of it employed in non-productive operations • Op. cit., "Entry into the Production •••• . - 87 - Table 6: COMPARATIVE EXPORT PRICES, EQUIVALENT RS LAKHS Machine India Taiwan China Korea, Rep. Center.Lathe 400 x 1,250 mm 5 kW 1.6 1.1 0.8 Center Lathe 500 x 1,500 mm 9 kW 2.8 1.6 Milling Machine 300 x 1,400 mm 5.5 kW 2.s 1.6 Milling Machine 300 x 1,500 mm 7.5 kW 3.5 1.9 Radial Drill SO x 1,500 mm 2.5 1.2 Tool & Cutter Grinder 280 x 600 1.1 0.5 Source: .. The Indian Machine Tools Industry, A Perspective Plan, 1983-93," July 1983. 3. Technology and Product Development 6.23 The development and production of machine tools in Indi3 had its beginnings in the· first World War. As a rule, the early machines did ~ot meet international stantlards. A serious attempt to manufacture machine tools was made just before and during the second World War, based on . . British developments and standards. Today, one of the largest machine tool manufactarers 1n lndia--Mysore Kirloskar, Ltd.--still produces machines developed SO years ago. · Its low-powered (0.75 kW and 2.2 kW) cone pulley lathes still have a market and a.re produced in large quantities--50 machines a month. Their production is partly based on the inch syutem, even though the metric system wa~ introduced in India 25 years ago. This example shows how the level of technology of machine tools in India differs widely. - 88 - 6.24 Another example is provided by HMT. In the mid-fifties, HMT built a European-oriented (Swiss-German know-how) machine tool plant. The stage of development was in accordance with the international technology of the time. To meet the growing demand for different types of machine tools, HMT decided to collaborate with many internationally recognized machine tool manufacturers in the following years. 6.25 The gap in technology between Indian and international standards was insignificant during that time. However, during the past 15 years, the application of numerical control has revolutionized manufacturing tech- nology, worldwide. Because of the rapid development in machine tool technology, it seems impossible to bridge the technology gap using India's own resources within a short time. The capability to develop machines that meet the high level of international standards will take time. Some companies are currently attempting to convert a mechanical machine into a NC-operated one. As the main spindle speed and tools still have to be changed manually, this solution can be considered as only an intermediate step and as unsatisfactory in the long run. 6.26 In general, the status of process technology in the Indian machine tools industry is comparable to that of industrialized countries 15 to 20 years ago. A few firms are planning to modernize their plants, and some others have started to do so and have made substantial investments, especially in the machining of medium and heavy castings. Machining centers used for this purpose will replace a number of standard milling, drilling, planing and boring machines. 6.27 In the machine tools industry, design and development of new products are the major way in which technology is created. As shown in Table 7, product development has a cyclical pattern that rises in reces- sions (the number of newly developed machine tools was highest during the recessionary periods 1967-69 and 1976-78) and declines in periods of high growth. A lesson derived from this pattern of product development is that times of hardship, when competition increases, tend to drive firms to engage in more intensive technology development. A firm that fails to develop new products and new process technology will simply lose its share in the smaller, more competitive market and be driven out. Thus, competi- tion is also a prerequisite for brisk technology development. An absence of that competition may be one reason a protected economy lags behind an open one in technology development. 6.28 Given this pattern, product development in India should have been brisk during the early 1980s, when the government began to 'expose firms to more external competition. However, actual performance did not correspond to expectations. For example, the number of newly developed machine tools during 1982-84 was far below the level attained during the recessionary period of 1976-78. The reason is that the change in the nature of techno- logy toward products requiring electronic controls and custom-design requires greater skills in the development of state-of-the-art machine tools than that were available in the Indian machine tools industry. To overcome this bottleneck, 14 leading machine tools producers have entered into foreign collaborations in order to assimilate advanced foreign - 89 -· Table 7: PRODUCT DEVELOPMENT IN THE INDIAN MACHINE TOOLS INDUSTRY Number of Period Newly Developed Percentage Growth of Machine Tools Production at Constant Value 1964-66 i6 15.5 1967-69 289 -6.0 1970-72 168 s.8 1973-75 137 9.9 1976-78 398 -1.0 1979-81 218 5.7 1982-84 213 3 .1 Source: Annual Reports, Indian Machine Tools Manufacturers Association. Note: Because of insufficient data, the figures for 1964-66, 1973-75 and 1982-84 cover 1965-66, 1974-75 and 1982-85 (1983 excluded), respectively. technology. However, a number of policy constraints remain which made it difficult for Indian manufacturers to respond to the more liberal price signals • . - 90 - - 91 - INDIA - INDUSTRIAL R.€GULATOaY POLICY STUD..,! * VII. Subsector Review - Automotive Products A. BACKGROUND 7.01 The Indian automotive sector consists of units producing passen- ger cars; light and heavy commercial vehicles; two- and three-wheeJ. vehi- cles; tractors; jeeps and automotive components. The industry began in the 1930s with the assembly of complete knockdown (CKD) kits by Ford and General Motors. The first two Indian vehicle manufacturers, Hindustan Motors and Premier Automobiles, started in 1942 and 1944, respectively. Local assembly of CKD vehicles dominated the market until 1953, when the government banned assembly activities by foreign companies. As part of an ex"l.icit GOI policy to promote entirely indigenous automobile manufacture, ve ..cle firms were required to achieve increasing levels of domestic content. In response to this domestic content requirement, component 1/ manufacture for the European vehicles being produced in India started In the mid-1950s, with financial support from European component makers, and grew rapidly. 7.02 The automotive sect~r in India is today one of the most highly indigenized in the world, with about 95 percent domestic content. By 1985, India's five car manufacturers produced 103,000 cars, 8 firms produced about 99,000 commercial vehicles, and 25 firms produced about t.2 million two/three wheelers. This chapter first describes the regulatory environ- ment which has shaped the struct.1re of the automotive markets and influ- enced firm behavior. Next, the impact of market structure and firm behavior on performance is analyzed. The charter concludes with a summary of recent policy reforms affecting the automotive sector, and includes recommendations for additional policy changes. 7.03 Until recently, capacity licensing restricted the entry, growth, exit and consolidation of domestic firms, and vehicles and components enjoyed complete protection from international competition as a result of a virtual ban on imports of vehicles and toos.t components. Equipment and machinery imports, foreign technology licensing and direct foreign invest- ment were closely restr~cted. 7.04 Domestic demand for cars is limited. The ratio of people per car in India is 788; only China, Bangladesh, Burma and Ethiopia have higher ratios. The largest segment of domestic demand is for two-wheel vehicles, sales of which increased from 69 percent to 80 percent of total vehicle * For an in-depth analysis of the automotive sector, including detailed recommendations regarding future strategic options, see forthcoming World Bank report, "Automotive Products Industry: Problems and Prospects.•• !/ The term "component" includes parts,- spares, automobiles, suo- assemblies, etc. - 92 - sales between 1975 and 1984. The fastest growth {28 percent per year during this period) was in mopeds, the low-cost end of the two-wheel market. The reasons for the low level of domestic demand for cars and two-wheel vehicles are e:~t~emely low disposable incomes, high production costs and domestic taxation; the latter in particular increases the retail retail price of vehicles by 40-70 percent over the ex-factory cost.2/ In spite of limited domestic demand, however, as a result of capecity licensing-induced restrictions on output, demand often exceeds supply, and waiting lists for the products of better firms are not uncommon. 7.05 In the case of commercial vehicles, demand is limited by a low level of investment in road expansion and maintenance. Medlum and heavy commercial vehicle {CV) growth is dampened by the predominance of rail transport rather than road transport for goods and people over long distances. The growth in demand for light commercial vehicles for urban passenger transport is constrained by competition from three-wheel ve~i­ cles. In the 1960s and 1970s, commercial vehicle sales increased, respec- tively, by 3.4 percent and 6.9 percent annually. From 1980-83, CV sales increased at an annual rate of 9.8 percent, but this growth slowed to 4.9 percent a year between 1983-85. By that latter year, commercial vehicle sales had reached 99,000 units; between 1974-84, roughly one-fourth of the market was for light commercial vehicles {up to six :ons gross vehicle weight), 72 percent was for medium trucks {6-16 tons), and 2 percent heavy trucks (over 16 tons). 7.06 Passenger transport pricing and ownecship policies have further limited the potential demand for commercial vehicles, particularly for medium and light ones which could be used for in-city transport in place of heavy, large buses which currently dominate. In some states (Maharashtra, Gujarat and Haryana), passenger transport is the exclusive monopoly of the state. In these cases, tight state budgets restrict the growth of fleets and vehicle turnover. Elsewhere, state transport corporations enjoy advantages over private operators in the allocation of routes.3/ In the case of both private and public passenger transport operations-;- fares are set by the state ~overnments and are of ten insufficient to cover operating 2/ In 1977, a study of the incidence of indirect taxation on the auto- mobile industry was conducted by the Association of Indian Automobile Manufact~rers. The tax component in the final retail price varied from state to state because of differences in local sales taxes and octroi. Interestingly, the rate of the sales taxes on cars--considered a luxury item--in some cases was lower than the tax on two-wheel vehicles. In March 1977, the excise duty on the Maruti car was increased to the present level of 25 percent ad valorem, while the sales tax is 35 per- cent. Import duties on imported components are 55 percent, recently increased from 40 percent, and auxiliary custom duties and import excise tariffs are 30 percent 9.0d 5 percent·, respectively. Thus, the total tax on the Maruti car exceeds 60 percent of the ex-factory price. !.J State transport corporations of ten have the priority in choosing a route and may be awarded exclusive rights; in such cases, private operators are restricted to routes not claimed by the public ones. - 93 - costs. Whereas public operators are generally subsidized by state govern- ments, unrealistically low state-set fares inhibit potential entry by private operators and constrain the ability of state transport companies to recover their operating costs. One result is that the public carriers have to operate their buses for longer periods before replacing them. Because of the limited domestic demand, the scale of production in light commercial vehicle assembly is generally suboptimal, and the growth of demand is sluggish. 7.07 In 1982, the GOI initiated a program to modernize and expand the automotive sector, spearheaded by a new public sector company, Maruti. Maruti was established to manufacture low-cost, fuel-efficient cars and light commercial vehicles. Prior to Maruti's entry, the automotive sector had been dominated by private firms. The modernization program stimulated by Maruti's entry led th~ GOI to relax the nates on capacity licensing and choice of product lines, approval of foreign collaborations, and access to imported components and technology. This relaxation of the regulatory and technology policies in the automotive sector has stimulated increases in output, modernization of product lines and productive techniques and the introduction of fuel-efficient, light-weight vehicles in India. B. THE REGULATORY POLICY ENVIRONMENT 1. Capacity Licensing 7.08 The major determinant of the structure of the automotive sector has been the industrial licensing system, which has constrained entry, growth and exit. Some relaxation in licensed output levels has been permitted from time to time in recognition of the fact that installed or rated capacity often exceeded licensed capacity. Between 1975 and 1980, manufacturers of commercial vehicles, scooters and ancillaries were allowed to produce in excess of licensed capacity, without limits on output, provided no additional investment in machinery was required; this provision for "regularization of excess capacity" was renewed in 1980 for two- and three-wheel and commercial vehicles. Starting in 1975, commercial vehicle and ancillary manufacturers were also allowed to expand lic.ensed capacity by 5 percent a year up to a maximum of 25 percent in five years (beyond the 25 percent expansion in five years that was normally permitted under the industrial licensing policy). This "automatic growth" rule was renewed again in December 1985. Beginning in 1981, firms were also allowed to .. produce freely in excess of licensed capacity provided the entire addi- tional amount was exported. In addition, starting in 1982, firms were allowe4 to expand licensed capacity by one-third provided that the highest level of output in the preceding five years, plus one-third, was higher than the licensed capacity plus 25 percent (the implication is that capacity utilization was 94 perr.ent). In 1986, this provision was relaxed, allowing "re-endorsement" of capacity by one-third provided capacity utili?.ation has r~ached 80 percent of licensed amounts. - M - 7.09 Each relaxation in the limitations on output was subject to restrictive conditions, including that the product not be reserved for manufacture in the small-scale sector; that certain export obligations be undertaken; and that firms abide by restrictions on !~cation and capital imports. The liberal growth provisions were not, however, allowed for Monopoly Restrictive Practice Act (MRTP) and FERA firms unless the product was on the list (Appendix I of the act) of products open to investment by those firms. (The only exception to the prohibition on expansion by MR.TP firms was in the case of automatic growth, which was allowed for non- dominant MRTP firms.) Further, any expansion requiring imported machinery or an increase in consumption of imported raw or intermediate materials required clearance by the licensing authorities. Until 1982, entry and expansion were generally tightly controlled. 7.10 One reason for this strict control on expansion was that, with annual domestic car sales of 103,000 (1985) and negligible export volumes, an unregulated market wo1•ld have resulted in a "natural" m:>nopoly for cars, with production by a single firm comprising the least costly means of supplying the industry's output. The barriers to entry and expansion have led, however, in most segments to scales of production that are small by international standards. The automotive sector is one of the most suscept- ible to scale economies. In the case of cars, for example, recent estimates of the minim\im economic scale are 200,000 to 350,000 units annually for each model, given flexible, highly automated assembly techniques.4/ Previously, production techniques employed mass production methods requiring expensive, special purpose machinery and thus even larger volumes per run were needed to achieve lower costs; in fact, 400,000 units a year per plant was considered the feasible minimum. Estimates of the economic scale for different activities range from 200,000 for some castings and assembly to more than 1 million for machining and stamping (Table 1). In the US, automobile output in 1983 ranged from 55,000 units (in a new Honda plant operating at low capacity utilization) to 4 million units per firm (GM). In other developing countries, such as the Republic of Korea and Brazil, output in 1983 by the largest firms in each country was 108,117 and 341,354 units, respectively. By comparison, the licensed capacity of India's two major car manufacturers was, until 1984, 30,000 and 18,000 units each and by 1985, total licensed capacity in India's five car manufacturers was only 185,000 units. The small scale of production by car manufacturers is reflected in the correspondingly small scale of production in many of the OE component firms supplying them. 7.11 In light commercial vehicles, which employ a mass production technology similar to that for cars, the annual output in India of 34,000 units (1985) is also small by international standards. Total installed capacity for these vehicles, which is now 87,500, is fragmented among nine firms, with the licensed capacity of the largest at 40,000 a year. A 1972 4/ Altschuler, The Future of the Automobile, Cambridge, MIT Press, 1984; OECD, Long-Term Outlook for the World Automobile Industry, Paris, OECD, 1983. - 95 - study of the economies of scale for automated mass-production technologies employed in car and light commercial vehicle manufacture estimated that the cost savings per unit were 1,520 percent for output levels between 50,000-100,000 and 5-6 percent for levels between 125,000-175,000.5/ Clearly, the small scale of production in cars and light commercial vehicles is a major factor in the high unit costs of these vehicles. Table 1: ESTIMATES OF MINIMUM EFFICIENCY SCALE (thousand units a year) Overall Complete Model Casting Machining Stamping Assembly Model Range Pratten 8/ 1,000 250 500 300 500 1,000 Rhys b/ - 200 1 000 2,000 400 2,000 White-Cf Small 260 400 200 400 800 University 1,000- 400- 500+ 200- 200+ 1,000+ Study Group d/ 2,000 1,000 400 With small economies thereafter Source: K. Bhaskar, Nichols, New York, 1980, The Future of the World Motor Industry. a/ c. F. Pratten, .. Economies of Scale in Manufacturing Industries," 1971. b/ D. G. Rhys, "The Motor Industry: An Economic Survey," 1972. c/ L. J. White, "The Automobile Industry Since 1945," 1971. d/ University of Bristol Motor Industry Research Group. 7.12 On the other hand, production of medium and heavy commercial vehicles (trucks and buses) is based on individual assembly and is not as scale-sensitive. In the U.S., the output of heavy commercial vehicle firms in 1983 ranged from 2,000 (VW) to 1.1 million units (GM). In the 6-8 ton range, which covers the largest segment of the Indian commercial vehicle market, scales of less than 30,000 can be efficient. At the heavy end of this vehicle range, production is even less scale-sensitive and required output volumes are smaller (3,000-10,000). (See Tables 3A-C at the end of the chapter for indications of international production scales.) Thus the 1985 output of 46,000 units by TELCO, India's largest heavy commercial vehicle manufacturer, compares favorably with international norms. 5/ D. G. Rhys, ''Economies of Scale in the Motor Industry," Bulletin of Economic Research (Nouember 1972): 91 - 96 - 7.13 The largest segment of the Indian automotive market is, as noted, two-wheel vehicles. The scales of prc;duction by the largest two-wheel firms are close to international levels. Bajaj Auto Limited, whose produc- tion in 1984 was close to 200,000 units, is one of the largest scooter manufacturers in the world (with its recent expansion, Bajaj installed capacity will reach 700,000). Nevertheless, t.here are a large number of firms in each segment of the two-wheel vehicle 1116rket whose production scale is low. There are seven firms producing scooters, with the 1984 output of the smallest two less than 5,500 units each. There are six motorcycle firms; the 1984 sales of the smallest reached only 273 units. Of the six moped manufacturers, the smallest three sold fewer than 10,000 units each in 1984. Capacity in the components sector is similarly fragmented. For example, there are 14 licensed manufacturers of clutch plates, compared to 4 in Japan, which has much larger volumes of vehicle production, and 7 licensed carburetor manufacturers in India, compared to 3 in Japan. · 2. MRTP Act and FERA 7.14 While the barriers to growth, particularly for MRTP and FERA firms, have often led to the fragmentation of capacity among firms, at the same time the barriers to entry have led to the concentration of markets. As mentioned above, regulatory barriers to entry and growth have been more binding on MRTP and FERA firms than on others. Licensing and MRTP clearance barriers to growth have reduced the actual competition among MRTP/FERA firms, since they have not been readily able to expand output to compete for market shares. Until recently, the barriers to entry inhibited ~ther MRTP/FERA firms that could have offered the most credible threat of competition to incumbents. On the other hand, as noted above, less restrictive entry and growth barriers for small and medium-sized fini'.8 have allowed the entry and survival of inefficient small units, with a resultant fragmentation in capacity among several firms. In almost every segment of vehicles and components production, the number of firms exceeds the number of tecl1nically and economically efficient units that the small size of the Indian market could support. 7.15 Until recently, there were four licensed car manufacturers, two of which (HML and PAL) accounted for close to 100 percent of the market (Table 2A.) The same high level of concentration was found in the two-wheel vehicles market, where 4 of the 25 licensed firms accounted for 65 percent of the market. The leading firm in each segment (motorcycles, mopeds and scooters) accounted for 42 percent to 63 percent of each market segment (Table 2D). In components, there were an average of 6-7 licensed manufacturers for each product and for some products there were as many as 20 licensed manufacturers. However, the leading component firm often had between an 80 percent and 100 percent market share; in other products, the main market share was held by the top two or three firms (Table 2E). Other companies had low market shares or, even though they took out an industrial license, never started production. For most components, the Herfindahl . ' - - Table 2A: SALES AND MARKET SHARES. CARS HML Premter (PAL}/ MUL SMPIL SAL Ambassador/Contessa Pac:tmtnt Marutt Gazet Bad•l/Dolohin TOTALS Growth Market Market Market Market Market Market .. s Vear Sate Share Sale Share Sale Share Sate Share Sale Shara Sales Share for & Vra. I lHo. I vehicles) '") I (No. I vehicles) '") I (No. un (No. hn I (No. on ("Total) IVehic lea) Vehicles) ivehtclesl I I I I I I I I I I 3,223 3,493 3,624 3,301 17 16 15 22 l I I I I 18,937 21,658 23,514 14,999 100 100 100 100 l ·I 3,676 I 3,850 1,170 I 2,674 I 2,263 16 15 4 8 6 I I I I I 23,284 24,834 27,869 33,140 37.383 100 100 100 100 100 I I l 5.6 I I 1969 2'. 333 61 '12. 185 35 f I 0 552 4 35,070 100 1970 23, 124 65 12,083 34 I. s21 1 35,128 100 7.5 1971 24,215 64 12, 765 34 I 131 2 37, 717 100 1 1972 1973 1974 24,574 25,383 20,558 63 64 56 13,614 l 13,875 I 14,360 35 35 39 I 686 603 · I 1 ,643 2 1 5 38,874 39,861 36,561 100 100 100 I 1975 9,477 41 I 13 ,436 58 I 153 1 23,066 100 8.4 ~ 1976 15,272 50 15, 167 50 I 163 30,602 100 1977 20,328 46 1 17,464 54 I ::: 37,903 100 I 1978 21.173 62 12.379 37 348 I 100 1979 1980 1981 1982 1983 17,117 22,056 22,178 22,804 24.667 59 71 54 52 54 11 • 675 8,933 18,675 120,842 120,751 40 29 46 48 45 I 100 I . 54 8 4 I 1 104 51 30 122 302 1 I 28.950 1····" 31,048 40.887 I 45,821 43,769 100 100 100 100 100 6. 1 1984 24,337 39 126,245• 42 110,719 17 919 2 I &2,220 100 1985 29.820 29 24. 720 24 48.410 47 103.000 100 27.0 Source: Association ot Indtan Automotive Manufacturers. The Automobtle.lndustrx, Statistical Proftle, Bombay, 1985. Table 28 SALES AND MARKET SHARES MEDIUM AND HEAVY COMMERCIAL VEHICLES ---TV PE OF I TEL CO Market AL Market PAL Market HML Market. SCL Market TOTALS MARKET Growth S P.A. V~d!.. ! ~HICLE I~. Vehtc1es) Share • !!.!.!!. Shtre • Se lea Share • ~ Share • ~ Share 1' I ~ $HARE U' to Total) for 5 Vra. 197L 1973 1974 T1 u 4..kS 17,019 18.202 15,984 78 73 67 ! 1. 797 2,539 3. 129 8 10 13 '· 766 2.492 2.851 It 10 12 1.387 1.871 1.933 6 7 8 121.969 25.104 23.897 100 100 100 I 191~ I 18.242 75 3.148 13 1.977 8 973 4 24.340 100 15.731 10 4.584 20 1.443 1 744 3 22.502 100 191b 1977 I 17.879 I 18.637 76 3.862 5.388 16 21 t.187 860 5 3 581 1. 145 3 4 23.509 26.030 100 100 t978 1979 t980 I 25.050 24,327 72 76 71 5.863 6.577 18 19 658 706 2 2 I 1.237 2.323 4 1 515 1 32.808 34.448 100 100 I l9tJ I I 31,735 70 8, 134 18 1.987 4 2.891 6 713 2 145,460 100 198.l I 33,421 11 7,832 18 1,510 3 823 2 43,586 100 I 36,014 80 7,966 18 558 1 406 •1 144.944 100 t983 1984 I I 31.068 80 7,072 18 I 344 1 421 l 38.905 100 ---- -+- I I •,673 67 2,333 27 375 4 163 2 8,544 100 I l91L t:su:. I 4,221 1973 1974 t97~ 1976 I I I I 1,493 ._972 ',483 1.005 56 55 62 62 3,014 I 4.309 4,915 37 39 36 35 358 407 139 88 5 4 1 186 261 '37 115 2 2 1 1 8,051 10,861 12.068 14.123 100 100 100 100 I . ,387 57 11.707 100 ' I t977 I 63 4. 171 37 92 1 1978 I ,814 57 5.782 42 I 81 1 46 113.669 100 100 .I ~ ,679 56 6,015 43 58 90 13,842 1979 1980 I I ,829 54 6,426 45 I 49 137 1' I f 14.441 100 100 t • I 1981 I ,939 63 6.880 36 I 96 141 I 119,056 I I 26 116,804 100 'il7 1962 I 1,544 57 1. t96 43 38 I I 6 100 ! 1983 ,398 64 6,406 36 41 1984 I ', 703 63 7,347 37 91 I ,851 20, 141 100 ---·-- ~-- - i I I 2.14• ! HH:l 1973 1974 1975 I rr u I I I cUld '. b9:l ',695 ,956 '• 725 74 b8 63 71 I I I I 4.130 5.553 1,350 7,457 14 17 21 20 I 2.850 3.258 2. 116 1 9 10 6 1.550 2.os1 i. 194 \.110 5 6 6 3 I 130,513 33. 155 134.758 36,408 100 100 100 100 4.4 1976 I '• 736 68 I 9.499 I 8,033 26 I 1.531 4 859 2 I I 1'6,625 35.2\6 100 100 I 1977 I '. 266 72 23 1.279 3 I 638 2 100 I 1978 I •,451 67 lll,116 28 I 941 2 1, 191 3 39.699 1979 I '. 729 10 111,878 25 I 716 2 1.327 3 46.650 \00 .• 156 66 f 13,003 27 I 755 2,460 5 515 1 148,889 \00 6.0 1980 I I 15,014 I 2.083 ' 713 f 64.516 I 1981 I I, ti74 68 23 3 3.032 5 l 100 19fU I ',965 71 f 15,028 25 I 1,548 3 849 1 160.390 100 HUU l ,412 75 I 14,372 23 599 1 412 1 62.795 100 19~4 I I,111 74 114.419 ,. 24 I 435 ' I 421 1 I 159,046 I 100 6.3 I I ~ - ...... ' • f Table 2C: SALfS ANO MARKET SHARES. LIGHT COMMERCIAl. VEHICLES 8TL Market I I M&M Market SMPU.. Mliri.et ....... Ma,.ket HML Market TOTAi..$ hrltet Growth ~ Vear sates (No. Share -m- , l Sales Share nro:- --m- mo:- -00- Salea Share ~~ (No. -T~T f•'•• Snare ---00- ,~ ~ TS'Fcitat) tor 5 Vra. Vehtcles.) IVehiclesl eh i c I es l \1 Veh\ctesl Vehtctea} 1972 I 5.044 3,460 49 I t. 832 12 1,326 19 1,452 20 l I 9,407 1.011 100 1913 60 t 70 •4 1,027 l2 - 1, '25 t ,079 13 41 I I \00 HH4 I 3.737 4,500 60 l,147 706 18 281 5 I l7 13 6,257 6, 751 100 100 l I 1975 67 773 11 771 1976 4,564 4,875 56 60 l,161 1 ,034 " 14 13 l,717 :n 20 708 556 "9 7 1 15 8, 165 J a.oeo 100 lOO 3 t977 1978 1979 l I 6,489 6,553 66 55 t,351 2,931 14 24 - l ,624 I 1 ,ae1 , 2,441 19 20 129 131 , I l ' 9,860 112,056 100 100 I 3,573 f17,06t 100 1980 1981 19&2 I I 9,766 8,615 n,&10 57 42 46 I 1,082 I a.474 21 34 33 3,491 4,TlVB OOMPONENT MANUPACfUR!RS, INTERNAi. HARDT STRlJCfURE, 1984 Pl'oduct lletened Majority of No. No. No. POl\l'•Fim or fol' Output Fl'OID Lice111ed Pit111e in Lettel'S of Concentration Small-Scale Small-Scale Pl'oduct Firu Pl'oduction Intent Ratio H•I11du tnduatey Pima !!tJine Puts Piato111 7 7 6 o.88 0.21 Substantial Production Pia ton Pi 111 8 7 6 0.82 0.19 Substantial Production Piston Rings 8 7 5 o.87 0.23 Gaskets 4 4 4 1.00 0.35 Substantial Inlet & Exhaust Valves 10 4 7 i.oo 0.42 Carburetol'S 8 7 2 o.91 0.32 Fuel Pumps (Petrol) 3 2 0 1.00 o.73 Fuel Pumps (m/cyl) 5 1 0 1.00 1.00 Fuel Puaps (s/cyl) 5 4 4 1.00 0.78 Fuel Pu1lrp Nossle Holdere 5 2 3 i.oo o.96 Fuel Pu1lrp Nossles 5 5 4 o.98 0.75 Fuel Pump ile111enu 5 s 4 0.99 o.78 Fuel Pump Deliveey Valves 5 5 4 0.99 0.87 Filters 7 5 l o.98 0.32 Substantial Pl'oduct ion Fly-wheel RilllC Geal'S 4 l 0 i.oo 1.00 Radiatore 7 7 6 0.19 0.11 Rese1ved Thermostats 2 2 0 1.00 i.oo Bi11etal Bearings 6 5 7 o.9a 1.00 Substantial Pl'oduction Electl'ical Parts Starter Kotora 7 4 6 1.00 o.s8 Generators 7 4 6 i.oo 0.56 Voltage Rer.ulators 7 s 6 o.99 o.43 Diatr'ibutors 3 2 l l.oo o.98 Spark Plugs 3 2 0 l.oo 0.79 Fly-wheel Magnetos 4 2 2 i.oo 0.54 Drive 1 Transmission and Steed!!& Parts Clutch Assembly 11 7 4 0.10 0.2 Clutch Platea 14 8 s 0.15 0.11 Clutch Facings 7 ? 4 Tie Rod Ends 5 3 6 l.oo 0.87 Steering Gears 6 2 4 1.00 o.s1 Gears 22 9 9 0.89 0.24 Pl'opellal' Shafts 8 4 3 1.00 0.41 u. J. Cross Kits 11 7 0 0.88 0.22 Axle Shafts/Asaembly 17 7 s o.88 0.23 Oil Seals 5 4 1 i.oo 0.55 Reserved Wheel 5 2 4 i.oo 0.12 SU8J!!USion and Braking Pa!!!. Leaf Si: .ings Reserved Shock Absol'bers 6 3 8 1.00 o.36 .Ur Brake Asaembly 3 2 2 1.00 o.76 Brake Assembly 2 2 4 1.00 o.61 Bl'alte Linings 7 4 4 1.00 0.35 . Equient Headlights 7 3 0 J.00 0.47 lliper Motol's 8 2 3 1.00 o.94 !lectl'ic Roms 7 3 0 1.00 o.53 Dash Board l111trumenta 4 3 0 1.00 0.39 ~: Auto111otive Component Manufactul'ers Aasociation, 11,dia. - 102 - index 6/ was extremely high (greater than 0.7), indicating a high level of concentration in the industry, despite the proliferation of firms and an explicit government objective of limiting concentration. In fact, although industrial licensing policies have in many cases led to the existence of more firms than could survive in an unregulated environment, they have not successfully prevented concentration in many segments of the industry. While the extent of concentration is perhaps lower than it would have been without industrial rggulation, the static efficiency cost of allocating production capacity and market shares among a larger number of firms has been high in terms of production costs and poor product quality. 7.16 As a result of more liberal entry for small and medium-sized firms and the recent overall liberalization of entry, there is now excess capacity (relative to domestic demand) in each vehicle segment. Total licensed capacity for cars is 184,240, whereas 1985 sales were 103,000; for commercial vehicles, the figures are 342,500 versus 99,000; for two-wheel vehicles, 3.7 million and 1.2 million. The distribution of capacity and the rate of capacity utilization among manufacturers are uneven: firms with large market shares often operate at high levels of capacity utiliza- tion. Generally, an MR.TP or FERA firm is the market leader in each vehicle segment. With greater financial assets and managerial experience, these firms usually have wider sales and servicing networks, and their products are usually preferred by consumers, so that there are long waiting lists for some of their products. Yet for these firms, capacity licensing policies made expansion difficult before the recent increases in licensed capacity. At the same time, smaller firms, whose products ar~ often of poorer quality, often operate at low levels of capacity utilization and may be unable to sell all their output. Bajaj Ltd., with an average 62 percent market share in scooters, until recently had a waiting list of seven years of booked orders on which deposits had been paid, in spite of the presence of 11 other scooter manufacturers (between 1972 and 1984). The same phenomenon is observed in the case of component companies. For example, MICO, a FERA company, operates at 96-181 percent of licensed capacity in its various product lines and has 86-100 percent of market shares; Western Thompson (also FERA) operates at 167 percent of licensed capacity and has a market share of 83 ?ercent; Lucas TVS (FERA and MRTP) operates at 97-117 percent of capacity and has an aver~e 75 percent market share·in its more than seven products. 7.17 The automotive industry has been characterized by conservative managerial behavior and technological stagnation, largely the product of the barriers to growth. Interviews with automotive firms which had recently installed new machinery and equipment for expansion revealed that the most efficient choices for plant and work station layout and process flow had been rejected in favor of more traditional productive te~hniques. Actual competition to increase market shares has been limited by the growth regulations, and potential new entry has been deterred by restrictive capacity licensing. The limits on competition have reduced the need for minor cost-saving or output-expanding investments. In many vehicle plants, The Herf indahl index is a measure of concentration arrived at by adding the squares of market share for each firm. - 103 - even minor investments in semi-automated production techniques, such as mechanized transfer of work-in-progress or moving bolsters that allow for more rapid press operations, were not made until recently because the large potential increase in output would not have been allowed. The marginal allowed increases in output have also inhibited major modernization, which normally req1tires large-scale {"lumpy") investment that can only be justi- fied if output is permitted to ~pand by more than 25 percent in five years. The effect on product quality has been significant: pressed sheet metal was dragged across the floor, work in progress could be seen lying around on plant floors susce~ti,le to being banged and dented, plant layout was not improved to minimize th'' number of operations or distance and time between operations, etc. 7.18 Another characteristic of Indian vehicle manufacturers is that, relative to international norms, vertical integration is high. In Japanese and US car firms, an average of 30-40 percent of the value of the vehicle is manufactured in-house, w:tth the balance bought from outside vendors. This strategy allows manufacturers to shift inventory and labor costs to suppliers and reduces the final assembly costs. In India, the opposite is true. Vehicle assemblers generally buy 30-40 percent of the parts and manufacture 60-70 percent of the final vehicle in-house. The reasons given for the high level of vertical integration are: {i) to reduce the cascad- ing impact of excise taxes (which are levied at each stage of production); and (ii) to maintain higher quality components than is possible from out- side suppliers, particularly in the case of products whose manufacture is reserved for small-scale industries. (The proposals to introduce a modi- fied VAT system should ease the former problem.) 3. Small-Scale Industry Reservation 7.19 Small-scale firms have benefitted from regulations that reserve for them the manufacture of about 60 automotive components as well as from policy-determined cost advantages that act as a barrier to the entry and expansion of medium-size and large firms. The presence of firms that are not sufficiently large to afford technological progress is responsible for the frequently poor quality of components, particularly in the replacement market. Indeed, with several products reserved for small firms (leaf springs, radiators and bulbs), the investment required to produce a high- quality product may exceed the Rs 4.5 milliou ($375,000) limit that defines small ancillary f i!'DlS for the purpose of the reservation. The reservation also of ten prevents economies of scale in production and R&D. 7.20 In Japan and the U.S., the component sector is usually comprised of a few large firms that make high technology, sophisticatec' parts {e.g., Nippondenso in Japan) and many small manufacturers that specialize in one or a few components. There is close collaboration, including equity parti- cipation and technology transfer, between the original equipment manufac- turers {OEMs) component firms, which are often highly specialized, and the vehicle assemblers. In India, the activities of small firms often dupli- cated those of larger ones. Ii- the case of items reserved for production by small-scale companies, the large firms are prohibited from taking an equity stake in the small one. The reservation protects small firms from - 104 - the threat of backward integration by vehicle manufacturers and from potential competition from medium-size and l~rge-component firms. Under this system, the incentive for small firms to maint~in product quality and to update technology and product desj.gn to meet vehicle manufacturers' stringent quality requirements is reduced. 7.21 Fiscal incentives (exemption from excise taxes), effective avoid- ance of sales and corporate income taxes, and segmented labor markets, in which the total wage package paid by small firms is substantially lower than the wages paid by large firms, encourage companies to stay even smaller than the Rs 4.5 million 8tnall-scale investment limit.1/ Such low investment levels make it difficult, if not impossible, to achieve accept- able levels of quality. The cost advantages enjoyed by small-scale produ- cers as a result of the fiscal benefits and freedom from the minimum employee restrictions are ~stimated by the ACMA to be on the order of 50-70 percent. 7.22 Before the introduction of broad-banding (see para 7.31), product-specific licenses led to a high degree of horizontal specialization in four-wheel vehicle firms. Each of these manufacturers produced one or two products or models, but car firms generally did not make trucks, and vice-versa. This regulation thus inhibited plant-level economies of scale. In contrast, for two- and three-wheel vehicle and component manu- facturers, horizontal specialization has been limited, and there has been widespread product overlap. The most successful two- and three-wheeler firms have produced motorcycles, scooters and mopeds (although the ratio of the firm's primary to total output has generally been high, often because output levels of one product were limited by product-specific capacity licenses). In components, some firms have become highly specialized-- producing one to three products--whereas other firms have been producing up to eight to ten products. Although most component firms have been turning out an average of six to eight products, these items generally fall within one product category (engine parts, electrical parts, transmission parts, etc.). 4. "Sick Industry" Policies 7.23 The remaining important r.egulatory instrument affecting the structure and performance of the automotive sector has been the extreme 1J Firms with investment in plant and machinery of less than Rs 200,000 generally fall into the "unorganized" sector, which means they are outside the purview of th~ Factories Act. The transactions of such firms are "Outside General Ledger," which means that they are generally effectively exempt from excise duties, central sales tax and local taxes, which amount to about 38.4 percent of ex-factory costs in the case of automotive components (AJtomotive Component Manufacturers' Association, ACMA estimates). Component products manufactured by these "tiny" firms generally fall into two categories: substandard parts and "pirate" (imitation brand name) parts. Substandard and "pirate" parts are exempt from quality regulation, and the ACMA estimates that sales of "pirate" parts comprise a significant portion of the replacement market for some products. - 105 - difficulty of exit. Although the Factories Act includes legal provisions for. bankruptcy, the "sick industry" policy has prevented firms from closing a loss-making unit (other than small units with less than 50 to 100 employees). Further, the procedural delays and restrictions involved in mergers and acquisitions by MRTP and FERA firms have virtually precluded healthy MRTP or FERA firms from taking over the licensed capaci~y and assets of a "sick" unit. These restrictions have prevented the consolida- tion of capacity and allowed the survival of inefficient units while restraining the grotrth of healthy firms. Because of regulatory rigidities in closing operations or divesting loss-makers, the sunk cost of an invest- ment has been increased significantly, a situation that has reinforced the entry barrier. The regulatory barriers to exit have further limited competition in the automotive markets and contributed to the inefficient performance of the sector. C. THE IMPACT OF THE REGULATORY ENVIRONMENT ON PERFORMANCE 7.24 The regulatory policies (including small-scale industry reserva- tion and MRTP restrictions) have constituted binding barriers to entry and expansion. They have created a situation of scarcity that has allowed uncompetitive behavior and "supra-normal" profits by successful incum- bents. Indeed, in the area of two-wheelers and heavy commercial vehicles, the leading firms have been financially quite successful; Bajaj Auto Ltd. has often been the most profitable company in India. In 1983/84, the ratio of before-tax profits to net sales for automotive and cycle makers was 5.8, compared to 3.9 for engineering as a whole and 4.4 for all industry.8/ The ratios of profits to net assets and net worth were, respectively~ 6.7 and 15.0 for automotive and cycle makers, 4.0 and 9.1 for engineering as a whole, and 4.5 and 11.0 for all industry. The perception that successful incumbents enjoyed "supra-normal" profits is evidenced by the response to the recent liberalization measures in the automotive sector. There has been a spate of applications for industrial licenses and foreign collabora- tion agreements to set up new units to produce light commercial vehicles, luxury cars and two-wheel vehicles by some of India's leading industrial concerns (from both the automotive sector and other industrial sectors). 7.25 The result of the regulatory barriers to entry, expansion and consolidation has b~en the absence of effective competition within the automotive sector. Before 1984 (when capacity was expanded throughout the sector), even though the demand for domestic cars exceeded supply (as evidenced by the waiting lists), HMG and PAL maintained sales and output at about 30 percent below licensed capacity during most of the period 1950-80. For many years, the government regulated vehicle prices, a measure that eliminated competitive pricing. Firms came to believe that price competition was beyond their control because of the high cost of raw materials, high incidence of taxation in manufacturing costs and the inabi- lity to expand production to take advant~~ of potential economies of scale. Source: Association of Indian Engineering Industries, Handbook of Statistics, 1985, New Delhi. - 106 - 7.26 Competition has been the most dynamic in the two-wheel vehicle market, especially the fast-growing moped segment. Market shares of firms making mopeds have fluctuated, with KEL's increasing, several new entrants capturing some share and MIL's share declining (Table 2D). In scooter~, price competition appears to have been a minor factor in influencing changes in market share. Between 1965-85 the prices for the Bajaj Ltd. scooter increased more slowly than the price of API's, a~d Bajaj's market share gradually rose from 41 percent to 63 percent, while API's share declined from 50 percent to 3 percent. However, consumer surveys indicate that the primary factor in competition was product differentiation in terms of quality and reliability and Bajaj's service and distribution network. 7.27 In the car segment, the entry of Maruti is the first example of explicit competition on the basis of price: Maruti's retail price has been 31 percent lower than HML's and PAL's. Maruti's entry caused the first significant shift in market shares in cars since the 1960s. Between 1960 and 1983, when Maruti first produced any volume, the share of Hindustan Motors had remained between 48 percent to 54 percent, while Premier's increased slowly from 35 percent to 45 percent. Between 1983 and 1985, when car sales more than doubled, Maruti captured 47 percent of the market, Hindustan's share fell to 29 percent and Premier's to 24 percent. Market shares in light and heavy commercial vehicles have been staile, with only minor and temporary fluctuations. 7.28 As to price competition among component manufacturers, regulatory policies have segmented this market into large and medium firms on the one hand and small firms on the other. The small-scale segment is very competitive, with high price elasticities. In the medium-size and large segment, there is evidence of monopolistic behavior, with a high degree of vertical and horizontal integration that allows the cross-subsidization of prices among products and that creates a strategic barrier to entry. 7.29 An analysis of domestic resource costs in automotive firms reveals that there co-exist both efficient and inefficient enterprises. Domestic ex-factory price and quality performance compare favorably with international comparators in the case of some two-wheel products and some components, such as $COoters, motorcycles, brakes, clutches, engine valves, and a few other components. In a sample of 27 firms, short-run DRC ratios were less than one--that is, ongoing activities efficiently employed domes- tic resources to earn or save foreign exchange--for 18 of the 28 activities measured. However, for many firms in the sample, the outdated product specifications made it difficult to compare Indian products with modern international vehicles and components. 7.30 In summary, the performance of the automotive sector has been generally poor. Output has grown very slowly, and capacity utilization has been low for cars and light commercial vehicles, despite waiting lists for some products. Exports of vehicles have been minimal, accounting for less than 3 percent of the sales of most companies, the reasons being poor quality and high prices. Exports of components have been more successful; they constituted the largest share of engineering exports during the 1970s and early 1980s. Growth of component exports has dropped off in recent - 107 - years, however. A 1983 study by ICICI of a sample of automotive exporters showed effective rates of protection for domestic sales ranging from -39 percent to 65 percent, with an average of 18 percent whereas protection for export sales averaged -11 percent. D. RECENT POLICY CHANGES AND MODERNIZATION 7.31 In 1982, the GOI, as noted, initiated a program to modernize the automotive industry. This involved a number of significant changes in policies and has led to liberalized entry, growth and access to imported technology. Component manufacture no longer requires an industrial license, except for MRTP and FERA companies; the manufacture of components and two-wheel vehicles by those companies has been delicensed where the investments take place in specified "backward areas." Entry into the vehicle market is still subject to industrial licensing; however, licenses have been issued relatively liberally recently. Vehicles and components have been added to Appendix I, the list of products in which investment by MRTP/FERA companies is allowed, and the additional clearance procedures for th•>se companies have been waived for all vehicles and components. The latter measure is important, as the MRTP clearance procedures, including public notification about investment plans, followed by semi-public bearings on the proposed investments, effectively allowed incumbents to lobby against new entries. The reduced barriers to entry and growth should lead to an increase in domestic competition in the component segment and, to a lesser extent, in the two-wheel vehicle segment. As a result of the new liberal policies, licensed capacity for cars has already increased to 184,240, for commercial vehicles to 342,500 and for two-wheelers to 3,731,900. 7.32 In January 1985, the GO! also announced a policy of "broad-band- ing" for two- and four-wheel vehicles that allows manufacturers to utilize their total licensed capacity to manufacture any product for which they have a license. As noted, previous licenses were product-specific and indicated the quantity of each product that could be produced annually. !road-banding should allow for economies of scale to the extent that installed machinery is flexible and can be shifted readily among mode~s. This will be particularly important in two-wheel vehicles. Recent invest- ments in this sector have been in plant and equipment that can be shifted flexibly among motorcycles, scooters and mopeds, which should also improve capacity utilization in this segment. 7.33 Although the basic policy framework has not changed, clearances for foreign collab~rations (Table 4) and imports of capital goods have been implemented liberally, and new investments using imported technology have taken place in almost all two- and four-wheel vehicle segments, as well as in some component products. 7.34 There have been new entrants into each segment of the market. In a significant departure from past policy, many strong incumbents (i.e., MRTP and FERA firms) have received licenses to enter into the automotive - 108 - sector for the first time or to take advantage of broad-banding to diversify into new products, e.g., from heavy to light commercial vehicles. Table 4: FOREIGN COLLABORATION APPROVALS FOR VEHICLES AND COMPONENTS 1980-82 1982-83 1983-84 1984-85 1985-86 Amount (Rs Million) 10,285 40,052 37,587 % Total FC Approvals 3.3 11.1 7.6 No. Vehicles 7 9 6 2 No. Components 2 2 16 7 Source: Indian Investment Center and Directorate General of Technical Development. 7.35 In view of the recent liberalization of capacity licensing in the automotive sector; the major policy-induced barrier to entry is now foreign technology licensing. "Free entry" would imply that an entrant could use the same productive techniques as incumbents. The GO! policy regarding imported technology, however, explicitly discourages multiple technology imports, the objective being to conserve foreign exchange. Once a parti- cular product design or productlon technique has been imported, therefore, the licensing authorities may rtject a new application. Because modern technology is generally not available in India, technology licensing is de facto a barrier to entry. As a consequence of the high cost of imports of components during the initial years of recent new projects, for example, several recent applications to set up or modernize car projects have been rejected, at least temporarily. 7.36 In the case of new vehicle models, manufacturers, particularly those with Japanese partners, had insisted that more than one Indian component manufacturer enter into a new collaboration agreement in the interest of assuring timely supply of components of acceptable quality. The component manufacturers complained to the government about this practice, and in most cases multiple imports of a particular technology or multiple collaborations by component firms and the same Japanese partner have been rejected. (On the other hand, Japanese suppliers also had an interest in receiving royalties from more than one Indian firm.) Thus, potential entrants into a sector in which the domestic technology is of ten 30 years out-of-date are faced with the prospect of developing the product domestically, a process that entails huge R&D costs. Even if the licensing barriers to entry were removed, as in the case of the delicensing of compo- nents, the ease of entry of the second firm is still substantially less than that for the first firm as a result of technology licensing, a situa- tion that co~strains competition in these markets. 7.37 Technology licensing also provides an incentive for excessive horizontal integration. Firms import a range of productive techniques and product designs both to achieve plant-level economies of scale and as a strategic barrier to entry, since it is relatively more difficult for the - 109 - second and third applicants to import the same product designs and tech- nology, since it would be considered to be available in Ind~a. Thus, firms may import facilities to manufacture six or seven different automotive components rather than expanding output and specializing in a single one. The proliferation of models has also fragmented capacity in component manu- facture. One way for entrants to get around the restriction on m·1ltiple technology imports is to enter into a collaboration with a foreign firm that does not yet have a collaboration agreement with an Indian firm and that has a slightly different product desjgn or productive technique. The result in the automotive sector has been a proliferation of vehicle designs and 100dels and a corresponding proliferation of collaboration agreements with foreign component manufa~turers. This excessive product diversity inhibits the potential for standardization and product-specific economies of scale among manufactures of components. 7.38 In licensing new foreign collaboration agreements, the government has recognized the potential for domestic nx>nopolies implied by the small size of the domestic market. Therefore, it has used technology licensing to stimulate domestic competition by approving several collaborations in each automotive segment or product line. However, there is a trade-off between trying to avoid a concentrated oligopolistic market structure and achieving scale economies in production and R&D, particularly in the case of cars and light commercial vehicles. Since the domestic vehicle market is completely protected from import compe~ition, and the barriers to entry and growth continue to reduce potential and actual competition within the domestic market, even with several players in the domestic market, firms may not be motivated to behave competitively and may attempt to share the market among themselves. Unless vehicle firms begin to export, it will be difficult both to achieve efficient scales of production and to motivate competitive, technologically progressive behavior. 7.39 AJJ noted, the remaining important change in policy has been the liberalization of components imports. Vehicle manufacturers using newly imported vehicle designs and new plants and equipment are allowed to import a declining percentage of components, usually starting with a maximum of 60 percent of the vehicle value, for a specified period, usually five years. When seeking approval for a new investment, if the project requires imported technology or foreign collaboration, the applicant must enter into an agreement with the Direc~orate General of Technology Development on a program to reduce the imported components so that a domestic content of 95 percent of the value of the vehicle, not including imported raw materials, is reached within a period of five years. Imports of component are subject to a low (by Indian standards) import tariff of 55 percent (increased in 1986 from 40 percent), whereas the import duties on raw materials such as steel commonly exceed 100 percent. 7.40 This phased manufacturing program is likely to lead to large increases in unit costs. Several studies of automotive manufacture in other countries have estimated the increases in unit cost with increasing domestic content (Figures lA-D). Unit costs depend on the volume of output, and range from 42 percent per unit for 100 percent local content of 50,000 vehicles per year to more than 200 percent per unit for 100 percent - 110 - domestic production of 3,000 units. In general, there is a sharp rise in unit costs when domestic content is in the range of 70-100 percent. The domestic content requirement inhibits specialization and potential economies of scale (as well as foreign exchange earnings) that would be possible from intra-industry trade. - llOa - Table 3A: NORTH AMERICAN CV PRODUCTION BY MANUFACTURER GM ,... ~ AMC~Mack e.-ur w..j lMl 447 556 71 124 154 19 9 1962 525 401 102 86 157 26 14 1965 6f8 458 121 110 180 28 16 1964 682 496 149 121 179 27 14 1965 82! 607 160 109 18!1 27 20 651 170 109 18' 1966 1967 1968 818 755 925 570 802 159 191 119 121 182 159 " 26 31 20 18 20 1969 941 834 181 95 173 33 25 1970 1034 952 '42 108 227 24 28 1971 1092 787 222 85 198 22 20 1972 1071 964 352 108 2~8 25 28 1973 1399 1114 401 132 228 32 34 1974 1254 1053 387 135 194 32 36 197S 1638 849 34S 135 117 13.5 29 1976 U26 1050 526 IH 127 22 25 1977 15:?2 U67 581 187 131 27 34 1978 1798 Ul6 582 218 141 19 '7 1980• 1000 700 250 130 150 20 30 1990• 800- 600- 200· 100- 100· 20 30 1400 1000 300 150 :?00 • Rep.-sent&tift forecuu Table 3B: HEAVY CV MANUFACTURERS IN WESTERN EUROPE 1977 1977 1977 , SaJa3 ~~ Prodlactloa Proflt/lom locadou Maia Otlier ('OOOs) •milliou • tailllaM IVECO (Fiat) Italy ~y 109.0 2990 76 Fnnce ll.enalllt sroup Fraarc - 39.S 3140 -51 Daimlu·Benz Ford Germm1 Britain Rolland - 195.7 40., 4446 1209 N/A 170 GMa.dford Britain - 33.9 558 NIA MAN OAF ) a-an, ffollaJld - - 16.4 u.o 1600 400s '6 12 Seddon· l Atlduon) (lntematlonal Britain - 1.2 N/A N/A Huvesm) ERF Britain - 0.6 68 3 Fodcn Britain - 1.1 94 5 Volvo Scanla {Saab) Sweden Swedai - - 25.0 19.9 850 1150 75 40 Bramh Leyland Peugeot-Citroen Britith Britain - Spain 22.5 8.8 8844 217 2 -58.8 3 Loss /Chrysler) 1. Productionowr 3..5 tonnn G\"W ~.Sal•• 4. 19;3 ind profits for totol commer zo 18 19 9+ tonnes• Tractor 44 4 38 5 .... 7 Crane ltllck 3 2 2 Special purpose wl.Jd~ 1 I I 2:<'7 2772 3035 I. C...-dmved van f-~~~_.., .... ~~~+--~~~--~~----.~~~- I 30 40 50 60 70 80 90 100 DOMESTIC CONTE111• (PERCENT) • ZR' •$'1;i ~ Jlttl • Actual production, April 1966 (5,700 vehicles ptr }ear 1. Source: Figure lA, lC and lD, J. Baranson, Automotive Industries in Develop- ing Countries, World Bank Staff Occasional Paper, World Bank, Washington, D.C., 1969; and Figure lB, Government of Australia, Australian Industries Assistance Commission, Passenger Motor Vehicles and Components - Post-1984 Assistance Arrangements, Camberra, 1981. - llOc - Figure lB: COMPETITIVENESS OP LOCALLY SOURCED CONTENT POR A TOYOTA COROLLA MOD~L Australian Price as a Proportion Of a World Price 8 ••••••••••••••••••••••••••••••••••••••••••••••••••••• A._,...______,...,...,...,......,________,...._...... 50 c 100% Local Cont1nt - llOd - Figure lC: MANl'FAC'Tt'RJ?\Ci rosn l!'\ LATP' AMERICA A~ A Fl'NCTJON OF OOMESTJC C'OSTE!liT, JA!'ffAR\' 19to~ 2eo---~ - -- --------~--...,.-------~- I 240~·~-------~--~----~--~~..__..---~ I I j 220.--~---1--~----------+----~--..--~---. ~ ; I 200--~---+~----1------JH---~-+------. a § ,,,, NEW ZEAL.ANO PltOJ€CTlO I .n.n..Y. IM7 ' ! 160 .......~------------+-------~----_,...~'--~ , .,, ,,"'°"-aRAZIL . ._._ OONtSTIC CONTENT !PERCENT\ - llOe - Figure lD: COST INCFEASE AS A FUNCTION OF LOCAL CONTENT 200 / 200 I I I I I I I I - ~ ~ u H Q) 150 I I I I I I I 1 I 144 ~ I I '-' I I Qi Cl) I I QI I I 41 I'JO I I "" () i:: I I H I I ~ I I 85 ti) 0 I I I u I / I / / I ii / I 4'~-1 / / . ~ ~ .~:,_.,,,"' ; / / sooo_..--:1~ .~·...-" // ..- iz~OO ~·- .- ,i, -- - - '1 P .R• * ,,.., ,.~/ / ,- / / / 42 I / / ,,, /., ,/" I ,/ --- ii 25,000 • ~~~~~-~ __ ,,_ n ~o.ooo V .~ R * ..,..,,. .,,....- , "" ~ """" ,,,..,. _.., ... _,,, ........ ~ .... ------ ..... ----- • • __ - ,,,-:;- 0 .-"'1~--r~-r---r--.,.....--.,......~.,.--~--~----.--- 10 20 so 40 50 60 70 80 90 100 Local Content (percent) *Vehicles Per Year Source: Asociacion de Fabricas de Automotores, !nforme Economico, Buenos Aires, 1969. - Ill - INDIA - INDUSTRIAL REGULATORY POLICY STUDY VIII. Subsector Review--Telecommunications Equiement A. BACKGROUND 8.01 Telecommunications services have been supplied in India for the last 130 years, and by 1947 there was a fairly effective, albeit very limited, telephone/telegraph system, operated by private companies under license. The system was concentrated in the larger cities of Bombay, Calcutta. Madras and Delhi, and only 659 locations in India had telephone services. 8.02 The system expanded considerably since independence. Between 1948 and 1985, total installed telephone exchange capacity increased from O.l to 3.3 million lines, connected direct exchange lines from 0.83 to 2.9 millio~ and telephone sets from 0.115 to 3.8 million. Yet, in spite of an average growth rate on the order of 10 percent in these indicators (which has decelerated to 7.5 percent since 1980), India has fallen behind other developing countries in expanding its telecommunication resources. Although average density increased fifteen-fold from 0.024 connected direct exchange lines (DELs) per 100 population to 0.36 as of March 1985, this level is still among the lowest in Asia, and is well below that of Malaysia (6.3), the Philippines (1.2), Thailand (1.1), Sri Lanka (0.7) and Pakistan (0.6). 8.03 Moreover, there are major shortcomings in the network itself. Between 1980 and 1985, the average waiting time for a subscriber to be connected increased from 2.5 to 3.5 years, while the telephone waiting list expanded from 334,000 to 829,000, given network growth of 7.5 percent a year compared with demand growth of 10 percent a year over the period.I/ Further, the telecommunications system basically offers only standard voice and some very limited data communication facilities, and none of the other services normally available in countries at a similar stage of industrial development (Telebras of Brazil, for instance, delivers 50 different types of telecommunications services, ranging from dial-in general information about the time, weather, etc., to teleconferencing and other business- oriented services). Most firms list the difficulty with communications as a major factor restricting their growth, both domestically and especially in the export market. l/ As of March 1985, 78 percent of expressed demand (working lines plus registered applicants) for telephones was satisfied, down from 85 percent in 1980. These figures understate actual demand, however, because many people are reluctant to pay the advance deposit (of rupees (Rs) 5,000-8,000 with remote prospects for a connection. Moreover, no waiting lists are kept in areas without·service. DOT's Economic Research Unit projects that the waiting list will double by 1990 to 1.6 million. - 112 - 8.04 Finally, the quality of service is also inadequate. The percent- age failure of local interchange calls attributable to equipment malfunc- tion or system congestion during FY85 ranged from 8 percent in Bombay to 42 percent in Calcutta; the rate for.well-operated networks is only 2-3 percent. The failure rate for subscriber trunk dialing averaged about 50 percent nationwide, compared to international norms of 25 percent. Poor quality has been caused, inter elia, by poor maintenance, lack of spare parts and tools, shortages of long distance facilities and ineffective traffic management. A major factor, however, also appears to be the supply of substandard equipment by domestic manufacturers.~ 8.05 Under India's Industrial Policy Resolution of 1956, the manufac- ture of telecommunications equipment was reserved for the public sector. Three enterprises were given exclusive licenses to produce equipment for the DOT: Indian Telephone Industries Ltd. (!TI), established in 1948, which produces telephone exchanges, transmission equipment and telephone instruments; Hindustan Cables Ltd. (HCL); and Hindustan Teleprinters Ltd. (HTL). These three entities have grown rapidly over the last three decades and have accumulated substantial production and technological capabili- ties. IT! manufactures a wide range of telecommunications products at five locations: Strowger and crossbar exchange equipment at Bangalore and Rae Bareli; telephone instruments at Bangalore, Naini and Srinagar; transmis- sion equipment at Bangalore and Naini; and electronic (trunk) switching equipment at Palghat and, in the future, local switching at Mankapur. HCL manufactures large and small diameter coaxial trunk cable, small pair jelly-filled cables, plastic-insulated cables for internal wiring and copper-clad steel wiring for its overhead lines at Rupnarainpur, whereas it produced both jelly-filled and dry-core cables at its Hyderabad Plant. HTL makes electromechanical page and tape teleprinters, printer reperforators and tape transmitters and recently began assembling (under a phase manufac- turing program) electronic typewriters and teleprinters. 8.06 Assured, however, of protection from competition by the indus- trial licensing system, as well as by binding trade barriers and guarantees that all their output would be purchased by the DOT under a cost-plus pricing arrangement, none of these enterprises has faced a proper structure of incentives to minimize costs or meet DOT's quality standards and deli- very schedules. In addition, efforts to introduce new products into India have been constrained by a combination of poor R&D management, which has failed to respond effectively to production requirements, and by the pursuit (not necessarily by the most effective means) of technological ~ In FY85, problems with subscriber equipment (mainly telephone sets and internal wiring) accounted for 36 percent of reported faults, telephone switching equipment for 34 percent and the external cable network for 10 percent. The main steps the Department of Telecommunications (DOT) has taken to improve quality have involved widespread replacement of faulty dials on telephones manufactured by IT!, a domestic public enterprise; special efforts to improve the maintenance of older exchanges; rewiring of cabinets and pillars; and introduction of tllOre advanced equipment, such as digital switching and jelly-filled cables. - 113 - self-reliance. In a number of cases. product development could have been accelerated and greater technological learning and maturity achieved through an effective transfer of technologies from outside the country, had the system not led to attempts by domestic manufacturers to recreate estab- lished designs on their own (see para 13). 8.07 As a result, ITI is one of the few remaining producers in the world of first- and second-generation (respectively, step-by-step and crossbar) switching equipment, whereas fourth-generation digital electronic technology dominates world sales. For its part, HCL is behind most firms in other countries in switching from paper-insulated dry-core cables to plastic-insulated jelly-filled cables (which are more water-resistant), and it has, until recently, been unable consistently to meet DOT's technical specifications for jelly-filled cables (such as those involving water pene- tration tests). HTL's main product line is an outmoded, though reliable, electromechanical teleprinter. 8.08 From the users' standpoint, the government's policies and regula- tion have had substantial static and dynamic costs. The major effects, however, of the shortcomings have been to hold back the Indian telecommuni- cations system and to constrain the ability of producers to serve govern- ment, business and personal needs eificiently.~ 8.09 Recognizing these problems, in March 1984, the government ini- tiated a major shift in policy whereby the private sector was allowed to begin manufacturing subscriber terminal equipment (e.g., PABXs, telephone instruments, teleprinters, etc.) and peripherals for data communications systems. Private firms could also acquire up to 49 percent equity partici- pation in joint ventures tii1ith central/state governments to manufacture switching and transmission equipment. Further, the government eased entry and expansion into the telecommunications cable segment of the industry. The GOI's target is for 20 percent of telecommunications equipment to be made by private and joint-sector firms by 1990 compared with 4 percent in 1985. 8.10 The impact of the policy change is evidenced by the relatively large number of firms that have expressed an interest or have committed themselves to enter production and by the execution of expansion plans by private and joint-sector firms. As a result of the lowered entry barriers, there have also been signs of a tentative change in strategic behavior by the public sector enterprises in terms of an increase in market-oriented 3/ For a discussion of the characteristics and critical problems of the network, see India: Ninth Telecommunications Project, World Bank Staff Appraisal Report, World Bank, Washington, D.C., July 1986; and c. Brundenius and B. Goransson, "The Quest for Technological Self-Reliance - The Case of Telecomm11nications in India," processed, February 1985. - 114 - technological activities.4/ There has also been greater concern over moder~izing product lines-in those segments expected to be targeted by entran~s. which will not be as saddled as are incumbents with outdated designs and equipmer.t and overmanned production lines. At the same time, the public enterprises have elaborated plans to move into areas character- ized by high financial or technological barriers to entry that sf ford greater protection from private sector competition. Finally, they have placed greater emphasis on quality and delivery schedules. The latter has in no small measure also been occasioned by a progressive tightening of procur.ement guidelines and practices by the DOT, particularly with respect to the 11)0r.~ critical items such as cables. 8.11 Although the new policy and resultant entry or potential entry of new firms into certain segments of the industry have led to perceptible changes in the structure of the subsector and in firm behavior, there has been little indication that companies will respond by shifting their supplies or substantially improving their performance. Moreover, not all the planned supply response implicit in the new policy will necessarily evoke competition that improves firms' behavior. Some expansion projects of the public firms may be so large as effectively to pre-empt new entrants from competing. Such might be the case with HCL, to the extent that it m~ves into the production of small pair cables, and with ITI in the area of telephone instruments. 8, 12 The GO! might want to reassess the timing of t;1e implementation oi these projects in order to avoid crowding out the private sector and to eli~it a rositive response to the new policy. From this perspective, the DOT'might consider signaling the subsector that it will not award preferen- iial. treatment in procurement to a single firm, except in the award of small development orders for new entrants. B. THE IMPACT OF PAST POLICIES 8.13 India has, as noted, been producing telecommunications equipment in the public sector for the last 30 years. During this period, many of the policies have intended to reserve the "commanding heights" of the economy for public enterprises, so as to attain a large measure of self- reliance, the principal means being extensive import-substitution. Other government objectives have been the deconcentration of economi~ activity toward "backwar.i areas" and the expansion of employment. Most of these 4/ In the case of IT!, for instance, the DOT has traditionally paid for R&D projects on a cost-plus basis. Together with the fact that 90 percent of its output was sold to the DOT, marketing was little more than a mere sales activity, at ITI, and efforts at new product identification, feasibility studies and market assessment were generally skipped. The more competitive environment has led IT! to ' consider amortizing the R&D expenses of its products, and to establish a marketing department, with a far greater say in the company's R&D resource allocation. - 115 - strategies have had adverse effects on the performance of the equipment subsector. Protection from competition was assured by the public sector reservation provision and trade restrictions that precluded imports, as well as by procurement procedures that virtually excluded other entrants. Such unrestricted protection has led to a pattern of managerial behavior characterized by few systematic efforts at cost-minimization and improve- ment in quality. Ultimately, it has jeopa~dized a major objective of the GO!: the development of a telecommunications network capable of delivering services of acceptable quality to large numbers of people. 8.14 The relationship between subsector policies and performance is discussed here first in terms of the consequences on plant size and the scale of operations. Although individually the producers (particularly ITI) are large, production at the plants has often been insufficient to achieve minimum efficient scales. This situation has applied, for instance, to the manufacture of private automatic exchanges, telephone instruments and copper cables. By way of illustration, HCL's old unit at Hyderabad has an installed capacity of 1.27 million conductor kilometers (ckm), Rupnarainpur's 1.6 million ckm, both of which are below the minimum efficient scales of production, which are on the order of 2-3 million ckm. For telephone instruments, whereas ITI's 1983-84 Bangalore and Naini output of 336,000 and 211,000 instruments was within efficiency bounds (200-500,000 units a year), Srinagar's 85,000 instruments was not. 8.15 It must be concluded that the policy of self-reliance in the context of India's limited market has led to excessive diversity of output by each producer. This policy, combined with the fact that some plants have been regionally dispersed to assist "backward" regions, have led to losses in economies of scale (and to agglomeration) of production activities. 8.16 In addition to their impact on plant size and the ext~nt of specialization, government policies have affected other elements of the structure of the subsector. Insurmountable entry barriers, a foremost being the reservation provision of the Industrial Policy Resolution of 1956, until recently precluded the entry of new producers. Finally, the DOT's procurement system has reinforced the barriers to entry by allowing only those orders that public enterprises have been unable to fulfill to be open to international and more recently to local bidding. It is useful to examine in greater detail the impact on the subsector's behavior and performance of the policy-induced barriers to competition. 8.17 Management of production facilities at both ITI and HCL has been quite poor. In-plant environmental conditions (lighting, air cleanliness, dust particle control and noise) have been unsuitable for the production of goods with a low fault-tolerance such as telephone instruments (there are exceptions to this problem, such as ITI's new Mankapur plant). In add!- tion, large numbers of individuals have been working in small spaces, and there have been few transit areas. The bulk of production has taken place at older plants that have not been properly maintained, and the state of the machinery at various plants has varied tremendously. It appears to have been normal practice (although not necessarily company policy) to run - 116 - machines as long as they will go on routine, first-order maintenance, leaving the second- and third-level checks until the machine has a major breakdown.~ 8.18 The weakest part of ITI's and HCL's operations has been produc- tion engineering, especially with respect to tooling and manufacturing methods. The basic manufacturing processes at lTI have remained nearly unchanged since the beginning of its foreign collaboration agreements, although production engineering management could have led to improved operational efficiency through the increased use of mechanical components with low-cost automation, roll feed for presses, centered carbide dies for the high volume, repetitive press work and modified molding dies with hot runners for plastic parts. Similarly, in the area of assembly, jigs and fixtures could have been used along with pneumatic tools with presettings for fastening. 8.19 In terms of product design, almost all of ITI's current produc- tion is or will soon be obsolete by world standards, including the electro- mechanical automatic telephone switches, the transmission (multiplexing) equipment not required in a digital electronic network, the rotary tele- phones, etc. This problem also affects both HCL, which is years behind schedule in converting from paper-insulated dry-core cables to plastic- insulated jelly-filled cables, and HTL, which produces an outdated mechan- ical teleprinter of limited capabilities designed 25 years ago. In sum, domestic firms have not kept pace with changes in production methods or designs. 8.20 The poor technological conduct of the producers has been caused by a multiplicity of factors. First, until recently, R&D projects have been defined without due consideration for production needs and market possibilities. Moreover, they would run for very long periods and suffer from overdesign. Then, by the time the necessary design and engineering information had been translated into production data, the product was 5/ For a more detailed discussion of the management of production facilities as well as aspects of production engineering and technological development in the subsector, see India: The Telecommunications Equipment Manufacturing Industry, An Interim Report, The World Bank, Washington, D.C., November 15, 1984, and the Consultant report on ITI, HTL and HCL prepared for the Industrial Technology Developm~nt mission of October 1985. - 117 - outdated.6/ This division between R&D and production plans and activities is accentuated by the fact that, again until recently, ther~ was little in the market or policy environment to penalize management for failing to utilize R&D resources properly, to abide by schedule commitments or to respond to customer requirements. An additional problem is interference by government agencies in the technological assessment process, which has also led to delays in product development •.?_/ 8.21 More generally, in choosing technologies and their development, the GOI's strategy has incorporated a variety of sometimes conflicting objectives, such as maximizing the degree of technological self-reliance, through means that have often proven ineffective and that have eventually led to even larger degrees of technological d~pendence. Moreover, insuffi- cient attention has been paid to the fact that R&D in certain fast-changing technologies that involve large fixed and sunk ~~sts (such as switching equipment) have substantial economies of scale, as well as resource, capa- bility and effort thresholds. If a firm (or a country) is unable to amass those resources and to cross the thresholds, there is little to be gained from (re)creating a technology. A strategy of self-reliance that also involves major efforts at transferring, adapting and improving existing technologies can provide significant learning opportunities and can be conducive to a greater degree of technological maturity. However, those 6/ The 4 GHz channel system ran 10 years from conception to field trials. The PCM took 7 years to be fully indigenized, even though its specifications were well-known internationally and had been standardized for 15 years. Another example is the Indian crossbar project (ICP). The crossbar electromechanical switching system was introduced into India in 1964 through a technology transfer agreement between !TI and ITT (Belgium) that was to last 7 years and included the establishment of a factory in Bangalore with a capacity of 100,000 lines a year. The original design proved quite inadequate to Indian environmental conditions and usage patterns; however, not until 1971 was a task force set up under TRC to suggest solutions to the problems with the Pentaconta system, and then three more years elapsed before the project was initiated. Four years later, despite 200 engineers from TRC and IT! and over 40 changes in the electromechanical systen at a cost of Rs 8 crores, the project was declared finished from a design standpoint. However, it then took three more years and an additional collaboration agreement with the Belgian ITT for the system to enter production at the Rae Bareli plant. It thus took a total of 17 years ... to improve and upgrade a foreign design that was initially inappropriate for India and to bring the project's output into production. During these years, electronic switching systems were developed and marketed, rendering the electromeehanical crossbar system obsolete in any event. 7J In an attempt to spur indigenous development, the GOI, for example, blocked an international technology transfer agreement for the production of translation bays for UHF radio stations. The result was a production backlog that prevented the full use of the imported UHF equipment. - 118 - benefits accrue only to the extent that resources commensurate with the complexity of the technology and the speed at which it is changing are economically available for the necessary adaptation and improvement, which must be completed in timely fashion.8/ 8.22 Certain policies have had a direct bearing on management behavior and firms' performance, apart from their impact on the structure of the subsector. In particular, DOT's procurement policies, as noted, have not established a correct structure of incentives for cost-minimization, improvement in product quality and adequate technical performance. Toge- ther with the barriers to entry, these policies allowed the management of public enterprises to operate conservatively, unchecked by either actual or the threat of competition. In the_ case of !TI, for non-stabilized or prototype equipment manufactured in pilot production, prices have been determined by actual costs plus a 10-percent profit margin; for stabilized items, the DOT has based the prices on prorated production overhead costs and standard costs for materials and labor. Departures in actual costs relative to the standards should lead to reduced profit margins. In practice, however, total year-end variances have not been completely charged against revenue. Instead, they have been the basis for revising the standard costs, a practice that mutes any incentive for cost-minimiza- tion. In addition, the bonus system has linked incentive payments for managers and operators to achievement of the target levels of production, and not to cost control. The same cost-plus features basically have applied to both HCL and HTL. Unlike with ITI, however, the norms for materials consumption and machine utilization for HCL have not been tech- nologically fixed, even in principle. The 1984 standards, for example, were calculated as an average of actual costs for the three years between 1979 and 1982. Note, however, that even those rules have not been fully enforced, inasmuch as the DOT's cost check unit has not been verifying that the costs have been consonant with the pricing agreement. 8.23 Another dimension of procurement has been the method of payment. The penalties for delivering non-sequential supplies (that is, not in the order of their logistical packaging) have been insufficient, particularly given the magnitude of ITI's delivery problems. Nor has the DOT adopted a method of payment that provides incentives for timely supplies on estab- lished penalties for liquidated damages. As a consequence, ITI's delive- ries of telephone exchanges, interworking and transmission equipment have of ten not accorded with the logistical packing sequences that are required for efficient installation and commissioning. On other occasions, ITI's output has simply been insufficient to meet DOT'~ requirements. 8.24 It has been shown that awareness on the pa~t of management and workers is very important to proper quality control. At ITI and HCL, 8/ In the ICP, both tTI and TRC incurred substantial RliD costs for a number of years, at the end of which the design was outdated. Similarly, by the time the electronic analogic switching system was complete, the technology frontier had moved to the superior digital switching. In both cases, there were simply not enough resources and accumulated capability to keep up. - 119 - neither group has been pushed by appropriate behavioral incentives and constraints to undertake quality assurance. A case in point has been the inability of HCL to meet DOT's specifications for jelly-filled cables.9/ Although DOT's specifications are quite stringent, they are not unattain- able, as some of HCL's competitors (such as Finolex) have shown recently. 8.25 From 1971-72 to the end of 1984, the DOT relaxed the standard for HCL several times, with only minor price penalties. Comparison of the timing of even these minor penalties and actions taken by ~Ct's management indicate that only after DOT has taken forceful actions have its concerns been addressed. Only when the DOT initiated price reductions in earnest in 1982 and implemented zero profi~ provisions in November 1983, did HCL's management undertake concrete steps to improve cable quality. The DOT also set a March !986 deadline for taking substandard cables, an additional measure that forced HCL to focus on the high rates of rejection. 8.26 It is clear from this review that past policies have had a sub- stantial impact on the technological and economic behavior of producers. Largely policy-determined entry barriers have led to and sustained markets that until recently were both non-contestable and uncompetitive. The nature of these markets has adversely influenced the conduct of suppliers. In combination with DOT's procurement guidelines and practices, they have led to substandard goods, to prices that incorporate large and avoidable inefficiencies and to deliveries that are untimely and non-sequential. Regulatory policies have also influenced the size of individual plants and the production runs for specific products. Other policies have produced a pattern of geographic dispersion of plants that has precluded economies of scale and agglomeration. In an attempt to achieve import substitution across-the-board, still other policies have stimulated the production of too wide a variety of goods; they have constrained horizontal specializa- tion and led to excessive vertical integration. Finally, the lack of selectivity and scales in R&D has caused large delays in product upgrading and development. 8.27 Many of the regulatory policies have had adverse consequences on performance. Although the growth rates of sales have been high (16 percent a year for HCL and 12.9 percent for IT! between 1976/77-1982/83), produc- tion has been quite inefficient. First, the prices for some domestically produced equipment have been above international levels (CIF prices before • 9/ The following statistics explain the dimension of the problem of poor quality. Between January and May 1985, all samples of 100-pair jelly-f ill~d cables produced at Hyderabad were rejected in water penetration tests, and the rate of rejection for samples of 50-pair cables was in the range of 65-100 percent. The results at the Ruprainampur plant were 40 percent for the 50-pair and 25 percent for the 20-pair cables. - 120 - custom duties). More important, product quality for most products (partic- ularly terminal equipment and cables) has been substantially inferior to what is available in the international market •.~/ 8.28 Performance could be substantially improved by better management of production facilities, application of updated production engineering methods and procedures, a program of investments in balancing, debottle- necking and other areas of production engineering, and implementation of quality assurance programs. Finally, product designs have generally been outdated because of the long lead times involved in product development and insufficient efforts directed at effective technology transfer, including adaptation and upgrading of designs and techniques. Inferior designs have been adversely affecting the technical performance of equipment. C. RECENT POLICY MOVES 8.29 In March 1984, the government announced a revised industrial policy. It allows the private sector to manufacture subscriber terminal equipment (for example, PABXs, telephone instruments, teleprinters, auto- matic dialers and message recorders) and peripherals for data communica- tions systems (see para 5). The policy also permits the private sector to hold up to 49 percent of the equity in joint ventures with central/state governments for the manufacture of transmission and switching equipment; further, it encourages joint and private sector entry into the production of jelly-filled cables. Since the inception of these reforms, a substan- tial number of firms have applied for and obtained letters of intent (LOis) eventually to be converted into industrial licenses (!Ls). By product group, the following has been the response. 8.30 Copper Cables. Production of copper cables has been opened to private and joint-sector Indian firms, and since 1984, at least three producers (Finolex, Traco and Delton) have come on-stream. Recently, letters of intent and corresponding licenses were issued to six private sector firms for a total capacity of 2.75 million ckm and to nine State Industrial Development Corporations (for 5 million ckm). By the end of 10/ The entry of competing imports has been restricted by both tariffs, which average about 55 percent for finished telecommunications equipment, and DOT's specifications (which in the case of cables, for example, are quite stringent). Imports, however, could have been used to discipline local producers. The DOT might consider building inventories of some imported items (such as optical fiber cable and large-pair jelly-filled cable) to be certain that: (i) their installation does not lag because of delays in establishing reliable local production, and (ii) there is a credible threat of a market loss if local producers do not meet DOT's quality specifications and/or fall outside acceptable price boundaries. - 121 - 1987, still other producers will have set up plants or expanded their capacity (see Table 1)._::t Table 1 PRODUCTION OF JELLY-FILLED CABLES PROJECT IMPLEMENTATION BY A SAMPLE OF PRIVATE AND JOINT-SECTOR FIRMS (as of April 1986) Licensed Capacity Production Application Technology (million ckm) Start-up for Expansion Supplier Punjab Cab;.. ,s o.~ Mid-1987 2 mil. ckm AEG Kabel Germany, FR Finolex Cables a/ o.s End 1986 1 mil. ckm M. Essex us UPCOM Cables 0.5 End 1986 1 mil. ckm CEAT Italy Traco Cables b/ o.s Mid-1987 General Cables us Sterilite Cables 0.5 Mid-1987 M. Essex Vidhya Telelinks o.s ••• Ericsson Cables Delton Cables b/ 0.25 ••• • •• Tidco 0.5 ••• • •• Source: DOT and GOI. a/ For Finolex, "end of 1986" refers to full utilization of licensed capacity. b/ Both Traco and Delton Cables already produce cable • • 8.31 Entrants have, however, been restricted to capacities of about 0.5 million ckm each, well below the minimum efficient scale of produc- tion. This limit constrains their ability to produce a wider range of cable products, particularly those in the higher pair range at competitive prices. In fact, recent studies demonstrate that minimum efficient scales are now on the order of 2-3 million ckm, and the Directorate General Tech- nological Development itself noted informally that for a product mix that includes cables larger than 400-pair, capacity should be at least in the range of 1-1.5 million ckm. Even though firms can perhaps be financially 1J..I In case of expansions, letters of intent are being granted once 60 percent of licensed capacity is built. Significant expansion takes an average of two years from the moment the letter of intent is issued. - 122 - viable at o.s million ckm, that capacity limit will restrict them to small-pair cables and restrict the manufacture of larger pairs to HCL. Recently the DOT reversed its policy of restricting larger pair acquisi- tions to HCL. Until then, the procurement and licensing policies for copper cables had assymetrically segmented the market, with HCL dominating the larger pair segment. It will take some time for a more competitive market to develop in larger pair cables. 8.32 Competition in the area of smaller pairs could, on the other hand, become constrained if HCL were to decide to expand its output and to cross-subsidize its sales with the monopoly profits from its larger pair operations. The way out of this conundrum is both to raise the capacity limits to 2-3 million ckm (or to delicense the product altogether) and to move all copper cable procurement toward competitive bidding, once dev6lop- ment orders allow new producers to meet the DOT's specifications for larger pair cables. 8.33 The configuration of the market in the longer rtm is an open question. If changes in capacity and procurement policies effectively promote e~try, by the late 1980s the market should be quite competitive. HCL is expected to have a 7-8 million ckm capacity by 1990 including approved projects and those under implementation. Private and joint-sector firms may together have 6-7 million ckm. Given that total DOT require- ments, particularly after the cuts in the original Seventh Plan estimates, will be substantially below full capacity output, producers will have to market their output competitively in order to survive. 8e34 Optical Fiber Cables. Both MPEDC (Madhya Pradesh State Electron- ics Development Corporation) and HCL intend to enter this market, which so far has not been open to the private sector. HCL has proposed establishing a factory that could supply all the domestic demand for both optical fiber cable and for line systems. Although HCL's choice of production technology for the fibers is one that does not optimally require scales much beyond current demand projections, historical experience has shown that problems of managerial or x-inefficiency may plague HCL, particularly where it is the sole supplier, with DOT as the captive customer. 8.35 In this context, DOT procurement policies will be a determintng force behind HCL's behavior. To the extent that the DOT resorts to dit~ct purchases from HCL (and other enterprises), they should be based on commer- • cial contracts that stipulate technical specifications (including minimum quality parameters) and delivery schedules, with penalties for failing to meet them. Competitiveness in optic fibers depends, however, not only on an adequate structure of incentives, but also on the volume and efficiency with which R&D resources, p~oduction engineering, quality control and assurance are used. In the face of a rapidly moving technology frontier and enlarged resour~e requirements, it will be difficult for HCL and other producers to become and remain competitive. 8.36 Terminal Equipment. This category of product has been declared open to private and joint sector firms subject to centralized procurement of technology. In the case of telephone instruments, the new policy has - 123 - led to 85 applications, with 35 letters of intent issued, of which 15 are expected to be converted into licenses. If those licenses are fully utilized, total production capacity WQuld be in the range of 3-7.5 million units a year, far exceeding the projected demand of 0.6 million instruments a year in 1990 (after cuts in Seventh Plan projections). ITI, tmtil now the sole holder of !Ls for this product, has licensed capacity for 850,000 instruments and installed capacity for 525,000. It produced 550,000 units in 1984-85 and recently concluded two technology transfer agreements with FACE Standard of Italy for the manufacture of 1 million instruments a year. 8.37 The large number of new entries for the production of telephone instruments and some excess capacity in the future may lead to a more competitive environment. Although licensed capacity is consistent with the minimum efficient scales, given conditions in India of 200-500,000 units a year, the absence of adequate exit policies might lead to capacity fragmen- tation. Improvements in the overall performance of the segment will there- fore depend on both freer entry and exit. Finally, performance will also be affected by the behavior of I~ If it moves to pre-empt entry by quickly implementing its expansion and modernization plans, on the assump- tion that the DOT will be awarding it preferential treatment for procure- ment, ITI might continue to monopolize the market, to the detriment of end users.!:!:_! 8.38 Three telephone instrument designs that the DOE ehortlisted and that it was testing locally were selected for local production to allow standard.ization of components. Most firms that were given letters of intent chose one design, to be obtained through a lateral transfer of technology. The process of centralized procurement of technology and its diffusion through lateral transfer could, of course, constitute a signifi- cant barrier to entry if the chosen licensee tacitly blocks the disclosure of the critical technical information to its (potential) competitors. It is thus an open question to what extent (capacity) licensed producers of instruments will have access to the chosen designs on a non-discriminatory basis.13/ 12/ That the ITI has long perceived its market as captive, even though it is organized under the form of a bilateral monopoly, is an indication of the symbiotic relationship between ITI and DOT. 11.I The rationale for centralized procurement of technology is the notion that by restricting the number of technology agreements, products and parts will be standardized. Economies of scale can be realized, and the range of equipment to be maintained by the DOT could be lessened. There are, however, other factors that argue against a centralized approach. In particular, it reduces the options available to . entrepreneur& in establishing technological links with foreign firms, and thus it constitutes an additional entry barrier for both local and foreign firms. The DOT could, while pursuing standardization as a way to simplify maintenance, simply provide type approvals for the input jack in subscriber equipment, and thereby devolve itself of maintenance responsibility. - 124 - 8.39 PABX equipment provides another example of the positive response to the new liberalized policy, although new capacity has yet to material- ize. The GOI has issued letters of intent for the production of 900,000 lines to 18 producers, 13 of which applied after the new policy was announced. A total of eight producers are expected to convert their letters of intent into licenses; two of them have already been recommended for conversion. In addition, 48 producers have taken up technology licenses with the Center for the Development of Telematics (COOT) for its recently developed PABX, and letters of intent were expected to be issued by Mey 1986.14/ Thus, aggregate output may far surpass the total demand for PABX (andi>AX), which is expected to reach 180,000 lines a year by 1990. This notional excess supply will constitute a strong incentive for efficient management of both entrants and incumbents. 8.40 On the other hand, the fact that licensed capacity for each producer is on the order of 50,000 lines opens the possibility that there may be excessive fragmentation.15/ In practice, according to DOE offi- cials, capacity licensing will constitute a major barrier to neither entry or expansion, and it has been suggested that there will be no upper limit on capacity. The absence of policy-induced barriers to entry and expansion creates an environment where achievement of optimal production scales is quite feasible, particularly if producers entering the market are not permanently protected from domestic or international competition and are permitted to exit if necessary. 8.41 The new policy has also drawn a positive response from producers of terminal equipment. For rural exchanges (RAXs), the GO! has licensed six joint-sector firms (in addition to ITI}, each for a capacity of 50,000 lines a year. The technology is being centrally procured by a special committee made up of TRC, IT!, the DOE and some licensees. The committee has floated tenders, and 10 potential foreign collaborators have applied, of which two or three will be chosen.16/ In the case of electronic tele- printers, RTL is establishing a plant'"'"'With an annual installed capacity of 8,000 units a year. In additien, DCM and Hindustan Communications have J.!!..I As in the case of telephone instruments, PABX technology is subject to centralized procurement; this practice may constitute a binding barrier to entry for capacity licensed firms wishing to use non-indigenous designs (although 3 foreign collaborators have already been selected for the purpose of technology transfer). This constraint is a distinct possibility in view of the existence of two national PABX designs. One was developed jointly by ITI and TRC (DOT's Telecom Research Center), and ITI is to add it to its present c~pacity of 40,000 PABX lines of the old Strowger variety. The second is CDOT's design. ]J_/ It is estimated that efficient scales are closer to 100,000 lines a year, whereas in the US major plants make as many as 2-3 times. 12_/ In addition, ITI has developed a 128-line RAX that has been in field trial for the last two years, and CDOT will be bringing out a more recent design. - 125 - 8.42 Transmission, Multiplexing and Switching Equipment. The new policy allows joint-sector firms with majority public sector ownership to produce this equipment. However, private sector participation, even as restricted to a limit of 49 percent, has yet to materialize. In the area of transmission, ITI remains the sole producer of open wire carrier and coaxial ~able systems, VHF, microwave transmission and frequency division multiplexing equipment. Together with HCL, and under a common technology transfer contract, ITI will also be involved in the production of terminal equipment for optic fiber media. Five joint-sector electronics firms have entered the UHF field, and their products are currently being tested. Because of the small production runs that tend to characterize most trans- mission equipment, fragmentation of output is not expected to raise unit costs.17 I - 8.43 The manufacture of switching equipment is still de facto reserved for ITI. In addition to its traditional Strowger and crossbar switching equipment, ITI has acquired CIT-Alcatel technology for (i) large digital exchanges for local networks and (ii) trunk exchanges for a first (Mankapur) and possibly a second (Bangalore) plant of 500,000 lines each. It is estimated that those plant-specific scales are not adequate for the planned levels of vertical integration, although Bangalore is expected to be less integrated than Mankapur, whicht for locational reasons, has sub- contracted few of its activities. 8.44 For a phased manufacturing program (PMP) of five years duration, ITI would need to take full advantage of available scale economies and might want to concentrate on the manufacture of components at one plant (Mankapur) and assembly operations at another (Bangalore). The plant- specific capacities would be 1 million lines (in components and equipment, respectively)t levels that increase the probability that the backward integration required by the PMP is completed successfully. On the other hand, improvements in production efficiency may be off set not only by transportation problems, but also by the fact that ITI will again have a virtual monopoly in a captive market, a situation that in the past has led to gross x-inef ficiencies. D. SUMMARY • 8.45 The point to stress is that any firm needs certain market- generated constraints to become efficient. At the same time, firms should have the flexibility to react to market changes--by threatening to or ]:]_/ In fact, the cost advantage that firms would normally have in those engineering-intensive product areas might be partially offset by the presence of substantial economies of scale in the production of components. Insofar as most demand is derived froin a still relatively ' small consumer electronic segment, most components would have high costs, a situation that adversely affects the final price of transmission and multiplexing equipment. - 126 - actually entering new market segments, by adding capacity to their traditional lines and by leaving economically less profitab!e niches. Measures that decrease policy-induced (and other) barriers to entry, growth and exit not only will change the structural composition of the subsector, but will also influence the specific form through which firms insert thamselves into the market. By stimulating more aggressive behavior by individual producers, these measures should ultimately bring about an overall improvement in economic performance. 8.46 Table 2 summarizes the response to the new policy as expressed in the number of letters of intent and industrial licenses issued prior to and after March 1984 for both terminal as well as transmission, multiplexing and switching equipment. As noted above, there has be~n substantial interest in establishing new production lines; it has been greatest in terminal equipment, particularly telephone instruments, as well as in answering and recording machines, PABXs, facsimile transreceivers and two- way radio and allied equipment, where the barriers to entry have been lowered the most. In the are~s of transmission, multiplexing and switch- ing, c~ntinuing restrictions on the entry of private firms, technological barriers and market size are still major constraints to entry. Neverthe- less, for a number of products in these areas, the new policy has appa- rently attracted considerable attention, as in the case of multichannel radio communications equipment, satellite communications equipment, multi- plexing and antennae for communications equipment • • - 127 - Table 2 LETTERS OF I~NT (LOis) AND INDUSTRIAL LICENSES (!Ls) ISSUED PRIOR TO AND AFTER THE MARCH 1984 . POLICY l\NNOUNCEMENT (as of May 198~) LOIs t.Ois/ILS LOIs LOis/ILs Product Pre-March 1984 Post~arch 1984 A. Terminal Equipment Telephone Instruments 6 1 35 1 Telephone Answering/Recording Machines 5 Push Button Telephone Dialers 2 Cordless Telephone 4 PABX 5 1 13 1 RAX 4 1 4 Two-way Radio Comm and Allied Equip. 10 7 5 2 Facsimile Transrecei~ers 15 Pay Phones 4 Teleprinters 3 1 B. Transmission, Multiplexing and S~"itching Eguip~o,.!!l Other Types of Exchange (DTAS, ESS, Strowg.) . 4 4 l Powerline Carrier Comm. Equi?• 4 4 l Multiaccess Rural Radio Telephone Equip. 2 2 2 1 Muitichannel Radio Comm. Equip. 2 4 Satellite Comm. Equip. (SPC, Amplifiers) 1 1 4 1 Microwave Comm. Equip. 3 2 Coaxial Line Equip. 2 2 Multiplexing Equip. 6 4 4 1 Open Wire Carrier System 2 2 Ground-Based Long-Dist. Radio Collllll. Equip. 2 2 Mining Comm. Equip. l 1 VF Transmission Equip. 2 2 l Antenna for Comm. Equip. 2 1 Source: Department of Electronics. ' \ - 128 - 1': I - 129 - INDIA--INDUSTRIAL REGULATORY POLICY STUDY IX. Subsector Review--Electronics * A. BACKGROUND 9.01 The world electronics industry has displayed very rapid techno- logical development during the past 15 years. A number of new products have been introduced, and the development of process and product technology has helped reduce production costs dramatically. For example, when 64K RAM was first introduced in the late seventies, it sold for about $100, whereas the prices in 1985 were $1 to $2. Similar developments have taken place i~ the area of computers. Computing power has increased very rapidly, and the prices have dropped by a large order of magnitude. At the same time, the computing capabilities of the most powerful computer available, measured in computational speed (arithmetic operations per second), have increased from 270 in 1951 for UNIVAC I and 33,700 in 1961 for IBM 7074 to 9,100 1 000 in 1972 for CDC CYBER 176 and 800,000,000 in 1981 for CDC CYDER 205.l/ While the biggest computers have been gaining in speed, smaller machine'Shave been improving in performance per dollar. Comparison of an 8-bit single- board microcomputer, Fairchild F-8, representative of the 1970s technology, with the IBM 650, a first-generation vacuum tube processor of the mid; fifties, shows that while the two machines are roughly comparable in computing power, the modern microcomputer is smaller, cheaper and vastly more reliable. While the IBM 650 cost $200,000 in 1955, the Fairchild F-8 microcomputer with terminal cost l~ss than $1,000 in the 1970s. 9.02 In the Indian electronics industry, many segments are, however, still in the infant stage. Despite the government's stress on the strate- gic importance of electronics, this industry has lagged behind inter- national technology developments. Neither VCRs nor picture tubes for color TVs are yet produced in India.2/ In the semi-conductor industry, only one firm is producing simple chips-for pens and watches, although there are plans for 64K RAM and 16-bit microprocessors. Computers, mostly micro- computers with 8-bit microprocessors, are assembled in India with imported components and locally manufactured cabinets. However, many locally produced electronic items are of poor quality and high in price. For example, Indian producer prices for color TVs are rupees (Rs) 5,000 (US$400) and for microcomputers with an 8-bit microprocessor, Rs 30,000- . 100,000 (US$2,500-8.000). The comparable international retail prices range * For a more detailed analysis of the Indian electronics sector, see Electronics Sector Report, World Bank, forthcoming. l/ U.S. Congress, Washington, D.C., "International Competitiveness in Electronics", 1983. :!:} Three or four firms are expected to begin production of color picture tubes in 1987. - 130 - between US$150-200 and $500-1,000, respectively. In view of the generally high rate of excise duties and sales taxes in India, a comparison of Indian and international retail prices would reveal even greater price differences. 9.03 While the Indian electronics industry manufactures a wide range of products, from consumer electronics to computers and aerospace and defense materials, it has not achieved the depth or degree of price- performance characteristics that pertain in the industrial economies and some other newly industrialized economies. The following discussion outlines the impact that industrial regulatory policies have had on shaping the structure and influencing the behavior and performance of Indian electronics firms. B. REGULATORY POLICIES PRIOR TO RECENT CHANGES 9.04 Until recently, India's policy toward the electronics industry emphasized self-sufficiency, indigenous technology development, public sector production of key products in areas such as communications and defense, and small-scale production in many electronics segments open to private investment. The role of foreign technology and equity was con- strained; only in certain segments were Indian firms allowed to acquire foreign equity or technology. The industrial licensing policy emphasized the geographic dispersal of production. 9.05 Two of the six electronics segments--defense, and telecommuni- cations and broadcasting--were reserved for the public sector until recently. Electronic components, consumer electronics, control instru- mentation, industrial electronics and computers 3/ were open for develop- ment by private, puhlic and joint public-private-enterprises. However, licensing policy has in practice of ten favored public and joint sector companies over private parties. Monopoly Restrictive Practice Act (MRTP) companies were subject to the more restrictive clearance procedures normal- ly applicable to those companies. In particular, the policy for consumer electronics was weighted strongly against MRTP companies and in favor of small-scale firms. In the area of computers, large companies could not sell more than five units a year if their expenses were less than Rs 300,000, and no company could produce and sell computers worth more than Rs 20 million a year. In keeping with industry-wide delicensing parameters, investments of less than Rs 50 million (about US$4.2 million) in land, building, plant and machinery did not require a license, provided no capital goods were imported, no foreign collaborations or technology were involved and the product was not reserved for small-scale industry; larger projects required a license from the Department of Industrial Development. All electronics projects required approval from the Department of Electro- nics (DOE). In the case of capital goods and other technology imports, the clearance procedures of the Directorate General Technical Development (DGTD), including the indigenous angle clearance, which required local 1J Large computers are, however, reserved for the public sector until 1987. - 131 - . advertisement of specifications, generally favored local technology, capital goods and inputs. 9.06 The DOE has had an increasing role in helping formulate indus- trial and trade policies in the electronics subsector. For the small-scale sector, the primary responsibility for approvals lies with the Development Commission of Small Scale Industries (DCSSI). 9.07 Until recently, import policy was very restrictive, providing high levels of nominal protection for final products. Traditionally, India's import-export policy protected both intermediate inputs and end products that were manufactured in India. Imports of raw materials, compo- nents and spare parts were classified into five categories: banned, limited permissible, automatic permissible,4/ canalized and OGL. Imports of consumer electronic items were for the most part banned entirely, while most consumer electronic components and certain professional components were in the limited permissible category; traditionally, duties of 75-135 percent applied to such imports. The automatic permissible category pre- viously included many electronic raw materials and components produced indigenously but in insufficient quantities. In the area of computers, the import duty was as high as 200 percent, and only large computers priced higher than $8,000 could be imported. For certain capital goods, if the value of imports was more than Rs 2 million, the imports required a license, and the licensing procedure included the indigenous angle clear- ance. (The purpose of this procedure was to provide an opportunity for indigenous ·manufacturers to meet the requirements of the advertiser, and local manufacturers were given preference over foreign sources. If equip- ment were available locally, it was generally preferred regardless of any price differential between domestic and imported versions). 9.08 In 1983, the import duties on all electronic components were reduced from 158 percent to 75 percent. In addition, recently the duties on imports of capital goods were reduced as part of the new electronics policy. Previously, imports of computers were allowed only for "actual users." Other users could get imported compc:ers only through government- designated canalizing agencies, e.g., the Electronic Trade and Technology Development Corporation (ETTDC), Semi-conductor Complex Limited (SCL) or Computer Maintenance Corporation (CMC). 9.09 The policy regarding imports of foreign technology was very selective: they were permitted only in "sophisticated" and high priority industries. However, in the area of computers, no companies were allowed • to import technology and software. The main considerations when reviewing applications for imports of technology were the technical competence of the applicant, magnitude of the operation, economic viability of the project, availability of indigenous resources, and the applicant's future technolo- gical capabilities. The basic objective was to develop technologies suited to the needs of the country within a given timeframe and to increase the degree of self-reliance. 4/ The automatic permis~ible category was eliminated in the 1985-88 import-export policy and its items shifted to the open general license (OGL) and limited permissible categories. The duty on limited permissible electronics imports is now 75 percent. ... 132 - 9.10 Foreign investment was allowed where it was to be accompanied by a transfer of technology needed by the country or if the project were export-oriented. Prior to the new electronics policy statement of 1985, the normal ceiling for foreign investment was 40 percent of total equity capital. A higher percentage of foreign equity could be considered in priority areas if the technology was sophisticated and not av3tlable in the country or if the venture was largely export-oriented, i.e., was located in an export processing zone. Foreign share capital had to be provided in cash, with no provisions for tied imports of machinery and equipment or payments for know-how and trademarks, etc. Payment for technology and know-how could be made ·in the form of annual royalties or a lump sum or both. The rates for royalties depended on the nature of the technology and, as a general policy regarding foreign investment, were normally limited to 5 percent and a period of five years. (As a result of this limitation, royalties were frequently factored into the price of imported equipment or components). Higher royalty rates and longer durations might be considered in exceptional cases, again where the technology involved was sophisticated or a major part of the production was exported. Employment of foreign experts was generally discouraged but could be permitted, if considered necessary, and they were allowed to repatriate a part of their earnings up to defined limits. C. IMPACT ON INDUSTRIAL DEVELOPMENT 9.11 The Indian electronics industry has been ch-iracterized by the dominant role of the public sector, limited use of foreign collaboration and foreign technology, fragmented production base, low volume of foreign trade, lack of econOtliies of scale, and low quality and high prices for many products. 1. Role of the Public Sector 9.12 In the electronics industry globally, rapid technological deve- lopment for the most part has been the result of private initiative. In India, on the other hand, public involvement has been substantial, partic- ularly in lerospace and defense. Some items, such as very large scale integrated (VLSI) circuits and mainframe computers, are still reserved for production by public sector firms.Sf Aside from the communications and aerospace/defense equipment segments of the industry, the shares of public sector firms in 1981 were 9.0 percent for consumer electronics, 35.4 percent for control instrumentation and industrial electronics, 30.1 percent for computers and 28.0 percent for electronics components. The share of public sector firms is meant to increase during the period of the Seventh Five-Year Plan. The Seventh Plan indicates that the public sector will be responsible for some one-third of the total production of 5/ As noted, only until 1987 in the case of computers. - 133 - electronics during that period. In products reserved for manufacture by public sector enterprises, there has been a reduction in competition and an increase in market concentration. 2. Limited Use of Foreign Collaboration and Technology 9.13 Historically, foreign collaboration has played a minor role in the electronics industry in India relative to other newly industrializing countries. A study conducted by Ashok Parthasarathi in 1978 showed that the share of transnational corporations in electronics output dropped from 10 percent in 1972 to 2 percent in 1977. Between 1974 and 1979, there was an average of only 11 foreign collaborations in the private electronics sector each year; in most cases, equity was not involved, and where it was, generally the share of the foreign equity was less than 20 percent. Only in one case was it greater than 50 percent. In the public sector, only 15 collaborations were approved between 1974-79. The low share of foreign capital in Indian electronics output has been primarily the result of the restrictions on foreign equity imposed by the government with a view toward achieving greater self-reliance. 9.14 Recently, however, this pattern has been changing. From 1980-83, a total of 53 public sector and 87 private sector foreign collaborations were approved. In 1985 alone, under the liberalized electronics policy, . 128 foreign collaborations were approved. 9.15 It is interesting to look at India's policy relative to that of other developing countries. In the Republic of Korea (South Korea), although foreign firms and joint ventures have played a minor role in consumer electronics, their share in the total production of components and industrial electronics is significant: 57 percent for components and 39 percent for industrial electronics and computers in 1983. Brazil, on the other hand, has followed a selective strategy regarding foreign equity and imports of foreign technology. Since it began a program to develop a domestic electronics industry, it has allowed foreign firms to continue to operate in Brazil but has restricted them to segments of the market not reserved for national firms. Those reserved areas involved large and rapidly growing segments in which production using older generations of technology was possible, such as· data processing equipment and microcom- puters. Foreign firms continued to dominate in the field of large machines and those requiring the newest generation of technology. 9.16 In 1974, India required that foreign firms reduce their equity position to a minority share, and most foreign firms (including IBM) withdrew from the market rather than take that step, which would have required a loss of control over proprietary technology. Now the import of technology is regulated through the licensing of technical and equity collaborations with foreign partners and imports of technology and machin- ery. Technology imports have been limited in electronics and from 1981-84, approvals of foreign collaboration for electronics averaged 7 percent of - 134 - total foreign collaborations, while the import of capital goods for elec- tronics averaged 1 percent of the total. One reason is the technological conservatism of many of the public R&D laboratories aru, institutes that are responsible for indigenous technology development, since these agencies are often consulted in clearing import applications. Often the R&D capability of public laboratories is ded~cated to developing indigenous technology, with little focus on assimilating foreign technology. Indian technology would be enhanced if the public R&D capability were also used to improve the capacity of Indian firms to absorb imported technology. 9.17 In Brazil, national firms have simultaneously followed three types ot strategies for acquiring technology. Some have imported equipment and reverse engineered it to produce it domestically, sometimes incorporat- ing design modifications to suit local conditions. Others have licensed the technology from foreign firms. Still others (particularly in the public sector) have relied on their own designs and have developed products indigenously. All three strategies have been permitted and encouraged. In India, on the other hand, the government has evinced a strong policy preference for indigenous technology development. 9.18 In general, foreign collaboration in the electronics sector should be structured to make maximum use of the advantages offered by internatio~al firms. Successful use of foreign collaboration depends, by and large, on two conditions. One is the degree of protection for domestic production. If domestic production is heavily protected, foreign companies tend to exploit th~ protected domestic market. In this case~ the prices charged may exceed international prices by several times, and the techno- logy employed in a joint venture is of ten second-best--adequate to compete with indigenous technology, but not the most modern or internationally competitive. The second consideration is the speed with which foreign technology can be assimilated and absorbed. If this process takes a long time, foreign partners do not feel compelled to bring in newly developed technology. 9.19 It is also very important that foreign collaboration agreements require that the foreign partners introduce new technologies to domestic manufacturers. Japanese R&D capability, which today plays a leading role in world technology development, started with the assimilation of foreign technology. To discipline the behavior of foreign firms, they can be constrained contractually to sell pr9ducts equivalent in price and perform- ance to international products in the local market, with prices held to within some margin of the international price. For example, the Foreign Investment Commission of Mexico recently concluded an arrangement with IBM whereby the prices charged in Mexico could not exceed international prices by more than 15 percent, and IBM is to introduce its most advanced products into Mexico within six months of their commercial production in the us. In Brazil, foreign firms are required to export an amount sufficient to cover their component import requirements. 9.20 Foreign collaboration is an important source of technology trans- fer. However, foreign firms are generally reluctant to part with their technology unless they have an equity stake in production. Another - 135 - advantage· of foreign collaboration is the worldwide marketing network of multinational corporations, which local firms with foreign equity can take advantage of. For example, Taylor Company, which manufactures conversion instruments in India, sells one-fourth of the instruments producdd to its foreign partner. The foreign partner then distributes the imported instru- ments to its branch companies all around the world. There is also a learning effect: the management skill and working attitude seen in the joint venture partner can spill over into domestic firms and serve as a reference point for local management. 3. Fragmented Production Base and Limited Foreign Trade 9.21 The electronics industry in India is not internationally special- ized, and India produces a range of products in most electronics segments. Whereas in India consumer and component electronics together constitute some 45 percent of total electronics production, in South Korea both sectors constitute more than 80 percent. As a small portion of domestic production is exported and domestic demand is largely met by local prodr.c- tion, the ratio of imports to domestic demand and the ratio of exports to domestic production are very low in India. As Table 1 shows, these ra~ios vary between 7-8 percent in India, whereas they are more than 50 perc~nt in South Korea. Moreover, in India, the ratios are particularly high for semi-conductor chips and integrated circuits--82 percent and 98 p~ccent, respectively, of the total production of semi-conductor chips and integrated circuits are exported--while 64 percent and 85 percent of domes- tic demand for those products are met by imports. In Japan, the ratio of exports to domestic production is 51 percent, while the ratio of imports to domestic demand is 11 percent. Japan runs a huge surplus in the electro- nics trade, while South Korea's balance of trade in electronics goods is roughly in equilibrium. In Brazil, of the total domestic production of $1,700 million in 1984, $881 million were produc-d by national firms, the remainder being local production by multinational corporations (Table 1). 9.22 This comparison makes at least three important points. First, policy in India has stressed production for the domestic markets: domestic demand has traditionally been the major consideration in allocating licenses and determining the optimal size of a project. In South Korea, electronics policy has stressed specialization and maximizing production in key industries for international markets. Because production has been for the world market, the size of the firm has had to be large enough to reap .. economies of scale. Concentration on a small number of products with scale economies bas been an essential feature of this outward-looking strategy. 9.23 Second, Indian industry has benefited little from inter/intra- industry specialization, whereas South Korea has sought to specialize in international markets. To survive the harsh competition in those markets, a country with a small domestic market has to focus on a small nunber of products. This strategy is particularly important in an economy such as India's• where the small size of the domestic market is a bottleneck to the - 136 - Table 1 PRODUCTION AND EXPORTS IN INDIAN ELECTRONICS INDUSTRY COMPARISON WITH SOUTH KOREA AND JAPAN, 1982 (US million dollars) Domestic Domestic Production Export Import Demand B/A C/D (A) (B) (C) (D) (%) (%) Brazil ~/ 881 1.s (1, 700) b/ India ~/ 1,016 74 88 ~/ 1,030 8 8 Computer Systems, 1984 88 Korea, Rep. 4,006 2,144 1,979 3,841 53 51 Semi-conductor 188 154 75 118 82 64 Integrated Circuits 671 658 160 188 98 85 Japan 43,243 22,087 2,667 23,823 51 ll Sources: a/ 1984 sales of computers and peripherals refer only to total domestic sales, excluding exports. Domestic production and exports refer to production and exports by national firms. b/ Domestic demand for all electronics for South Korea and India is obtained by indirect calculation, i.e., domestic production and exports + imports. S:./ The electronics figure for India is for 1981. ~/ Refers only to "official.. imports and does not include items carried in under accompanying baggage rules; the latter amount may be substantial. ,, sustained growth of the industry under competitive conditions. The domes- tic demand for electronics in IJdia reached only US$1.S billion in 1984, roughly one-fifth the size of the South Korean market. 9.24 Third, given the widely held view that the figure for Indian imports is underestimated, the electronics trade was larger than it appears in Table 1. The trade deficit shown there occurred even though India tried to indigenize virtually all electronics products. South Korea, on the other hand, has concentrated its production on a small number of items, yet its trade balance has still approached equilibrium. - 137 - 9.25 The size of the domestic market in Brazil is similarly small. Yet in 1~84 the production of computers and peripherals by wholly national firms w~s $881 million, or about 52 percent of total domestic production. By comparison, the level of production in India was $88 million, but pro- duction by wholly Indian firms accounted for almost 98 percent of output. The Brazilian government has adopted a strategy of selective promotion of the domestic computer industry, an approach that was predicated on progres- sive development of the industry, beginning with the des:.gn and assembly of data processing systems and then later moving onto component manufacture. Production was segmented between national firms and foreign firms; as noted, some market segments were reserved for wholly national firms, which were encouraged to grow. Foreign firms were thus contained, without a forced conversion of equity. Government procurement policies stimulated domestic demand, focusing on segments populated primarily by national firms. In the early stages of the development of the computer industry, public sector procurement accounted for about one-half of domestic demand. With the market reserved for national firms and rapid growth in the reserved segments, the market shares of foreign firms declined except in the large computer segment. Policies blocked the additional entry and growth of foreign firms. 4. Economies of Scale and International Competitiveness 9.26 The scale of operations in the Indian electronics industry is small by international standards. The reasons include the focus on produc-. tion for the small domestic market and industrial licensing policies that limit output. The policy reserving several consumer electronics products for production in the small-scale sector has resulted in a fragmentation of capacity in this segment among many firms. In consumer electronics, there are 20 units in the organized sector and over 3,000 in the small-scale sector. Although the 20 firms in the organized sector and the top 50 small-scale firms account for about two-thirds of production, the annual production of the leading firms together is less than the monthly output of the leading firms in the international market. 9.27 Some examples demonstrate the lack of economies of scale. Recently, 51 firms were licensed to produce 500 VCRs annually, compared to the leading South Korean firm, which had a capacity of 0.5-1 million units. In the case of color TVs, each leading South Korean firm prvduces 2-3 million units a year, while in India 86 firms in the organized sector and 368 in the small sector produce an annual total of only 10 million, while a leading Indian firm prvduces only tens of thousands of units. This lack of economies of scale can be seen in xhe components industries as well Table 2 compares estimates of scales of production in leading international firms with suggested minimum efficient scales for India. In the case of television sets or VCRs, Indian producers could be efficient at scales of production which are smaller than international norms. In other products, such as integrated circuits and PCBs, scales of production in India that are smaller than international scales would lead to unacceptable unit cost increases. Column three outlines minimum production scales to be applied - 138 - by the government in reviewing licensing applications. Firms would be encouraged to increase output up to these levels·. Generally, these "mini- mum operational capacities" are smaller than both international scales and estimates of production scales that could be efficient in India. Expansion beyond the ••minimum operational capacities" would still require government clearances on a case-by-case basis. 9.28 The GOI discourages small-scale firms from producing components, because it feels that for a components firm to supply quality goods for consumer electronics at reasonable prices, it should be large enough to reap economies of scale. Despite this restriction, the capacity of compo- nent electronics is still very fragmented. For instance, ICs are produced by two firms whose capacities are 4.5 million and 2.5 million units, res- pectively, whereas the international scale for IC production ranges from 50 million (without wafer fabrication) to 200-600 million (with wafer fabrica- tion) for bipolar SSI/MSis and MOS LSI/VLSis, respectively. 9.29 The retail prices of consumer electronic items in India are much higher than those of similar products in the international markets, in many cases, two to three times higher. In part the reason is the high import tariffs and domestic excise and sales taxes, but more commonly high domes- tic prices reflect managerial inefficiency and the lack of economies of scale. Furthermore, the quality of locally produced items is of ten poor by international standards. 9.30 India and Brazil share similar problems with scale, ~s both have domestic markets that are small relative to economically efficient plant sizes and to support the R&D needed to compete with international produ- cers. The issue in both countries is the trade-off between creating competitive markets by allowing several entrants, a condition that will lead to cost-efficient and technologically progressive behavior, versus achteving scale economies in production and R&D in a market in which economic scales of production might lead to a natural monopoly or oligopoly. 9.31 The two countries have approached this dilemma di=ferently. In Brazil, free and unrestricted entry by national firms using domestic and imported technology has led to competitive market structures, although at the cost of fragmented capacity. As a result, although competitive condi- tions have led to dynamic fluctuations in market shares, improvements in factor productivity and substantial price reductions over a period of a few years, the computer industry is in general not competitive in the interna- tional markets, on the basis of price comparisons of products with compara- ble levels of performance. To maintain and improve the competitive condi- tions needed to stimulate the production of machines with competitive price-performance characteristics, Brazil's strategy has been to encourage increased horizontal specialization and vertical disintegration, to focus on the need to develop exports to achieve static and dynamic scale econo- mies, and .to encourage increasing amounts of investment in R&D and manpower development. India could benefit from the adoption of a number of elements of Brazil's strategy. 139 - Table 2 Suggestei MiniJwm Efficient Product/Canponent Scale for India a/ World Scale Ef OOE/Got't Minima as of 5/5/86 • Television Sets, CDlor 200,000 1,000,000- 50,000 (SSI)- 2,000,000 100,000 Val w/ o Tape Deck & &!a:i 50,(X)() 100,0CIJ 50,000 Val w/ Tape ~ck & Head 500,(X)() 1,000,CXYJ 300,000 CRT, B&W, Chlor 500,(X)() 1,500,CXYJ 500 ,000 (B & W) 1,000,000 2,000,000 1,000,000 (Color) EPBAX 50,000 100,000 50,000 (20,000 for Finns tsing a-dot Tedm.ol~ ICs (bipolar) SSI/MSI w/o Wafer Fabri. 5,000,000 50,000,000 30,000,000 w/ Wafer Fabri. 200 ,000 ,000 600 ,ooo ,000 ICs (IDS) IS!/VLSI w/o Wafer Fabri. 50,000,000 5,000,000 4.5 Million for Assembly ani Test w/ Wafer Fabri. 200,000,000 200 ,000 ,OCXJ (Fabricat:ion could te 4 Times as High) PCBs Dbl-sided 20,000 -at. 20,000 -at. 20,000 mi. Film ResistDIS 300,000,000 1,000,000,000 100 Million by 200 Million Has Been Deliamsei Film Capacitors 50,000,000 500 ,000,000 50 Million Electrolytic Caps. 75,000,000 100 ,ooo ,000 75 Mill:ion Tantalum Caps. 100 ,000 ,000 500,000,0CIJ <10 Million Ceramic Caps. (disk) 500,000,(X)() 5,000 ,000 ,CXXJ 500 ,000 ,000 (E'.xistirg is 50 Million 13ut Not ~ Tnvestnent in the 100 ,000 ,000 But Elq:ect to gp up tD 200) M..tl.ti-1.ayer Ceramic 200 ,000 ,CXYJ 1,000 ,ooo ,oco "' Soft Ferrite 500 Tuns Hard Ferrite 1,000 Tuns Notes: 'Ihlevision sets - Assunes axial lea1 canp:ment imertion, in-circuit test~ arrl two-shift operation. Val - World scale is 1:ased on secon:l-tier Japanese finns that subcontract fre ccst of fre nedlanical canp:ments an:l always purchase tre heads. a/ Suggestei scales are those that would oo appropriate mininllm scales for India, ard are conrultant's estimtes. b/ Conmltant 's estirmtes. - 140 - D. RECENT POLICY CHANGES AND THE SUPPLY RESPONSE 9.32 In March 1985, the GOI initiated a set of "integrated policy measures on electronics". In addition, the Department of Electronics recently drew up the Seventh Five-Year Plan for electronics. In it, toe target is growth of domestic production at an average annual rate of over 35 percent, to reach a level of output of Rs 100 billion in 1989/90, equivalent to a more than five-fold increase in five years. 9.33 To achieve this ambitious target, the new electronics policy sets forth a number of policy measures to encourage investments in production, R&D and manpower development. These measures mark a significant departure from the past and introduce new concepts in industrial policy. One is the use of competition as a means of putting pressure on domestic producers to reduce their costs and prices. Another is the shift from a ceiling on manufacturing capacity to a floor on minimum capacity, with the emphasis on achieving economies of scale. Last, physical control is to give way to fiscal control, with excise and customs duties used to regulate behavior in the market. While there are still elements in the policy/procedural frame- work that will constrain healthy growth, this new policy is a positive step toward enhancing the international competitiveness of the Indian electro- nics industry. 9.34 The new policy measures and their implications are as follows. First, liberal growth in capacity is allowed once satisfactory use of existing capacity is achieved. While there is no upper limit on capacity, a minimum investment is in many cases required to assure economic viability and production at international costs and quality. For electronic compo- nents and entertainment electronics, no license is required for entry or expansion, provided that no foreign collaboration or imports of capital goods and foreign technology is involved. In such segments as entertain- ment electronics, electronic toys and games, computer peripherals, electro- nic test and measuring instruments, and discrete semi-conductor devices, "broad-banding," or flexible utilization of overall capacity limits among different products, is possible, so that firms can now shift production among items as the market changes. Moreover, all segments of electronics except !Cs and consumer items are exempt from MRTP clearances, so that all firms except those collaborating with foreign partners are now allowed to enter the industry or expand in accordance with normal licensing proce- dures, but without additional MRTP clearances, regardless of their asset size or share of the Indian market. This exemption, combined with the removal of the upper limit on capacity expansion, is a major indication that the importance of scale economies in electronics is being recognized. 9.35 The phased manufacturing program (PMP) was established to expe- dite the indigenization of components and materials and accordingly to i~crease local content while reducing dependence on imports. Submission of a PMP outlining a firm's plan to increase Sition and strong marketing skills could afford the advertising and product pto~otion expenditures needed to justify the modernization investment. . 10 • .3;)· For coarst: and medium quality fabrics, the composite mills have not been able to compete with the powerloom sector. The latter enjoys a n~mber of productioa cost advantages which are mostly derived from diffe- r~ntial regulatcry policies. First, powerlooms are decentralized, use unorganized non-union labor, and thus pay wage and salary rates that are . less than' a thtri of the rates paid in organized mills. Second, mill cloth ·, . attracts excise duties whereas powerlooms, classified as "tiny" sector units, enjoy a .number of fiscal advantages not available to organized sector firms." Third, modern technologies have implied increased opera- tional speed, 't'educed labor/machine ratios, and a reduced number of opera- tions. Therefore, they would have had to be accompanied by reductions and redeployment of the existing labor force. In organized mills, labor retrenchment has been resisted by unions and has been possible only at pro- hi bi ti ve cost. Thus organized mills have been unable to reduce the wage 'component of weaving operations. Powerloom sector firms enjoy greater iabor flexibilit} · .'\nd lower working capital and inventory costs than organ- izerl mills. As a consequence, particularly in view of the poor financial conNSUMPTION OF READY-MADE GARMENTS AND HOSIERY. Ready-Made Garments Hosiery 1980 1983 1980 1983 Cities - Quantity (Meters) 1.86 1.64 0.85 0.90 - Value (Rupees) 28.32 33.40 9.69 12.15 Large Towns - Quantity (Meters) 1.08 1.18 0.89 0.97 - Value (Rupees) 16.78 32.07 7.22 11.52 Small Towns - Quantity (Meters) 1.03 1.16 0.81 1.00 - Value (Rupees) 19.75 28.95 7.17 11.10 Source: Consumer Purchases of Textiles, Vol. II, Textile Committee Marketing Research Wing, Ministry of Commerce, Bombay, various issues. garments. Per capita consumption of ready-made garments has risen steadily over time but was still only 1.2 and 1.6 meters per person for small towns and cities, respectively, in 1983. The outlay per person amounts to about Rs 30 per person per year (Table 1). 11.04 Domestic sales of four ksy items (men's shirts and trousers, ladies' blouses and frocks) amounted in 1983 to 8.4 billion rupees (Rs) (Table 2). The value of exports of apparel had been only Rs 150 million in 1970 but reached Rs 6.4 billion in 1983 and 10.7 billion in 1985 (Table 3), an increase of 33 percent a year over the past 15 years and a 29 percent a year increase in the past two years in current rupee terms.2/ Garment exports rose from an insignificant level to become 12 percent of India's manufactured exports in 1984/85, or 9 percent of total exports. However, these exports remain very heavily concentrated in cotton garments, are very seasonal and dependent on fashion. Few are made from synthetic, woolen or blended fabrics (Table 4). r '!:_/ Total Indian merchanaise exports, by contrast, increased by only about 16 percent a year over the last decade and 14 percent a year in over . the past two years in current rupee terms • - 165 - Table 2: INDIA--DOMESTIC MARKET FOR SELECTED GARMENTS 1981 1983 Quantity Value Quantity Value (Mil. pieces) (Mil. Rs) (Mil. pieces) (Mil. Rs) Shirts/Bush Shirts - Total 75 2,154 87 2,556 - Cotton 50 857 53 989 Trousers - Total 25 1,699 28 1,W9 - Cotton 09 364 09 441 Half pants - Total 39 428 59 720 - Cotton 32 295 47 457 Blouses - Total 79 711 105 1,087 - Cotton 71 582 93 881 Frocks - Total 61 1,693 106 2,897 - Cotton 45 960 71 1,383 Vests - Total 192 1,226 239 1,787 - Cotton 190 1,219 236 1,749 Underwear/Knickers - Total 40 247 80 501 - Cotton 39 224 70 466 Sweaters/pullovers - Total 20 1,192 45 2,250 - Woolen 16. 1,004 3.5 1,906 Socks - Total 05 80 11 108 - Cotton 05 55 09 91 Source: Consumer Purchases of Textiles, Vol. II, Textile Committee Marketing Research Wing, Ministry of Commerce, Bombay, various • issues. '"1 - 166 - Table 3: EXPORTS OF INDIAN APPAREL BY MAJOR IESTINATION .• (100 million rupees and million pieces) Countries with Restraints Other Countries Total USA EEC Total USSR Total Value Quant. Value Quant. Value Quant. Value Quant. Value Quant. Value Quant. 45 141 1981 65 199 1982 63 187 1983 23 10 22 63 49 145 7 14 15 48 64 193 1984 29 79 28 80 64 176 11 22 21 54 85 231 1985 35 74 34 97 78 193 17 33 29 63 107 256 Source: Handbook of Export Statistics, Apparel Export Information Council, New Delhi, various issues. Table 4: FABRIC INFORMATION ON INDIA'S GARMENT EXPORTS (100 million rupees and million pieces, percentages in parentheses) Cotton Synthetic Woolen Total 1983 Quantity 175 (90.3) 12 (6.6) 6 (3.0) 193 (100.0) Value 54 (84.4) 6 (9.1) 4 (6.6) 64 (100.0) r 1984 Quantity 206 (89.4) 17 (7.5) 7 (3.1) 231 (100.0) Value 73 (85.6) 7 (8.6) 5 (5.8) 85 (100.0 1985 Quantity 222 (86.8) 27 (10.4) 7 (2.7) 256 (100.0) Value 87 (81.8) 13 (12.6) 6 (S.6) 107 (100.0) Source: Handbook of E~ort Statistics, Apparel Export Informati<1n Council, New Delhi, various issues. - 167 - 2. Structure of the Subsector 11.05 Garment manufacturing is undertaken almost exclusively by small- scale units using local fabrics. In 1985, approximately 9,000 small units produced for export, each with machinery and equipment worth less than 3.5 .. lakh rupees and employing less than 100 persons. An unknown but very large number produced for the domestic market. 11.06 The fragmented nature of garment manufacturing is the result of a number of factors. First, the reservation of garment manufacturing for the domestic market to small-scale producers. Second, the technical obsoles- cence and poor quality control in cloth production resulting in lower priced short pieces (fents) and labor considerations. Large-scale garment manufacturers would have to pay at least 20 percent extra for long pieces of defect-free fabric. Third, the competition among fabr:r.c retailers and tailors, is such that tailor-made clothes are as inexpensive as ready-made ones. Only as a large proportion of the population wear stitched apparel and buyers switch to ready-made cloths could production become consolidated to any marked extent. Even then, it seems likely that the inflexibility of labor policy (see below), coupled with the relatively lower wage rates for small-scale units, will still lead entrepreneurs to choose a fragmented production structure. 11.07 As to export manufacturing, the organizational structure of the subsector is explained by relative fabric costs and labor and other regula- tory policies in the face of the highly seasonal and variable pattern of demand for the cotton-based garments. Other factors are trade restrictions and input costs, the specific marketing arrangements that characterize the subsector, the pattern of quota allocation and infrastructural constraints. 11.08 Although export producing units are free of the small-scale reservation, they generally avoid larger scale, centrally located assembly facilities. The proportion of exporter~ that have their own organized assembly factories is very small. Those producing garments for export outside the Kandla Free Trade Zone generally do so through a putting-out system, whereby they cut the material but then use "third-party" f abri- cators on a commission or piece-rate basis to assemble the garments. The garments are then returned to the central units to be finished and packaged • • C. KEY POLICY ISSUES 11.09 Variability of Demand and Labor Policies. The fragmented structure of production is in part the outcome of efforts by individual producers to avoid the large central labor requirement and to maintain managerial flexibility in the face of the rigid labor regulations.~/ 11 Under the Industrial Disputes Act, an employer with more than 100 employees must have government permission to retrench (permanently lay-off) workers. - 168 - While the small-scale, fragmented assembly system clearly involves more labor time per garment, it does not have a significant adverse effect on overall labor costs, as the per hour rates are lower in small scale units. Producers in other Asian countries, including Sri Lanka, Bangladesh and Hong Kong, have shown, however, that the labor content of garment· assembly can be substantially reduced by larger scale production units with an assembly line organization. The labor time required to manufacture a shirt, for example, can be more than halved in this way. In the late 1970s, several export manufacturers in India did in ract attempt to operate central asse~bly units, but they reverted to small-scale, commission-based assembly. The difficulties these manufacturers experienced in closing their large-scale assembly units when outside the Kandla Free Trade Zone (in Gujarat State) have been a major deterrent to others that might consider establishing similar plants. 11.10 When faced with a rapidly changing and seasonal market, as India's garment exporters are, managerial flexibility is essential. The market for exports of Indian apparel is so seasonal that more than 80 percent is shipped during the first three months of the calendar year to meet the summer market in the northern hemisphere. Moreover, by concen- trating on non-standard cotton garments, the Indian export industry is relatively exposed to fashion and seasonal changes. An unseasonably cool European summer, as occurred in 1982 and 1983, for example, largely explains the absence of growth in Indian garment exports in those years. Permanent organized factories to supply such a variable market would mean idle capacity for large parts of the year. In contrast, contracting "third-party" fabricators limits the potential losses from idle capacity. This system also affords flexibility by avoiding the rigidities imposed by the labor laws and associated restrictions on exiting. Small-scale fabrication units, on the other hand, often operate for limited periods each year and are designed to fit in with other seasonal opportunities and work obligations. Thus, the current structure of the apparel industry has evolved as a means of meeting the variable market opportunities for the type of non-standard garments exported from India, while conforming with domestic industry regulations. 11.11 Trade Restrictions and Input Costs. The trade policies that restrict effective access to .blended and synthetic fabrics at internation- ally competitive prices have had an important influet:ce on the composition • of exports: they have forced them to be made almost exclusively of locally produced cotton fabric. As such, these policies have severely restricted the scope for exports in standard garment lines, such as business shirts, ( and hence have increased both the volatility and the seasonality of the demand for garment products. The cost of locally produced synthetic and blended fabrics or equivalent imported fabrics with duties is so much higher than the international prices that their use would render exports uncompetitive (see Table 5). Moreover, the inadequacy of the duty drawback and cash compenRation systems to refund indirect taxes has severely limited the use of thes~ fabrics for exports. • 11.12 While there is a legitimate reason to question the widespread use of imported material inputs to produce quota-restricted exports if the objective is to maximize net foreign exchange earnings, the ineffectiveness - 169 - of these compensatory mechan:J.sms has meant that the flexibility of produc- ers to diversify their exports and thereby reduce the limitations imposed by foreign import quotas has bean sharply x·~duced. Although the cash compensation for garment exporters was re~e~tly raised from 7.5 parcent to 10 percent, and replenishment (REP) licenses incr~ased to 15 percent of the value of exports, these rates continue to be well below what they would • need to be to offset the cost penalty associated with the purchase of local inputs. 11.13 While the current arrangement for gaining duty-free access to imported inputs to manufacture for expo~ts has worked reasonably well for the rupee trade, where lead times are greater and design standards rela- tively stable, they have not been effecti.11e for other exports. In part, Table 5: INTERNATIONAL PRICE COMPARISONS FOR SELECTED INPUTS, FY86 Landed Cost Product/Item Domestic of Duty-free Price Price Imports Ratio Polyester Filament Yarn (Rs/Kg) 57.3 26.4 2.2 Polyester Staple Fiber (Rs/Kg) 48.0 17.0 2.8 Dimethyl Terephthalate (DMT) (RsiKg) 2.l.5 8.0 2.1 Monothylene Glycol (Rs/Kg) 16.0 6.6 2.4 Cotton Yarn: 14s 20.0 24.0 0.85 20s 19.0 25.0 0.75 30s 27.0 29.0 0.92 40s 29.0 35.0 0.83 Source: Work done in the New Delhi Office of the World Bank. the reason is the delays and difficulties associated with the import-export ' procedures and controls and in part the location of the Kandla Free Trade Zone ( FTZ) •';_! !!.._/ Current regulations, for example, prever.t the establishment of a pro- duction unit, even 1£ 100 percent export-or~~nted, in and around Bombay. "• ·, ) 170 - \._I 11 .. 1'4 For established non-FTZ exporters, the flexibility to expand seres and/or move into new lines of production under the advance license and new pass-book system has been limited because of recurrent delays, wh~!eas certai~tty and quick turn-arounds are needed. For new exportei·s, . \ ' \ these ;itrangements are not available and constitutes a barrier to entry • l . However, it should be noted that such schemes are currently operating quite • efficientl:t in other Asian countries. Bangladesh, for example, has be·3n able to .st·;ceamline the import-export procedures for garment manufacturers 11 .•,'I ., ; to, such 'an· extent that they have been able very rapidly to develop exports • ''' b:t.sed entirely on imported materials. '' \ ' ll.15 Marketing Arrangements. In contrast with the highly concentrated marketing arrangements that are common for exports from other developing countries, India's apparel exports are commonly managed through family links and personal contacts. As a result, order lots are often "tailor- made" smaller batch, non-standard garment items concentrating on light- weight summer fashion garments rather than on large-scale standard business items such as blended shirts. 11.16 India's tradition of restricting foreign investment and collabo- ration together with the inadequacy at compensatory mechanism for indirect or new exporters has also limited the scope for establishing international ·marketing channels and has been another factor in the structure of the garment manufac~uring industry. Even if manufacturing continues to be concentrated in small scale units, trading should be concentrated to take advantage of economies of scale. 11.17 Quota Allocation. Although the performance of India's garment exports has been impressive when compared with that of almost aay other subsector, it has been outstripped by the other major exporters of garments in Asia. Regulations that impeded the growth and diversification of exports prior to the time when the major Western markets imposed their import quotas are expected to have a long-term adverse effect on exports. The reason is that the quotas are based on past performance, and India's early export development was low because of policies restricting imports of materials and foreign interconnections. The failure to develop these export lines effectively based on blended and synthetic fabrics prior to the widespread introduction of the import quotas has been and will be quite costly in this regard. · 11.18 Once the quotas on garment exports were imposed, the manner in which they have been allocated within India has itself fostered or main- tained the fragmented structure of garment manufacturing. At present~ 65 f peccent of the quota is allocated according to past performance, while 25 percent goes to first-come, first-served small orders, 7 percent to manufacturer exporters and 3 percent to central/state corporations. Thus, there is a very large amount of quota fragmentation. 11.19 Unless an efficient means o~ transferring quotas among different producers is maintained, this fragmentation will adversely affect perform- ance over time by lessening the incentives for cost reduction and asso- ciated market expansion. AlL10uJ:!h quc.tas have been reasonably trans fer able - 171 - and have, for shirts and blouses, yielded prices ranging from Rs 11 to Rs 40 per piece over the past two years, the costs of this system have adversely affected performance. For example, the price of the quota has varied considerably over time and across countries. The transfer price of the quota for men's shirts to the Netherlands in April 1986, for example, was less than half that prevailing one year earlier. 11.20 As with any bureaucratic system, the issuance and acquisition· of quotas also absorb valuable executive time and hence reduce the time avail- able for more creative pursuits. Moreover, the price for mistakes in the documentation concerning the license applications can be quite high. A minor omission, for example, can delay a license for six weeks, by which time the foreign buyer ref uses to pay the full price because of late delivery. 11.21 One feature of the quota allocation system that has helped increase foreign earnings has been the principle of floor prices for th~ per unit FOB value of exports in various categories. To the extent tl1at the industry is almost exclusively based on indigenous materials, thi~ system amounts to an approximation of a floor on domestic value added. However, the system only ensures that ;~reign exchange earnings do noc fall below a certain figure. If the explicit 0bjective was to maximize the foreign exchange earnings from quota e~ports, the quota would need to be allocated on the basis of the net foreign exchange earnings per unit of exports. Unfortunately, informational limitations seem likely to prevent an auction system with futures contracts, designed for this purpose, from providing the nec~ssary stability in production planning. However, it would be desirablo to increase the incentives for raising the per unit net foreign exchange earnings. Such an incentive could take the form of a foreign exchange retention scheme, whereby a rising proportion of the net· foreign exchange earnings could be retained by the exporter, with wide discretion as to their use. 11.22 The concentration of export "incentives" on exports by manufac- turers, coupled with the lack of trading or export houses with a strong interest in market development and stable long-term links with domestic producers, have inhibited the expansion of indirect exports. This situa- tion should be rectified so that market intelligence and development can be deployed more efficiently. 11.23 Infrastructural Constraints. Infrastructure such as power, transport and communications have also had an important bearing on the • industry's structure and firms' behavior. The frequent interruptions in the power supply in India have been an added deterrent to the establishment of large-scale, power-based units. Further, tight export scheduling, coupled with the risk of delays in sea transport, have resulted in a large part of garment exports being air freighted. This mode is more expensive and is also afflicted by frequent delays because of the scarcity of cargo space. At present, the onus of proof of a need for additional cargo capa- city, whether it be on charter flights or scheduled airlines, rests with the exporters. Naturally, i t is very difficult to establish a shortage in advance. - 172 - 11.24 It is apparent that there is an urgent need to provide exporters with better infrastructural support in the fo=m of continuous and adequate power, communications and transport. Without effective communication systems, for example, it is impossible to build market links. D. PERFORMANCE 11.25 The wide array of factors such as labor regulations, trade restrictions, marketing arrangements and infrastructural limitations that have given rise to the very fragmented structure of garment manufacturing and trading adversely affects performance. As a result of the structure of production, the labor and transport content is high, and quality control is quite poor. 11.26 The production structure, although it evolved in part in response to market variability and seasonality, has itself become an impediment to export diversification. The needed shift of garment exports away from small batch non-standard item~ 1nto standard ones is only possible if quality control is enhanced i:.o that the market requirements for large orders of high and uniform quality can be met. That condition will be difficult to achieve under th~ current fragmented assembly system. In fact, the greatest difficulty that the present industry structure poses for further export development may not be any increase in per unit cost, but t~ ~ adverse effect on quality control. 11.27 Quality could be raised with relatively little investment in new equipment. However, quality control and buyer confidence are likely to remain a problem as long as the structure of production is fragmented and the quality control in fabric production is so poor. These factors also explain why the development of a local market for infants' and children's wear is more advanced than that for adults' clothing. Until garment production is organized into larger scale assembly line units, a change that depends upon the supply of competitively priced long pieces of defect-free fabric, unnecessarily large amounts of labor and transport will be used in production. By increasing unit costs, these factors indirectly restrict local sales and the profitability and expansion of production for the international market. •