The pervasive effects of high taxation of capital goods in India

Poliy, Research, and External Affairs _ WORKING PAPERS Industry and Finance Operations Asia Regional Office, Country Department IV (India) The World Bank August 1990 WPS 433 The Pervasive Effects of High Taxation of Capital Goods in India Findings and Conclusions from a Sample of Projects Francois Ettori India's heavy duties on capital goods biur the incentive signals from the tariff structure. In practice, that structure favors import substitution of intermediate products from heavy industry and discourages exports. The complex protection structure should be simplified, with priority to slashing the duties on capital goods. Thc Pol,c> Resurch, and I xu:rnal Affairs Complex disrnhuses PRI W'orking apers to dissarnruate the findings of wotk in progress and to encourage uhe etchanrge of ideas among Bank staff and all orhers Intercsted in developwment issucs These papcrs carry the names of the author, mflect on1v Lhetr views, and should he used and cited accordingly The findings, interpretatios, and ccnclusions are the authrr' own 1heN sh.Z:d nr 5w attrNiiehud mo Lhe Bornd ak iks HiBard nf Director is imanagerni, or a.r) of u.s meroher crmnnecs  Policy, Research, and External Affai(s V IA Industry and Finance Operations WPS 433 This paper --- a product of thc Industry and Finance Operations Division, A .ia Regional Office, Country Department IV (India) --- is part of a larger effort to undcrtakc a comprehensive review of intlia's trade regimc and( oxlicics and to man kerecomnmncndations for liieral /.ait011 of tradc policics. Copics are available free froni thc World Bank, 8 18 I11 Strect NW, Washington DC 2()33. Please contact Francois Eltori, room DI-049 , cxtension 80324 (39 pages with graphs and tables). Some 60 industrial projects (chiefly in tihc producingi intermcdiates and inputs for down- chemical and cnginecring subsectors) finainced stream subsectcrs -- and witiin cach subsector. by thc Developmnent Finance Institutions in India in 1988 and 1989 were analyzed. The major Nominal protection rates (as reflected by finding is that levying the heaviest dutics on domestic to world price ratios) averaging 40-50 imported capital goods has deeply distorted percent, are substantially lower than average industrial incentives anid harmed industrial tarif'f collection rates (c()-70 percent) and much competitiveness and exports. lower than ol'ficial tarif'f;s (120-140 percent). The wide variations in protection rcflect a complex With taniffs on capital goods averaginig 80 system comprising many exemptions and ad hoc percent (except for clcctronic industries equip- tariffs. mnerit which pays about 4() percent), Indian projects are generally 40 to 50 percent more Tariff'reform is urgently necded. T'ariffs ;spensivc than thyc \ould bc under f'ree tradc, should primarily providce protection arid incen- and( Up tO 80 perCet ntore expensive in capital- \ives, V ith only' a seconidaryv unction of gcnerat- iniitesive projects. ing pub!ic revcrnuC. First, tariffs should be slashed, and imports liberalizied, on capital 'I hc high iu esint cosLs require a comnill- goods, tO\; ard a un;lornim tariif of 15 percent and sator) ef'fc cti\ c roictlionl aB erlging 3() percent f-ull cxcmiption for projects exponing at Icast halt to allow industrial pIrojX ds to carn rcturns at least ol output. [or intncniediates and othcr inputs. equal to lihose available undter trc trade. I Ic lo- most tari lf exempt ions should be eliminated. cvc r, about haiil the' projects igenral lv those import regirreCS unifiLd, anll(d larilfs aligned on producinig final goods) rccix\e cf'L,ctive p)rotec- :ollcction rates tom,ard rc(luced levels averaTgingt tionl significatlilk lo'.cr than111 thle com,)pensator\ .40 percent. elf'cliVye protect ion, and gcncrai loAer profits than tiose of foreign comi petitors Public rccnuC should he generated increas- inolv throughl trade-nCuiral instrunitnts (profit Nomllinal proicction \'arics Awidel\ belt cen ta;Xs aInd itiirCet taxes suchi as O(M \A'' and lh,ectciors - frl'01om 25 pericent for 'itn.:l goods colnsumliption VAT ). idi(ustries to 610 to oS perceni lor indu trics Thle PRE Workirng PaIer Series di,wsen,n:n~ ihe nndinc nf ork U1dL r vay in Lhc Bank's Policy. Rescarch, and Liternat AffEir,s Con(ipcx. An ohI- t(" 1Ie scr:ez is is to,-t thcse fil-tling', ot iii iki.,. e%en if prekenlttion-s are l.ss than futlly polishd.F The fintings: in[Crp)retalini vid Lnol(u';ons in thes. papers do noi nek.es.;ri!r represent ofiOcial Rank Tolihc r'tJlked bK [i' ikl)9i,criin.itwno (Cenitr  Table of Contents Introduction 1 General Purpose and Approach 1 Comparative Profitabilities of the Domestic and 3 International Markets Impact of Input Costs on Competitiveness and Profitability 4 Impact of Investment Costs on Value Added and 5 Effective Protection Impact of Inputs and Investment Costs on Competitiveness 9 and Nominal Protection Towards Lower Protection and Tariffs 13 Conclusions and Suggestions (Summary) 15 Annexes 21 Graphs 39  Introduction 1. The Industry and Finance Division of che India Department of the World Bank (AS4IF) undertook in 1989-1990 an in-depth review and analysis of the Trade Regime of India and of its effects and implications on the performance of India's manufacturing sector. Vast amounts of statistics, other analyses and literature were assembled to that purpose, with a view to recapitulate and synthesize the available knowledge on the subject and to draw a relatively comprehensive picture of the protection structure, growth, cost competitiveness, external trade and structural evolution in each of the major subsectors of Indian manufacturing. Prominent among the available sources of information and prior analysis were the Bank's 1987 CEM, Garry Pursell's Review of Effective Protection of Indian Industry (mimeo, unpublished), and ICICI's 1985 study of export-performance of ICICI-financed companies. 2. A significant obstacle, however, in carrying out the AS4IF's review was the scarcity of updated data on prices, costs and profitability in Indian industries relative to international markets and comparators. The most recent years for which the other available analyses and data were providing reasonable estimates of prices, protection and profitability in industry were covezing the period 1980-1986. Moreover, in computing their estimates with the best possible accuracy, these other studies paid less attention to the analysis of the internal workings and dynamics of the protection structure, i.e. the degree to which protection feeds upon itself and calls for further protection when one group or category of products (e.g. , intermediates, capital goods) is used or further processed by a downstream subsector. The purpose of this Working Paper is to contribute further to the existing studies by analyzing the internal dynamics of the protection structure on the basis of recent data valid for the Late 1980s. General Purpose and ADDroach 3. Up-to-date information and data of the type required for estimating the Competitiveness and the Effective Protection (in particular, relative domestic to international prices for inputs and outputs) of industry normally requires lengthy and extensive surveys of a representative sample of manufacturing enterprises. Due to time constraints principally, such survey could not be undertaken. As the best substitute, detailed data of the type required were extracted fro- a sample of some 60 appraisal reports prepared by ICICI and IDBI for projects which they finaneed in 1988 and 1989. These projects were made in 25 manufacturing subsectors (as ciefined in the 115 - subsector classification of the Indian economy, which comprises a total of 66 manufacturing subsectors), focussing principally on the metal and engineering industries (including electronics) and to a lesser extent on the chemical and related inoustries (synthetic fibers and textiles, phrrmaceuticals). 4. The project data were assembled with a view to providing quantitative indications, and whenever possible answers, to the following questions:  -2- (a) What are tr_ respective profitabilities of the domestic and international markets for Indian industries?; (b) Hov much of Indian industry's lack of international competitiveness stems for the extra costs paid for inputs (in the form of duties on imported inputs or high prices for domestic supplies)?; xc) What is the impact of India's investment costs, grossed-up by duties and taxes on imported equipment and by higher prices of domestic machinery, on the Value Added and the Effective Protection of Indian industry?; (d) What is the minimum level of nominal protection required by Indian industry to compensate for the extra costs paid for its inputs (item b) and for its investments (item c)?; and (e) To which extent are the actual nominal and effective protections received by Indian industry in concordance vith the levels of protection stemming from items (c) and (d) above? 5. The results of the analysis presented below should be interpreted with sowe qualifications. Firstly, the parameters and data used in DFI project appraisal reports tend inherently to err to the favorable side with respect to comparative prices (domestic versus CIF) for the projects' outputs. Secondly, the projec- tions implicitly assume that the projects will be operated efficiently to minimize production costs (inputs, labor, capacity operating ratio, ...). In practice, actual production costs are often higher than projected during appraisal, and domestic ex-factory prices are then increased as much as permitted by domestic competition and the degree of protection provided by the Quantitative Restrictions and the protection tariffs levied on competing imports in India. For these reasons, the data are most likely to underestimate the actual domestic profitability, the price uncompetitiveness, and the effective and nominal protections presently experienced by Indian industries. i/ Nevertheless, they provide an indication of the incentives and disincentives steoing from the protection structure as they are perceived ex-ante by the project promoters and financiers. In this sense, they should provide useful indications and benchmarks for a reasonable assessment of the issues raised in para. 3 above. 1/ The same favorable bias is a first factor for underestimating the effective protection. Moreover, available data ef appraisal reports could not permit to separate out the excise (or CVD) taxes from the ex-factory or landed prices of inputs to be purchased for the projects' operations. Estimating and separating out such excise or CVD taxes on the basis of the official tax schedules would have been misleading, due to the pervasive exemptions and ad-hoc rates applied in India, and was not done. To this extent, the comparative prices of inputs used in the Effective Protection computations are slightly overestimated, and the resulting Effective Protection rates are further underestimated.  -3- ComRarative Profitabilities of the Domestic and International Markets 6. About one-fourth of the sample projects were projected to export a significant share of their output (including one 100% Export Oriented Unit), and 6 other projects (one-tenth of the sample) would export a marginal share (5% to 10%). Only half of the significant exporters show a favorable (i.e., below or close to 1) ex-factory to export (FOB) price ratio (the Nominal Protection Coefficient, or NPC) 2/ for their output. O.her significant exporters (mostly in chemical and associated industries) have ur.favorable domestic/FOB price ratios, ranging between 1.4 and 2.3 (Annex I). The marginal exportets have output NPCs ranging between 1.14 and 1.84. Among the non-exporters, a group of 8 potential exporters (see para. 9) have favorable output NPCs below or marginally above 1 (0.80 to 1.14), and all other non-exporters (half or the sample) have price ratios ranging between 1.20 and 1.94. 7. Production costs per unit of output were determined for each project, on a full cost basis as well as on a marginal basis (i.e., before depreciatian, interest and administrative overheads). Profitability of the domestic market (as percent of output prices) was determined over the full production cost (i.e, domestic price minus full unit production cost). The full cost profitabi.ity and the marginal profitability (over the marginal production cost) were computed for exporters and non-exporters, as % of t'he international price (FOB in cases of export, CIF otherwise). The results, detailed in Annex I, are summarized below: Table 1: Craprative Profitabitity of Damstic wnd Internatiowal Markets Export Sawipi Export to Output FuLl Cost Profitability I() Marginal Category Share JS Output CS) MPC Dom_stic InternationaL Profitability (X) Lice PricC of Intnlt Price None 53 0 1.55 11.2 41.2 -12.2 Potentias 14 0 1.00 14.7 12.2 38.5 Marginat 10 8.0 1.39 0.7 *24.9 1.3 Significant 23 45.9 1.42 16.4 13.0 15.5 OveraLt 100 11.7 1 43 12.9 -24.6 3.6 Note: Unless otherwise specified, aLL indicators in this and folLowing tablee are simple (ureighted) avcrag". Source: Annex I 8. The projected price competitiveness, as measured by the Output NPC, does not appear to constitute the explanatory factor for the decision to export. The output price ratio is similarly high (1.4 to 1.5) for the categories of non- exporters and exporters. The only difference, which may carry the full explanation, between these two categories is the marginal cost profitabilitv of international prices: -12% for non-exporters, +15% for significant exporters. Moreover, CCS income increases the marginal cost profitability of significant 2/ The economic mechanlsms by which these price ratios (NPCs) are generated are not analyzed in this note. Suffice to say here that NPCs are the result of the various effects of import Quantitative Restrictions and tariffs relative to the degree of domestic competition and supply/demand balance permitted by the regulatory policies.  -4- exporters by only 2.6 percentage points on average. j/ This confirms the analysis and findings of the Bank' Export Strategy Report (6663-IN, Mar-h 1987) which established that Indian industry generally exports a marginal share of output when the export international price permits a reasonable profitability at marginal cost. The category of marginal exporters earns a small marginal cost profitability of 1.3% on FOB prices, and does not expect to receive additional CCS income. These levels of export profitability as anticipated in 1988-1989 compare favorably with those prevailing in 1978-1980 as recorded by ICICI for a sample of industries. !/ 9. The seconG category of Table 1, dinominated potential exporters, constitutes some sort of anomaly. They are characterized by having not only the most favorable price competitiveness (averagj NPC of 1), but also a full cost profitability over international prices which is simil&r to their profitability in the Jomestic market (12% versus 14.7%). Nevertheless, thi appraisal reports for these projects do not indicate plans or even intentions to export. These projects are generally in engineering industries, with some in chemical industries based on local primary resources. A major characteristic of these projects is their relatively high Value Added content (their VD%/Output ratio in international prices average 54%, as compared to 32% for the other project categories), which givss them more room for cutting down on the cost components of Value Added. As a matter of tact, all projects of this category have negative Effective Rates of Protection (EPRS), averaging -29% as opposed to an average +50% for the other project categories (Annex I), which suggest high operating efficiency (according to appraisal report projections). The reasons for not exporting nor considering exports are not mentioned iri the appraisal reports for the..e projects, and might be case-specific. They may relate to quality standards and issues, particularly for final engineering goods. This apparent anomaly would deserve during a subsequent mission some investigation with the DFIP which might lead to worthwhile findings. impact of Input Costs on ComDetitiveness and ProfitabilitX 10. A major source of uncompetitiveness in Indian industries has been the higher prices paid for operating inputs relatively to international prices. The main reasons for these higher prices are: (i) the tariff duties paid on imported inputs; (ii) the uncompetitive prices of domestic supplies; and (iii) the non- deductible excise taxes. On average, inputs used by the projects have prices 2/ These low levels of CCS, as explicitly recorded in the DFIs' appraisal reports, are much lower than the official CCS rates. It is quite possible that the exporting projects will actually receive CCS incomes higher than those taken into account by the DFIs. Nevertheless, the high NPCs and negative profitability of some projects in the category of substantial exporters cast doubt on their real intentions and future achievements regarding exports. In particular, it is difficult to see how six projects with highly negative profitabilities of exports on full cost could export 50% to 75% of their output. The reasons why these projects intend to export such major shares might be case-specific and would deserve further investigation. iJ Export Performance of ICICI - Financed Companies (1978/79 to 1980/81), Bombay, 1985.  - 5 - 49% above international prices. The d, estic to international price ratio (NPC) for inputs decreases from 1.58 in tha ..irst category (non-exporters) to 1.38 for substantial exporters, as shown below: Tflse 2: Iict of IrPut CoStS rd Taxes am Profitability Export Averi Inpujt Taxes as X Adjusted ProfitabiLity /g of Category IrFout NPC of gutout Intnt. Price Internatioml Price over: Ful k Cost 1laralnAt Cost Mor-q 1.58 39.0 -0.7 28.2 Potential 1.43 14.6 30.5 56.8 Marginal 1.413 27.3 2.2 28.7 Significant 1.38 21.1 8.1 36.7 Overall 1.49 30.3 6.0 34.2 /I Adjusted by remving taxes x price surcharge from irput costs. Source: Arwex II. The additional costs stemming from the higher input prices represent on average 30% of the output's international price. This cost handicap is significantly higher for the non-exporters (39%), and markedly lower for the potential exporters (only 14.6%) due to their high Value Added content and their use of relatively cheap domestic resource-based inputs (e.g., Maize, Aluminum). 11. A simulation of the "tax-free" production costs (excluding the surcharges paid for inputs, i.e., the tariffs on imported inputs, the price differentials on local ones, and the non-deductible excise taxes) indicates that the competitiveness and profitability of the projects relative to international prices would be greatly enhanced by eliminating these surcharges. Practically all projects which had a negative marginal profitability on international prices would generare a positive tax-free marginal profitability under such conditions. Furthermore, the pro?ortion of projects which would ear from international prices a positive profit on their full production cost wo, i triple (from 20% to 63%). In the category of non-exporters. this proportion would increase dramatically from 0 to 50%, i.e., half the projents which have no prospect nor potential for export under the present circumstances could become substantial exporters with a reasonable positive profit margin if they could procure their istputs at international prices. It confirms the major finding and recommendation of the Zxport Strategy Report, regarding the critical importance of access to inputs at international prices for any successful export policy. Impact of Investment Costs on Value Added and Effective Protection 12. One uncomoon feature of India's protection structure is the high lerel of customs tariffs and taxes levied on imported capital goods. Correlatively, the domestic capital goods industry supplies the rest of the industrial sector at prices generally higher than international prices for comparable capital goods. Moreover, the price ratios for capital goods increase with the technology level of capital goods and equipment (e.g., CNC machine-tools have NPCs exceeding 2 as compared to 1.2-1.4 for standard machine-tools). 13. The high cost paid by Indian industrial projects for their capital goods (either local or imported) implies that the projects' Value Added in domestic prices should normally include a larger amount of capital ren'imeration (interest, depreciation, and return on fixed assets). This by itself contributes, ceteris  paribus, to domestic Value Added exceeding Value Added at international prices, and to a positive Effective Rate of Protection (EPR). In fact, the difference between a project's financial return on capital (or KRR) in domestic prices and its economic return (or ERR) in international prices under a free trade regime is shown (see Annex III) to be approximately expressed as follows: MRR - ERR - C. ((1 + UJ) - 1] y 1-s where: MRR is the (pre-tax) financial internal rate of return; - ERR is the economic internal rate of return; * y is the ratio of the Investment Cost in domestic prices to the Investment Cost in economic (international) prices (this ratio is greater than 1 in India) 5./ - s is the share of Labor in the Value Added in economic (international) prices; - EPR is the project's Effective Rate of Protection; and - C is a parameter specific to the project, i.e., depending exclusively on its parameters in world prices. It follows (cf. Annex III) that in order to ensure to an investment project a financial pre-tax return at least equal to the economic return achievable under a free trade regime, the project should receive a Compensatory Effective Protection Rate (CEPR) at least equal to: CEPR - (l-s) (y-l) i/ 5/ The factor y represents in fact the ratio of both Investment costs required per physical unit of output (ton or piece). It is thus the product of tuo factors: y - x.z. Factor z, the ratio of Investment costs expressed in domestic and international prices, captures the effects of capital goods pricing policies. Factor x captures the effect of x-inefficiencies in selecting the project's production capacity. In subsectors characterized by substantial economies of scale such as chemical iri'ustries, the investment cost per ton of output, and thus x, increases by about 25% each time the capacity is .reducea by half. Many projects in India have capacities far below international standards for MES, and have x factors significantly above 1. However, in order to focus the analfsis on the effects of pricing policies for capital goods, x is implicitly taken to equal 1. i/ This formula for CEPR can be interpreted by using one unit of VA in international prices as the numeraire. With this numeraire, the additional VA in domestic prices (CEPR) compensates for the renumeration (l-s) of the additional investment cost (y-l). Moreover, the formula for CEPR becomes (1-s) (y-l)+t, whera t is the share of income tax in the VA (in international prices), if the after-ti- financial return is to be equalized to the economic return.  . 7 - Moreover, e.he quantities KRR-ERR and EPR-CEPR should normally have the seme sign, i.e., be simultaneously positive (or negative). 14. Factor y in India is significantly above 1 for several reasons. Firstl, collected tariff duties on imported capital goods for projects are high. They average some 70% (Table 3). Only electronics industry machinery enjoys a lower tariff duty o ;os. Other industries pay an average tariff duty of 80% on their imported equipment. Secondl locally procured capiLal goods, generally representing the major share of equipment, carry purchase prices which average 40% aoove international prices. _1 Thirdly, other goods and materials usnd in investment projects, cement in particular, are charged substantial excise and other taxes which increase further the financial costs of invsstment above international equivaler.ts. 15. Two estimates of factor y were computed for the sample projects. The first estimate yl was computed as the ratio of the financial investment cost to the same adjusted for tariff duties on imported equipment and for the high prices of local equipment (the first two factors described in para. 14). This factor yl captures only the impact of the high prices of capital goods. The second estimate y2 is the ratio between the financial investment cost and the 'economic' investment cost (used for ERR computation), and captures not only the effect of capital goods prices but also the price distortions of other goods and factors (e.g., cement, interest during construction, ...). In a sense, the factor y2 represents an upper limit of the cumulative effects of all forms of investment taxation in India. 16. The sample data confirm that investment costs in India are significantly higher than international costs, and that the taxation of. capital goods is the major factor accounting for that, as illustrated below: Tabte 3: Effect of Investfint Costs on Effective Protection /I Export Tariff (t) on Investwent Cost Factor Cqpersatory Actual Projects (X) in Ca"soorv Imorted Machinery xi x2 EPI (CEPR) /b EF Excess/Shortfatl (in S) (in X) None 74.2 1.41 1.55 30-41 60 65/35 PotentiaL 76.1 1.52 1.60 45-52 -27 0/100 marginaL 56.8 1.36 1.52 23-33 54 33/67 Significant om 0 LR 2740 -jt 23/7 OveraLl 70.5 1.40 1.55 30-42 30 42/58 of which. ELectronics 3$.9 1.27 1.40 15-22 7 Others (A .) 78.7 1.43 1.58 34-46 34 /a Actual EPRs In this tabLe are weighted by VAi /b The Lower (resp. higher) value of the indicated range correspords to y1 (resp. y2). Source: Arnea IV 17. These results call for a number of observations and remarks. First, exporting projects (whether marginal or significant exporters) pay lower duties on their imported equipment than non-exporters (about 60% versus 75% on L' Derived from the conversion factors used by the DFIs for local capital goods in their ERR computations. This 1.4 ratio is reasonably close to our own estimates.  average). IL/ This tariff differentiation has an impact on the y factor, which is slightly but significantly lower for exporters (yl averages 1.34) than for non-exporters (average 1.45). The yl tactor indicates that the taxation of imported and local capital goods in India makes industrial investments 40% more expensive than they would be if they had access to capital goods at incernational prices (this caxation impact is only 27% for projerts in electronics), Other distortions and taxations captured by factor y: add about 15 percentage points of extra-costs (difference between yl and y2) 18. Secon, the surcharge on investment costs due to Laxation of capital goods would reqiire a compensatory Effective Rate of Protection of 30% on average to allow Indian investment projects to earn returns equal to those they could earn under a free-trade regime. The amount of Compensatory Effective Protection stemming from capital goods taxation varies between exporters (average of 25%) and non-exporters (average o' 35%), and between projects in electronics (15%) and other sectors (34%). Other source; of distortion and taxation on investment costs would require on average an additional 12 percentage points of compensatory effective protection (only 7 percentage points in Electronics). 19. Thirdl, a large majority of non-exporters (65%) have actual EPR in excess of the CEPR (sea last column of Table above), as opposed to a niinority of exporters (marginal or significant exporters). Potential exporters all iiAve their EPR much below their CEPR. Import substitution is clearly favored by the structure of Effective Protection. 20. The substantial differences between Electronics and other subsectors with respect to duties on imported equipment and the resulting CEPRs suggest that Effective Protection varies markedly from one subsector to the other. The data presented in the previous Table, after reclassification by subsector (as opposed to export category), indicate indeed significant variations between subsectors in CEPRs and in the discrepancies between CEPRs and actual EPRs (see Table 4 below). Average CEPRs (adequately weighted by Value Added in world prices) range between about 15% for Electronics to 30-45% in most other subsectors and up to 45-55% in Synthetic Yarns. Furthermore, the comparison between CEPRs and actual EPRs indicates wide variations in discrepancies and uniequal incentives from one subsector to the other, though the CEPR/EPR discrepancy on average over all subsectors is small (and within an acceptable error margin of