Capacity Taxation: The Pakistan Experiment
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

In developing countries, an excess of industrial capacity is a disturbing phenomenon. Whereas at one time capital formation was believed to be the basis for economic development, recent attention has focused on the underutilization of existing capital stock.1 Thus, in Pakistan, where it is assumed that capital is critically scarce, the use of industrial capacity is reported to be much less than it is in the capital-rich United States—a paradox of no small significance.2 This finding has initiated a review of the accepted beliefs concerning economic development policies.3 Higher capacity utilization rates should be recognized explicitly as an alternative to saving, it is argued, and policies should be designed specifically to increase these rates as this “… holds great promise for increasing the level and rate of growth of income in underdeveloped countries.” 4

Abstract

In developing countries, an excess of industrial capacity is a disturbing phenomenon. Whereas at one time capital formation was believed to be the basis for economic development, recent attention has focused on the underutilization of existing capital stock.1 Thus, in Pakistan, where it is assumed that capital is critically scarce, the use of industrial capacity is reported to be much less than it is in the capital-rich United States—a paradox of no small significance.2 This finding has initiated a review of the accepted beliefs concerning economic development policies.3 Higher capacity utilization rates should be recognized explicitly as an alternative to saving, it is argued, and policies should be designed specifically to increase these rates as this “… holds great promise for increasing the level and rate of growth of income in underdeveloped countries.” 4

In developing countries, an excess of industrial capacity is a disturbing phenomenon. Whereas at one time capital formation was believed to be the basis for economic development, recent attention has focused on the underutilization of existing capital stock.1 Thus, in Pakistan, where it is assumed that capital is critically scarce, the use of industrial capacity is reported to be much less than it is in the capital-rich United States—a paradox of no small significance.2 This finding has initiated a review of the accepted beliefs concerning economic development policies.3 Higher capacity utilization rates should be recognized explicitly as an alternative to saving, it is argued, and policies should be designed specifically to increase these rates as this “… holds great promise for increasing the level and rate of growth of income in underdeveloped countries.” 4

When capital formation was still the dominant theme in development economics, tax incentives such as accelerated depreciation, investment allowances, rate reductions, and exemptions of income and product taxes played a prominent role in furthering a government’s industrialization objectives. While the literature on these investment incentives is impressive,5 little attention has been given to the possibility of using taxation to promote a fuller utilization of existing capital stock.

In the past, production incentives have been largely associated with agricultural taxes. For example, the incentive aspects of presumptive agricultural income taxes based on standard land yields or on standard rates of return from the capital value of land have often been stressed.6 A similar effect is ascribed to land taxes based on potential output or on the value of land determined as a function of potential output.7 However, as a rule the incentive effect is only an incidental (though welcome) by-product of presumptive assessment methods, which are probably introduced in most cases because it is difficult to tax the agricultural sector on the basis of actual yields and values.

Early incentive schemes relating to industrial production were designed to correct market imperfections such as those caused by the monopolists who, being in a position to set the price of their products, could earn excess profits by keeping their output below the socially optimum level. The most notable is the tax-and-bounty scheme of Joan Robinson, who suggests that monopolies should be induced to produce a purely competitive output by granting them a subsidy equal to the marginal cost minus the marginal revenue corresponding to that output. The excess profit (inclusive of the subsidy) would then be fully recouped through a lump-sum tax that would leave the post-bounty equilibrium undisturbed. However, she notes that the proposal would be impractical because of the indefiniteness and instability of demand curves.8

A more detailed incentive tax for production has been worked out by Knorr and Baumol.9 They suggest that the rate of economic growth should be accelerated through a tax-and-rebate scheme involving the imposition of a flat rate penalty tax on each firm’s value added, in conjunction with the allowance of a tax rebate (or subsidy) dependent on the rate of growth of the value added to goods that the firm actually sells. A similar tax was experimented with in Canada in the early 1960s, when an offset against taxable income was made available in the form of a tax credit on the taxable income from the sale of goods marketed in excess of the “sales base” (defined as average net sales in the three preceding years).10 Recently, the use of tax incentives for production has been recommended for developing countries by Vito Tanzi, who proposed a levy on potential value added, measured by deducting actual inputs from the full capacity output (determined through annual surveys) of manufacturing enterprises.11

With the exception of the brief Canadian experiment, these incentive taxes for production never left the drawing board—perhaps largely because of their administrative complexity and the uncertainty of their effect. Another possible reason is that there is little need in developed countries to induce the business community to do what is usually in its own interest and within its reach. On the other hand, various constraints may inhibit the growth of industrial output in developing countries, even though the necessary capital stock is already in place. Faced with this situation, the Government of Pakistan designed a tax to increase the level of industrial production through better utilization of existing capacity. It believed that firms taxed on the basis of their full capacity output instead of their actual production could be induced to use their plant and equipment more fully. In view of what appears to be at stake, Pakistan’s experience with “capacity taxation” deserves closer scrutiny.

This paper first outlines the conceptual characteristics of the capacity tax, its setting, and its historical development. It then describes and comments on the technical features of the scheme. Next it analyzes the likely economic effects of capacity taxation and offers some evidence on the development of capacity utilization rates after the imposition of the tax. This is followed by reflections on the revenue implications of the tax. Finally, the paper summarizes the major findings and conclusions to be drawn from Pakistan’s experience.

I. Concept of the Capacity Tax

The capacity tax was first imposed by the Government of Pakistan in 1966 in lieu of the excise duties previously charged on certain products under a system of production controls requiring the presence of tax personnel on factory premises. By taxing capacity output instead of actual production, the Government believed that the decline in average tax rates as production expanded would stimulate output. In addition, administrative procedures would be simplified to benefit both taxpayers and collectors, because the capacity tax obviated the need for production controls and excise personnel could be withdrawn from factories. At the same time a potential source of collusion and tax evasion would be eliminated, because continuous contact between tax officials and taxpayers would be unnecessary.

Initially the capacity tax in Pakistan was applied to three industries—cement, soda ash, and sugar—which were chosen because each produced a commodity that was homogeneous, uniformly priced, and manufactured from indigenous raw materials.12 It was believed that these criteria would facilitate the imposition of capacity tax. More importantly, these industries were not subject to the vagaries of import constraints and restrictions; hence, in principle, each firm would be free to plan its own volume of production. In 1968 two industries—vegetable products (ghee and oil) and cotton textiles—were added, although they did not satisfy all of these requirements. Vegetable product manufacturers depend partly on soybean oil imports for their production (the other raw material being cottonseed oil, produced domestically), and the products of the textile industry may differ widely in price and quality because of the wide range of fabric fineness or yarn counts. Since most of the problems experienced with the introduction and implementation of the capacity tax are attributed to the textile industry, more attention is given in this paper to it than to the relatively simpler systems applicable to the other products.

Production capacities were computed by the Central Board of Revenue, mainly from past and comparative physical production data as well as machine ratings, and these were published in The Gazette of Pakistan. Local tax officials ascertained the tax liability by applying the tax rate, expressed as a specific amount per unit of production, to capacity outputs. The annual capacity tax assessment was payable in 12 monthly installments upon presentation of a demand notice;13 default carried a penalty of 1 per cent a month. The annual production capacity, and hence the tax liability, could be reduced by regional allowances and adjusted with the installation of additional machinery or removal of old equipment. Abatements were granted if production had to be halted for reasons beyond a manufacturer’s control or if widespread industrial setbacks occurred, and a refund scheme was in effect for exports. A Review Board was set up to examine taxpayers’ grievances about notified capacities.14

The basic workings of the capacity tax are illustrated in Table 1 for a Karachi-based cotton fabric mill assumed to be operating with 1,330 looms for 897 shifts (each of 8 hours duration) annually.

Table 1.

Pakistan: Sample Computation of Capacity Tax Liability, 1969/70

article image

S.R.O. No. 61(R)/68, Gazette of Pakistan, April 22, 1968, p. 198.

Capacity adjustment if loom is installed or removed.

S.R.O. 120(I)/69, Gazette of Pakistan, June 28, 1969, p. 526.

After reduction of capacities with the regional allowance for West Pakistan of 10 per cent; see S.R.O. No. 61(R)/68, rule 3(5), op. cit.

The capacity tax forms an integral part of Pakistan’s extended system of excises that are levied under a tariff schedule consisting of 54 separately identifiable commodities and one service activity.15 The schedule comprises sumptuary commodities, benefit-based levies, commodities primarily justified to be taxed on revenue grounds, luxury items, and a fairly wide range of raw materials and intermediate goods. Most rates, including those on commodities subject to capacity tax, are specific. Excise taxes are an important source of revenue in Pakistan, contributing 40 per cent to total Central and Provincial Governments’ tax receipts, or 3.3 per cent of gross domestic product, in the fiscal year 1970/71 (Table 2). The products subject to capacity tax form a substantial part (9 per cent) of total tax receipts. Apart from the large tax base (cotton textiles being by far Pakistan’s most important industry), this high percentage can be explained by the high tax rates, as the opportunity was taken to merge existing duties into a single capacity tax rate when the new tax was introduced.16 The Government stressed, however, that additional taxation was not intended.17

Table 2.

Pakistan: Central and Provincial Governments’ Tax Receipts, 1970/711

(In millions of rupees)

article image
Sources: Tax receipts: The Explanatory Memorandum on the Budget, 1972–73 and The Budget in Brief, 1972–73. Gross domestic product: Central Statistical Office, Monthly Statistical Bulletin, Vol. 20 (January 1972).

The fiscal year is July 1/June 30. The figures cover the whole of Pakistan, before the separation of Bangladesh.

In decreasing order of revenue importance: land revenue, net wealth tax, estate duty, and gift tax.

Tobacco, benefit-based levies (gasoline, petroleum products, tires, tubes), revenue levies (synthetic textiles, yarn, wool, kerosene, tea, salt, soap, matches), luxury goods (batteries, bulbs, tubes, electric and gas appliances, cosmetics, perfumery, shoe polish, creams, glass products, beverages, hotel and restaurant consumption), raw materials and intermediate products (natural gas, paints, varnishes, plastics, steel products, jute manufactures, paper, tanned leather).

On cotton yarn.

Motor vehicle taxes, entertainment duties, and other miscellaneous provincial levies.

historical development

The capacity tax was first mentioned in the Finance Minister’s speech announcing Pakistan’s 1961/62 budget: “… to provide production incentives to industry … power is also being taken for levying duty on the basis of installed capacity. This alternative, which, wherever applied, will do away with the day-to-day excise control, will be adopted in respect to such excisable commodities as are found suitable for this treatment. We intend to adopt this system wherever possible and hope that it will give a great fillip to production. As the amount of duty payable by a manufacturer will not be related to the actual production, the more he produces the lower would be the incidence of the duty.” 18

In the three years that followed, little attention was given to the tax, except as an administrative expedient for collecting excise duty from small manufacturing units. Because the small industry exemption for textiles had been used as a tax avoidance device by larger units that split up operations, diverting their highly taxed production to smaller units,19 the excise administration gradually withdrew the exemption and started taxing small units—not on the basis of their actual production (square yardage), which would be difficult to assess, but by reference to the number of looms installed. Tax was levied in this manner on small cotton cloth manufacturing units with not more than 20 power looms.20 In 1963, the scheme was extended to small art-silk manufacturers who would henceforth pay duty at a flat monthly rate on the basis of installed capacity. The Government stressed the measure’s incentive effect on production, especially when there were three shifts.21

In 1964, the payment of excise duty per loom—considered to be a successful experiment because both revenue and production increased—was extended to art-silk units with up to 50 looms.22 More significantly, it was proposed that for all cotton textiles standard rates combining the excise duty and the sales tax should be worked out on a per-loom basis for each factory in the light of past production records. Mills opting for this alternative would pay the fixed rate throughout the year, irrespective of changes in production. Similar arrangements would be made for the vegetable ghee and cement industries, although these would pay the alternative rate per ton of productive capacity.23

The introduction of the capacity tax would be facilitated by the experience gained with the self-clearance procedures, introduced in 1962/63, that replaced the cumbersome production controls administered by excise personnel stationed on factory premises. These procedures permitted manufacturers at principal industrial centers to clear their own goods under the obligation to debit for the tax incurred an “account-current” that they maintained with the excise administration. The credits to this account were advance payments determined by the administration on the basis of such past production patterns as the output of various taxable categories of fabrics. Under the new procedures, excise staff was gradually withdrawn from factories, and the administration relied on documentation such as accounts prescribed for raw materials and on surprise inspections to verify compliance with the law.24

However, the capacity tax proposal could not yet be implemented. The Government thought that a simple amendment of Section 3 of the Central Excises and Salt Act of 1944 would be sufficient to impose the capacity tax, but taxpayers argued successfully in court that the tax was not an excise duty but was levied in lieu thereof. Therefore, a constitutional amendment was required to levy the capacity tax, and it was not adopted until March 1966. The Government then made cement, sugar, and soda ash subject to capacity tax from the beginning of fiscal year 1966/67. It described the tax as an “… incentive for increasing efficiency, maximising production, reducing tax evasion and lessening the burden of the tax collecting machinery both on the industry and Government.”25 Later it was also stated that the capacity tax would provide incentives for export, but arguments for this assertion were not given.26 Presumably it was thought that an exporting manufacturer selling part of his output domestically would be completely exempt from duty if his exports equaled or exceeded his notified capacity. However, the tax differential vis-à-vis his competitor, with the same output and capacity but selling his whole production locally, would remain the same as under the excise system.

Tax evasion was described in the budget speech as “… malpractices and collusion between some of the manufacturers and the lower officials of the Government departments.”27 There is evidence, particularly in the textile industry, that manufacturers sometimes used night shifts to produce excisable commodities with high rates of taxation, but did not report them because excise officers were not on the scene.28 The Central Board of Revenue is said to have estimated that revenue thus lost on some products ranged from 10 to 25 per cent.29 Since capacity estimates were in part determined by historical output data, it appears that the introduction of the capacity tax would not necessarily prevent this type of tax evasion; in fact, underreporting may thus have been institutionalized.

At the time of the 1966/67 budget, the Minister expressed the hope that it would soon be possible to extend the capacity tax to cotton and art-silk fabrics. So far it had been impossible to devise a suitable formula, presumably because these products showed wide variations in form, price, and quality.30 A committee was appointed to examine the matter, but its recommendations could not be accepted in time for the 1967/68 budget because of legal and technical difficulties. Instead, a central committee under the Ministry of Industries was set up to formulate rules and to determine production capacities for the textile industry. In this task it would be guided in the application to individual units by certain broad principles that would take into account “the national average annual production of each statutory category of fabrics or yarn on a per loom or per spindle basis, the average production of the unit itself, the production of a comparable unit, and … changes in production because of technological improvements, normal growth or other such factors.” 31 After the committee reported its findings and recommendations in early 1968,32 textile fabrics and yarn, as well as vegetable products, finally came under the capacity tax, bringing to five the number of industries to which this novel tax was being applied.

But the initial problems that had beset the capacity tax were not over. Almost all textile units filed with the Review Board applications for a review of notified capacities and subsequently went before the high courts. In response, the Government appointed a review committee to study the effects of the tax on production, prices, exports, and revenue, with particular reference to the alleged inequitable and discriminatory consequences of the system.33 Following the committee’s recommendations, the Government provided taxpayers with more complete information on how capacity was determined, widened possibilities for appeal, and decided to review notified capacities at four-year or five-year intervals.34

However, objections and appellate procedures continued. Partly as a result of the stay orders granted by the courts, arrears of tax accumulated which would not have been possible under the traditional excise tax system because goods could only be cleared after payment of tax. The backlog in collections amounted to some 15 per cent of annual excise tax collections at the end of 1970/71. These and other factors led to a diminished role for the capacity tax in the following year. Bangladesh abolished the tax altogether. In West Pakistan, the cement units and one of the two soda ash units were transferred to the public sector, relegating to the background the role of the capacity tax in these industries. The textile industry was given an opportunity to opt out of the capacity tax if the plant had been installed prior to 1956, presumably on the grounds that the obsolescence of machinery affected the feasibility of operating at theoretical capacity. Twenty-nine units subsequently reverted to the traditional system of paying excise duty on actual production. Sugar mills were given a similar option, and 17 units out of a total of 19 in West Pakistan requested to be taxed on actual production. In the case of the vegetable product industry, involving 23 units, capacity taxation was discontinued as of July 1, 1971.

However, in 1972 interest in capacity taxation revived. In October of that year, with the simultaneous promulgation of new rules, the Government brought the whole sugar industry back under the capacity tax.35 Thus, the capacity tax became applied to four industries in West Pakistan: cotton textiles (97 factories), sugar (19 mills), cement (9 factories), and soda ash (2 factories).36

II. Technical Features of the Capacity Tax

Although it may be clear that “capacity” represents a rate of production that can be attained rationally (economically) in the short run, given a fixed plant and equipment, further specification is obviously necessary. Is capacity a maximum (physical) production concept, an optimum (cost-related) frame of reference, or is it sufficient to use preferred operating rates? 37 In practice, even more important than the concept is the measurement of capacity. Generally, five different approaches have been adopted, separately or in combination: (1) a questionnaire inquiry as to what firms themselves regard as capacity output, (2) peak forward extrapolations of historical output data, (3) peak capital-output ratios, (4) capital outlays themselves, or (5) engineering data.38 This section examines the way in which capacities are determined in Pakistan. In addition, the rules for duty abatements when actual production falls short of target, the export rebate scheme, and a few other provisions are reviewed. Finally, the relative merits and demerits of excise and capacity tax administration are compared.

measurement of capacity

Basically, the Pakistan excise administration used three factors to determine annual production capacities: (1) estimates made by manufacturers themselves, (2) machine ratings, and (3) past production data. For sugar the recovery rate from cane formed an additional criterion, and for vegetable products output data of comparable factories were also used. More complicated rules were devised for cotton fabrics and yarn which, in addition, made reference to such factors as national average production, growth rates, and the hypothetical production of a profit-maximizing firm. The criteria used for each industry are summed up in Table 3.

Table 3.

Pakistan: Salient Features of Capacity Tax Rules

article image
article image
Sources: In order of listing, Statutory Rules and Orders (S.R.O.) 87(R)/66, 86(R)/66, 85(R)/66, 18(R)/68, 61(R)/68, and 62(R)/68; and Central Statistical Office, Monthly Statistical Bulletin, Vol. 20 (January 1972). Unless otherwise indicated, all data relate to the date on which the tax was introduced.

Actual production of the year in which capacity tax was introduced, unless otherwise indicated.

This is the abatement rule for work stoppages beyond a manufacturer’s control. For closures excluded from abatement and for the general disaster clause, see text.

As amended May 6, 1969 (S.R.O. No. 73(R)/69).

37,754 looms.

2,783,384 spindles.

Machine ratings—that is, engineering estimates readily available for the standard type of plant and machinery employed in the sugar, cement, and vegetable product industries—appear to be the most objective criterion. A manufacturer’s declaration about the production capacity of his factory refers to responses to industrial surveys or, more often, to requests for permission to import raw materials or capital goods. Since the import-licensing system was administered on a rated capacity basis, these requests probably overstated capacity. On the other hand, as noted earlier, there is some evidence that past output data collated by the Central Board of Revenue reflected underreporting to evade tax. By extension, this would also be the case with the maximum output of factories of identical capacity. In some cases this may have introduced a downward bias in capacity estimates, particularly because great weight was attached to past production figures.

Considerable uncertainty is introduced when the capacity concept relies on such characteristics as the percentage yield of sugar from cane. A not uncommon change in the recovery rate of, say, 2 per cent from one year to the next may mean a swing in production of some 25 per cent. A factory may then be crushing cane 24 hours a day, but it will be unable to increase its output beyond what inputs will yield. Disputes on this point led to a ruling by the Review Board that recovery percentages should be made dependent on the recovery percentage achieved by the most efficient unit in a particular geographical zone, due allowance being made for variations in soil, climate, irrigation facilities, and the manufacturing process.39 The Review Board also decided that the highest daily crushing capacity achieved in any year should be regarded as the norm but that the degree of obsolescence of the machinery should be taken into account for the individual units. Some variation may also be inherent in determining output capacities of cement units if output is based on clinker-crushing capacity. The addition of hardening agents may mean gains in output of from 4 to 6 per cent. Clearly, tax liabilities that cannot take account of substantial differences in interunit outputs are inequitable.

Considerable effort was put into the proper calculation of the annual production capacities of cotton fabric and yarn units. The Textile Industry Capacity Committee, composed of a chairman and three other members, undertook an intensive four-month survey of weaving and spinning activities in Pakistan by sending working groups (consisting of a textile technician and a cost accountant) to each mill. The groups worked under the supervision of committee members who determined annual production capacities on the basis of the standard number of shifts a year and the average output for each spindle shift (in pounds) or loom shift (in linear yards). To determine shift production, standard efficiencies and machine utilization rates per spindle and loom were computed, the latter being reduced by allowances for maintenance, repair, and technical depreciation. In the case of spindles, the standard number of shifts a year was set at 1,000 for West Pakistan and 921 for East Pakistan, reflecting the extra time lost because of more frequent shift changes (every four hours in East Pakistan, instead of every eight hours in West Pakistan). The corresponding number of loom shifts were 897 and 766, respectively. The production data thus computed were compared with the national average production for the preceding three years. To permit the cotton fabric industry to catch up with the new standards, a country-wide rebate of 20 per cent of capacity was allowed for the first year after the tax came into operation and 10 per cent for the second year.40

However, no rebate was granted to units that had already achieved the prescribed efficiency standard, as indicated by their average annual output for the preceding three years. Moreover, units whose production exceeded the notified capacity were taxed on the basis of the three years’ average annual output.

The Committee did not take into account any growth factor (defined in the capacity tax rules for the textile industry as “the past rate of improvement in production and likely increase in efficiency relating to improvement in technical, managerial, labour and financial factors of individual mills”), because the recommended annual production capacities were in excess of actual production figures. Similarly, “the technically possible maximum production potential” of a unit aiming at “maximizing its profits before tax” was not considered. Clearly, the output at which profits are maximized can hardly, if ever, be ascertained in the real world. At best, profit maximization is a process of trial and error and not a goal which can be achieved by following a textbook prescription.

After the capacities for each of the four taxable categories of cotton fabrics has been determined, tax is assessed on the basis of square yardage, the rate depending on the fineness of the fabric.41 The tax liability is adjusted if an additional loom is installed or if a redundant machine is dismantled and removed from the factory. The adjustment takes effect immediately if capacity has expanded, but if a loom is taken out of production, tax payments are corrected from the first of the month following the month in which the loom went into disuse. Taxable capacity is then increased or reduced by the production capacity of the particular loom, computed by dividing the annual production capacity of the factory for each of the four categories of cotton fabrics by the number of looms.42

In this connection, an important point to note is that downward adjustments in the tax liability are made only if a loom is dismantled and removed but not when it is temporarily shut down. On the other hand, the installation of new machinery results in an immediate increase in tax. The penalty for creating or operating with excess capacity is thus reinforced. Of course, once a firm is operating at full capacity, the prospect of upward tax revisions may act as a deterrent to new investment. Administratively, the tax adjusment provisions are an effective antiavoidance device, because machinery would otherwise be shifted around. Finally, for cotton textiles, the Government decided that capacities would be reviewed at four-year or five-year intervals. Notified capacities of all other industries were changed several times by the Central Board of Revenue and the Review Board before a final figure could be agreed upon with the taxpayer.

abatements

To alleviate hardship, the capacity tax rules provide for two kinds of relief.43 The first is based on production: if, for reasons beyond the control of the factory’s management (such as a cyclone or labor strike), actual production in a financial year falls short of notified capacity, the Central Board of Revenue, with the prior approval of the Government, may grant abatement at the rate and to the extent that it considers proper. This form of abatement, which is similar to the “disaster relief” provisions common in agricultural tax schemes, applies to industry-wide production failures. It was reported to have been put into effect in 1967 when a number of sugar mills were seriously affected by a drought that had substantially reduced the sugar yield of cane. At that time relief was granted if actual production fell short of notified capacity by 10 per cent or more.44

Under the second relief provision, applicable to individual cases, abatement is given for any day or shift for which the factory has to be closed due to circumstances beyond the management’s control at a rate corresponding to the tax payable per day or shift (see Table 3). Normal cleaning or repair operations, as well as closures for a period of less than six days at a time (and longer if the tax authorities have not been properly notified), do not qualify for abatement, as they had already been taken into account in determining machine utilization rates. To safeguard revenue and presumably to limit abuse of these rules, the Government prescribed that abatement would not be allowed if actual production exceeded capacity output; and, if a shortfall arose in production, the abatement would not exceed the difference between the duty payable under the capacity tax and the tax that would have been due if the capacity tax had not been levied.

Although the abatement rules generally facilitate the administration of the capacity tax, they do not lessen the extent to which actual production must still be ascertained nor do they remove the potential conflict between normal cleaning and repair activities and closures beyond the management’s control. It is tempting to evade the six-day rule by resuming (but not reporting) production before the period expires. More seriously, there is a curious incentive to postpone operations until the next six-day period, in order to qualify for abatement. Furthermore, factories may try to bunch production, particularly if they foresee supply bottlenecks, and then halt operations (claiming circumstances beyond their control). In practice, the abatement provisions may favor firms with fluctuating production patterns over similar firms with output spread more evenly throughout the year, as the former are in a better position to evade the rules. Finally, a general economic effect, often noted in connection with presumptive agricultural taxes, is that abatements reduce the incentive effect of the tax.

export rebates

To achieve the Government’s objective of freeing exports from tax, the administration allows exporters a rebate on the capacity tax rate.45 Although it sounds simple, this rule has led to some of the most cumbersome provisions of the capacity tax. Under the excise system a simple check at the time of export is usually sufficient to establish the claim and amount of the rebate, whereas under the capacity tax the rebate would be too little if actual production (assuming it is all exported) falls short of notified capacity; similarly, without explicit provisions to the contrary, too much would be rebated if exports exceed notified capacity for a particular category of fabrics or yarn.

In Pakistan excessive rebates have been the main concern. When goods are exported, the rebate is credited against the notified capacity of the exportable category of fabrics or yarn and deducted from the monthly installments due on the capacity tax. If exports exceed that notified capacity, the rebate is credited against the notified capacity of the next lower category (or if that credit is exhausted, against the capacity of the next higher category, but at the rate corresponding to the category exported). This procedure ensures that despite any increase in production beyond the notified capacity or any changes in the pattern of production from the notified pattern, the rebate remains restricted to the quantum per the notified capacity.

Obviously, the understandable desire to safeguard revenue results in substantial administrative ramifications,46 necessitating (1) a link between the exportable product and the notified capacity under which it has been produced—an administrative millstone as industry diversifies and the number of middlemen increases—and (2) the keeping of accounts for each factory, showing its capacity tax status and the rebates received. A solution would be to allow rebates on the basis of the country-wide average tax per pound of yarn or square yard of cloth, but this may unduly favor firms eligible for other rebates (for example, those granted on a regional basis). An obviously undesirable effect is that if available credits have been exhausted, manufacturers may be inclined to hold up their exports until the following year’s capacities are notified. Moreover, a shift in the composition of a firm’s export production may become more difficult because capacity estimates are based on historical production patterns.

other aspects

Through rate differentiations and exemptions, the Government has sought to use the excise system to achieve a number of social and economic objectives—such as a fairer tax burden distribution, the development of small industries and particular regions, and the control of prices.

These features have also been carried over into the capacity tax. Thus, for cotton fabrics and yarn a bracketed specific rate structure is in effect; the duty rates increase with the fineness of the fabric or the number of counts of yarn, and by extension, with the price of the taxable categories. Purportedly, the incidence of the duty would be progressive as a result of this measure.47 Prima facie, however, it is difficult to see how the progressive rate structure’s potential to improve the burden distribution can be retained once the capacity tax assessment is issued and any connection between tax payments and the nature of the taxable commodity (for example, coarse, medium-fine, or superfine fabrics) is severed. If recent tax/price ratios in the Islamabad market are an indication, the progressive effect is negligible.48 In March 1972 the tax expressed as a percentage of the retail price was 8 per cent for fine cotton fabrics but only 1 per cent lower, or 7 per cent, for coarse fabrics. Last, but certainly not least, the bracketed rate structure provides an incentive for manufacturers of coarse fabrics and low-count yarn whose tax liability has been determined accordingly to switch to finer fabrics and higher-count yarn.49

The capacity tax appears to serve about as well as the excise system for the administration of small industry exemptions and incentives to stimulate the industrialization of particular regions. At present, textile factories with not more than four power looms are exempt from the capacity tax. To allow for a lag in efficiency and a lower degree of utilization, notified capacities of cotton fabric units in East Pakistan were reduced by 20, 15, and 15 per cent successively for three years, commencing in 1968/69. In West Pakistan, the facility was limited to textile factories in Peshawar, Dera Ismail Khan, Quetta, and Kalat, which were granted a uniform allowance of 5 per cent for the entire period.50

The capacity tax rates for cotton fabrics and vegetable products are reduced “if the retail price is legibly printed or woven on the selvedge or border of every linear yard of all cotton fabrics,” or “legibly, prominently and indelibly printed” on each container of vegetable product. This curious device, introduced in 1966 and also applicable to a number of other excisable commodities, is meant to control retail prices.51 The measure has been less than successful, however, as goods are often sold above stated prices.52 As regards its feasibility, there appears to be little choice between the capacity tax and the excise duty system.

comparison of excise and capacity tax administration

The technical analysis in the preceding paragraphs permits a comparison between excise and capacity tax administration with respect to such basic processes as ascertaining the tax base, computing the tax liability, collecting and recovering the tax due, settling disputes, and ensuring the compliance and cooperation of taxpayers.

To be sure, the capacity tax obviates the need for the cumbersome production controls in effect under the traditional excise system. These controls, governed by complex provisions prescribing such things as the design of buildings in which excisable commodities are produced or stored and the movement of goods and personnel on factory premises, necessitate the continuous presence of excise staff during working hours. They involve considerable interference in the day-to-day operations of most factories. On the other hand, the capacity tax still requires excise staff to verify closures for abatement purposes, and, as an antiavoidance measure, to ascertain actual production in that connection. Moreover, prevailing arrangements require elaborate checks and balances to ensure that export rebates do not exceed liabilities for the capacity tax.

Compared to the sure and straightforward controls on production, the determination of capacity remains an inherently arbitrary exercise, no matter how much expertise and ingenuity are applied. Disagreements about the volume of actual production are of a factual nature, but disputes regarding the capacity of a plant have conceptual overtones, making them much more difficult to resolve. A taxpayer’s idea about the capacity of his factory presumably changes as often as any of its determinants. The danger then becomes real, particularly if tax rates are high, that he will regard the assessment as inequitable and will subsequently request abatement, withhold his cooperation, or resort to litigation. Whereas taxpayers hardly ever went to court in Pakistan to dispute their liability for the excise tax, many taxpayers filed petitions under the capacity tax. Obviously, the objections and appeals, as well as the increased correspondence with taxpayers and field units, require considerable time of regional and headquarters tax officials that could otherwise be spent on supervision and control.

Under the excise system collections are safeguarded by the provision that goods cannot be cleared without payment of tax, whereas the rules for the capacity tax permit the payment of the assessment in monthly installments. This deprives the tax administration of an effective enforcement tool, even though late payments carry a penalty of 1 per cent a month. Default in payments and stay orders of the courts resulted in a considerable backlog in tax collections in Pakistan, amounting to almost two thirds of capacity tax receipts in 1970/71 (15 per cent of total excise collections, inclusive of the capacity tax, or 6 per cent of total tax receipts). The considerable gains on the assessment side (particularly for the taxpayer) should therefore be weighed against the increase in appellate and tax collection work, even if the increased workload is temporary.

Tax evasion in the form of collusion between taxpayers and minor tax officials would probably not occur to the same extent under the capacity tax as under the excise system. Although there would still be the possibility of malpractices in the verification of abatements and the ascertainment of, say, the fineness of cotton fabrics as the basis for the export rebate, daily contact would be eliminated. Since actual production would not have to be reported to the excise administration, underreporting would similarly not affect the tax liability directly. However, to the extent that past production figures are one of the determinants in ascertaining capacity—and the Pakistan experience shows that they became increasingly the critical variable, perhaps because they were less open to taxpayer objection—underreporting is reflected in a lower estimate of capacity. Moreover, if past production figures continue to play a crucial role in the determination of capacity, it is important that current production data should be as accurate as possible. This in turn means that regular audits by the excise administration would still be indispensable under the capacity tax.

III. Economic Effects of the Capacity Tax

An intriguing aspect of the capacity tax concerns its usefulness as a production incentive. The originators of the tax believed that it would create a built-in reward for entrepreneurs who worked harder and more efficiently, as they would not have to pay tax on any part of their production in excess of assessed output. Since the marginal tax rate would be zero, the return on the incremental product would fully accrue to the entrepreneur himself. In their view, the penalty aspects of producing below capacity would act as a powerful disincentive, up to the point where actual production equals notified capacity.

This section examines that claim and also considers the price, income, substitution, and other effects of the tax. First, the theoretical incidence of a capacity tax is considered quite apart from the environment in which it is applied, although illustrations are drawn from the Pakistan economy. Second, a closer look is taken at capacity utilization rates of the Pakistan manufacturing sector and the attendant variables at the time the capacity tax was introduced. Third, the development of utilization rates after the imposition of the tax is reviewed for specific industries.

incidence theory

The change in the tax base may alter the nature of the tax for the individual firm. The traditional liability for excise duty varies with a firm’s production volume, whereas in principle liability for the capacity tax shows no such variations but is a fixed (lump-sum) levy. In other words, the excise duty is part of a firm’s variable costs, but the capacity tax belongs to its overhead or fixed costs. Another point of view, probably more realistic, is that the capacity tax is an excise duty on plant and equipment, or rather a property tax on business assets. The latter concept appears to apply most aptly to the cotton textile and vegetable product industries, where changes in the tax liability are a function of the number of machines. For continuous integrated production processes, such as cement, soda ash and perhaps sugar, the lump-sum tax idea has relevancy.

It appears that partial equilibrium analysis can be appropriately applied to the incidence effects of a capacity tax levied in lieu of excise duties. The Government of Pakistan did not intend to increase the amount of tax previously payable by the taxable industries; therefore, their supply and demand schedules or those of related industries should not have been appreciably affected. Moreover, factors of production would probably not change if incomes declined, in view of the highly specialized nature of the taxed industries. Finally, the number of untaxed substitutes to which demand could be diverted was small.53

The effect in theory, on output and price, of a capacity tax in lieu of a specific excise duty, may be identified as the combined effects of: (1) the abolition of an excise duty, and (2) the imposition of the capacity tax.54 In a competitive world, the abolition of an excise duty has an expansionary effect on industry’s output, while the price declines to an extent depending upon the supply and demand schedules. Under increasing cost conditions, the decrease in price will be less than the reduction in duty. Similarly, if demand is relatively elastic, price will decline less than it would if consumers were strongly attached to the product. Mutatis mutandis, these results are similar for a monopolist, except that the adjustment is affected through a change in the individual firm’s marginal cost curve rather than in the industry’s supply schedule.55

A capacity tax viewed as an addition to fixed costs does not have any effect on output and price in the short run.56 If a firm is in equilibrium before the tax is introduced, equating marginal cost and marginal revenue, this point should still determine the most profitable output (or least loss) after the tax is imposed. However, there would be long-run effects. In a competitive industry, the imposition of such a tax causes the average fixed and total cost curves to shift upward. Unable to cover total cost, marginal firms will be taxed out of business and leave the industry; subsequently, the average revenue schedule of the remaining firms will shift upward. The size of the ultimate effect will depend on long-run supply and demand elasticities, but under the assumptions the price increases along with the scale of operation of the remaining firms (the number of firms in the industry is reduced), although this need not affect their degree of capacity utilization. In the case of a monopolist, price and output do not change in the first instance, but over time, however, capital would be shifted out of the monopolist’s sector, if his profits net of capacity tax are less than the return that he can earn elsewhere (although this is unlikely if the monopolist still earns excess profits after the imposition of capacity tax). As a lump-sum levy, the income effects of the capacity tax may possibly induce entrepreneurs to increase their efforts.57

A capacity tax that is a function of the number of machines in operation, and is measured by their notional output, resembles a property tax on business assets based on physical quantities. A firm’s product would then be taxed indirectly through a production tax on one of the factors (machines) that produces it, and the incidence of such an equal yield levy on machinery would be in the nature of a specific excise duty on the product itself. The output-inducing effects of the withdrawal of the excise duty would be offset by a decline in production after the imposition of the capacity tax. Similarly, the initial decline in price would be offset by the rise in price that can be expected after the excise tax is replaced by a capacity tax. The exact extent of these effects again depends on supply and demand elasticities, and interfirm adjustments differ because of divergent machine-to-production volume ratios.58

A firm’s demand for machines—a derived demand depending on the demand for its products and the supply of substituting factors of production—will also be affected. Machinery already installed, being a highly specialized factor of production, may have to absorb part of the new tax, and its value would fall accordingly; to that extent the owners of the machinery would have to carry the burden. They may try to shift part of the new tax backward but, because much of the machinery is imported, that would probably be difficult; forward shifting is more likely. In any case, it is likely that the effective cost of the taxed factor (machinery) will increase. Assuming that the new levy fell entirely on capital, Winston estimated the tax per rupee of installed capacity and found that the effective cost increases in the cement and sugar industries were 12 and 24 per cent, respectively.59

Such an increase in factor cost may have a beneficial effect on capacity utilization, if it is assumed that its optimal level in a two-factor model depends, as postulated by Winston, on the relative prices of capital and labor.60 In this view, excess capacity may have been consciously built into capital stock as part of a perfectly rational profit-optimizing decision. As a result of the increase in the price of capital, firms will increase their desired utilization of capital stock rather than expand their existing stock; therefore, with a given investment budget, resources will be freed for capital investment elsewhere or, probably more important, for increasing raw material supply. Another effect (in most cases beneficial) will be that the increase in the cost of the taxed production factor (machines) will induce firms to substitute labor for capital. However, if capacity utilization is adversely affected by capital underpricing, direct adjustments would appear more appropriate. This could be achieved by a tax on capital equipment or, since capital is generally imported in developing countries, appropriate corrections in the exchange rate and import duty tariff.

However, the simplified assumptions underlying the foregoing analysis may be unrealistic in view of prevailing market imperfections. Thus, the existence of excess profits and monopoly prices (inter alia resulting from a scarcity of imported inputs), in conjunction with licensing and other direct controls, would make the ultimate effect of a capacity tax indeterminate.61 If entry into an industry is limited or blocked, because costs are high and credit facilities are largely in the hands of established firms, a capacity tax viewed as a lump-sum levy may strengthen the oligopoly, because the increase in fixed costs makes entry even more difficult. On the other hand, it can also be argued that the capacity tax makes production restrictions more costly and therefore forms an incentive for the individual firm to break the “agreement.” 62 Imported raw material inelasticities may mean that a firm’s marginal cost curve is kinked and intersected by the marginal revenue curve within the undefined range.63 The abolition of the excise duty would then cause the marginal cost curve to shift downward, but the output associated with the equilibrium level of production does not change; neither does price. Profits would increase and a capacity tax subsequently imposed would be paid out of profits with no salutory effect either way.64 Such a situation limits the effectiveness of a capacity tax, although it should be noted that the Pakistan Government excluded from the tax the industries dependent on imports. This point is further developed below.

explanatory variables of capacity utilization

Although the production incentive was one of the main reasons for imposition of the capacity tax, the industries to which this novel tax was applied were among those with the highest capacity utilization rates of the entire manufacturing sector. Cotton textiles, sugar, cement, and vegetable products all fall within the top one third of the ranking computed by Winston (see Table 4, column A). Cotton textiles, Pakistan’s largest industry by number of units and total turnover (Table 4, columns C and D), has the second highest utilization rate, operating at 70 per cent of capacity (94 per cent if not adjusted for the number of shifts) in 1965/66.

Table 4.

West Pakistan: Industrial Capacity Utilization Data, 1965/66

article image
Source: Gordon C. Winston, “Capital Utilisation in Economic Development,” Economic Journal, Vol. 81 (March 1971), p. 55.

These aggregated rates are a weighted (by capacity) average of industry utilization rates computed from annual production (value) and annual production capacity (adjusted to a two and a half shift level if the industry worked less than that) reported in Census of Manufacturing Industries 1965/66 (unpublished, West Pakistan Provincial Ministry of Industries).

Computed as in column A, but without adjustment of annual production capacity to a two and a half shift operation.

Summed for each sector.

Annual production in millions of rupees per reporting unit.

The proportion of the value of total raw material inputs purchased from abroad.

Imports as a proportion of total supply.

As a proportion of gross output at factor cost, f.o.b.

From Winston’s analysis, it is clear that the above-mentioned industries and others with high utilization rates have important characteristics in common that enable them to utilize their resources more fully even in the absence of tax incentives. First, they are less dependent on imported raw materials than most other industries (column E) and are therefore not subject to the perennial foreign exchange constraints or the vagaries of a raw materials licensing system that was administered on a rated capacity basis, thereby inducing the creation of excess capacity.65 Winston found that for every 1 per cent decrease in dependence on imported raw materials, capacity utilization rose by 0.267 per cent.66 A second characteristic of these industries is that they do not face demand competition from imports to the same extent as industries with lower uitlization rates (Table 4, column F). But there may also be a negative relationship on the supply side; if import substitution policies favored the creation of industries with competing imports, it is likely that at least some excess capacity was created.67

A number of other characteristics, such as exports (inducing demand expansion—Table 4, column G), firm size (economies of scale), capital-income ratios (reflecting profit-maximizing adjustments to prevailing cost patterns) and rates of growth, are also positively related to capacity utilization. More interestingly, competitive firms appear to have higher utilization rates than industries with only a few firms that can substitute either inventory accumulation or excess capacity for price fluctuations when faced with changes in demand.68 The textile industry is an interesting example: it is one of the most competitive industries in Pakistan and also has one of the highest utilization rates. Moreover, competition may have prevented manufacturers from installing excess capacity by importing (overinvoiced) capital goods in excess of requirements, which was considered an easy, although illegal, means of transferring capital abroad.69

Thus, by using independence from raw material imports as a criterion for choice, the Government in effect applied the capacity tax to those industries that did not need an incentive to increase capacity utilization. On the other hand, firms with low utilization rates may not be able to improve their performance, regardless of the incentive offered, because of constraints on input supply; however, since the tax was not levied on these firms, its practical effectiveness cannot be ascertained. As a general point, it may be stressed that capacity production presupposes that demand is strong enough to clear the market of the goods produced at capacity level operations and, more importantly in most cases, that there will be an adequate flow of variable inputs—raw materials and intermediate goods—at current prices. The effectiveness of any device used as a production incentive will be limited by the extent to which these two conditions are satisfied.

utilization rates after the capacity tax

Nevertheless, it would still be of interest to see whether any changes occurred in the capacity utilization rates of the industries to which the capacity tax applied in Pakistan. Without taking into consideration technological improvements, obsolescence factors, and changes in efficiency, Table 5 shows capacity outputs computed on the basis of standard machine ratings, number of shifts (days of production), and other technical data for the cotton textiles, sugar, and cement industries. It should be noted that the standards used are peculiar to each industry and do not conform to those employed by Winston, who in calculating capacity disregarded the firms’ ideas of full utilization, arguing that these did not necessarily reflect the presumed capital scarcity; instead, he took two and a half shift operations, each of eight hours duration, as the norm.70 In Table 5, for cotton fabrics the number of shifts has been set at the standard used by the Excise Department. Data for machine outputs conform to those used by provincial governments, although corresponding figures underlying the capacity estimates of the Excise Department for all of Pakistan are lower—being 190 pounds per spindle and 21.250 square yards per loom. For sugar and cement, the provincial governments’ standards were adopted.

Table 5.

West Pakistan: Actual and Capacity Production, and Capacity Utilization Rates for Selected Industries1

article image
Sources: Central Statistical Office, Monthly Statistical Bulletin, Vol. 16 (September 1968) and Vol. 20 (July 1972); and Government of the Punjab, Bureau of Statistics, Development Statistics of Punjab, Sind, N. W. F. P., and Baluchistan, October 1971.

Cotton yarn and fabrics became subject to the capacity tax on May 1, 1968; sugar and cement, on July 1, 1966.

240 pounds per spindle working 1,000 shifts per annum.

30,000 square yards per loom working 897 shifts per annum.

Average seasonal run of 200 days, downtime during the season 10 per cent, and assuming a recovery rate of 8.75 per cent.

Continuous operations of 330 days a year.

Estimates.

The development of utilization rates, that is, actual production as a percentage of capacity output, is illustrated in Chart 1. The cotton textile industry has slightly higher utilization rates under the capacity tax than before, although the upward trend is already discernible during the three years preceding the tax. The rates for the sugar and cement industries show no noticeable long-term improvement. More importantly, the trends and fluctuations in utilization rates can be largely explained by supply conditions and do not appear to be related to the capacity tax. Cotton textile production in 1965/66 inevitably declined because of crop diseases and the smaller area planted in the previous year on account of adverse weather conditions; the decline was also compounded by industrial unrest and breakdowns in power supply. However, the upward trend in production and capacity utilization resumed in 1966/67, when the supply of raw cotton increased because of improved imputs and the greater availability of irrigation facilities, as well as rising demand in export markets. After 1967/68, acreage and production leveled off somewhat, but record production levels were reached in the early 1970s, reflecting the farmers’ response to higher cotton prices following the 10 per cent government bonus of 1970.71 In addition to supply problems, capacity utilization of many mills is adversely affected by a lack of knowledge about textile engineering, inefficient use of machinery, absence of proper maintenance and temperature controls, lack of standardization and specialization (by range of counts and variety of fabrics), and ineffective managerial supervision. More generally, this suggests that the improvement of industrial extension services will enable firms to do what the capacity tax is supposed to force them to do. Again, the agricultural parallel is striking.

Chart 1.
Chart 1.

West Pakistan: Capacity utilization Rates, 1964/65–1971/72

Citation: IMF Staff Papers 1974, 001; 10.5089/9781451956375.024.A006

Sources: Central Statistical Office, Monthly Statistical Bulletin, Vol. 16 (September 1968) and Vol. 20 (July 1972); Government of the Punjab, Bureau of Statistics, Development Statistics of Punjab, Sind, N. W. F. P., and Baluchistan, October 1971.

The utilization rates in the sugar industry follow sugarcane supply patterns exactly. After bumper crops were harvested in 1965/66 and 1969/70, production and utilization rates increased. On the other hand, in the first year that the capacity tax was imposed, lack of winter rains and heavy frosts damaged the crop so severely that the normal recovery rate dropped considerably. The effect was carried over into the 1967/68 season, because the seeds were adversely affected. The sugar mills are also subject to supply and demand restrictions. Besides facing transportation holdups that reduce the sugar content of cane, mills cannot procure sugarcane from outside the zone alloted to them by the provincial governments. On the demand side the Central Government often prohibits the sale of sugar on the open market in order to control prices, causing stocks to accumulate at factories and causing farmers to divert sugarcane to gur or khandsari (crude sugar made outside the mills).

The utilization rates in the cement industry cannot be explained by constraints on input supply. Production increased steadily in the period under review, concurrent with the rise in demand because of the large-scale building programs that were initiated. The supply of raw materials was abundant, and capacity utilization rates were high.72

IV. Tax Revenue Implications

The response of the tax system to changes in income is an important concern in developing countries, since an income-elastic tax system enables governments to finance their growing expenditures in a noninflationary way. In Pakistan, excise taxes have shown a remarkable degree of income elasticity. Over a ten-year period ending 1970, revenue from this source increased 4.4 per cent for every 1 per cent growth of the economy, in contrast to the much lower income elasticity (1.2) of all other taxes and duties. The exceptional increase in excise duty collections can be accounted for by the fast growth of the domestic manufacturing sector and the frequent upward adjustments in tax rates.73

Prima facie, a capacity tax seems relatively less elastic than specific excise duties. Both the tax rate and the tax base (quantum of production) remain the same as the economy expands, provided investment in new machinery does not lead to an immediate revision of the tax liability. An attempt has been made to find some support for this proposition by computing quantitative production indices and comparing these with the indices of tax receipts as adjusted for rate increases that took place at the time of and after the introduction of the capacity tax (Table 6).74 Over a five-year period prior to capacity tax, cement excise collections increased by 58 per cent. At the same time, the production index showed a gain of 28 per cent. Thus, the flexibility ratio of excise to index of manufacturing is 2.07.75 A similar computation for a four-year period following the introduction of the capacity tax yields a ratio of 1.90. For sugar and vegetable products the ratios are 0.87 and 0.68 for the excise tax periods and 0.07 and 0.58 for the capacity tax periods (covering four and two years respectively). For cotton fabrics, the ratio is negative (−5.32) under excise duties, but receipts became even more inflexible (−6.28) after the capacity tax was introduced.

Table 6.

Indices of Production and Tax Receipts of Items Subject to the Capacity Tax

article image
Sources: Production figures: Central Statistical Office, Monthly Statistical Bulletin, various issues. Tax receipts: Ministry of Finance, Pakistan Budgets, 1969–70, p. 24, and The Explanatory Memorandum on the Budget, 1971–72 (pp. 4 and 5). Tax rates: Ministry of Finance, The Budget in Brief, 1969–70, pp. 78–91; 1971–72, pp. 121–30; and 1972–73, pp. 61–69.

According to the formula: Adjusted index of tax receipts = Index of actual tax receiptsIndex of tax rates×100(1961/62=100).

Based on revised estimates; not included in the flexibility ratio calculations.

Thus, the capacity tax appears to have been a less flexible source of revenue than excise duties, although allowance has to be made for the growing tax arrears that may not have reached a stable level. Whether this should be considered a drawback of the experiment depends on what happens to other taxes. If the capacity tax reduces the optimal capital stock-output ratio, allowing income to grow faster with a given investment budget, the base of other taxes would expand with yield effects that may offset the greater revenue inflexibility of the capacity tax.

To increase tax collections, the Government of Pakistan could adjust the tax base periodically or raise the tax rate. In practice, capacity estimates were revised rather frequently in Pakistan, reflecting decisions on appeals to the Review Board (particularly by the sugar and vegetable product industries; by the end of 1970/71 no decisions had yet been taken on appeals by the cotton textile industry), as well as direct action by the Central Board of Revenue, which undertook a full-scale revision of capacities in the cement industry.76 Of course, frequent adjustments of capacities work against the zero marginal rate effect of the tax. On the other hand, changes in the tax rate would not have that effect. To adjust rates automatically, they could be linked to price changes.77 Such a system applies to land revenue collections in Sind Province, permitting adjustments of tax liabilities in step with changes in the average market price of the products over a defined period.78 As regards the capacity tax, over the period under review only the rate applicable to vegetable products was increased—from 40 to 45 rupees per hundredweight of rated capacity in 1970/71.79

V. Summary and Conclusions

It is now possible to sum up some of the arguments and to evaluate the capacity tax in the light of the objectives it was purported to serve: (1) administrative simplification and convenience, and (2) increased capacity utilization. These conclusions regarding Pakistan are augmented by a few more general considerations.

incentive aspects

Partial equilibrium analysis suggests that if the capacity tax works as a lump-sum levy, there should be no short-run effects on output, except those induced by the income effect of the fixed levy. The long-run effects are largely indeterminate; although under competitive conditions the size of operations may increase, this need not involve a higher degree of capacity utilization. If the capacity tax is a function of the number of machines in operation, resembling a property tax on business assets, cost and supply curves of the final product are likely to be affected by its imposition; price will rise and output will decline. If it is assumed that the desired level of capacity utilization depends on the relative prices of capital and labor, by raising the price of capital a capacity tax will increase utilization of capital stock and induce entrepreneurs to substitute labor for capital. To make an overall judgment on incidence in Pakistan, the effects of the withdrawal of the excise duty that was previously chargeable—leading to an increase in output and a decline in price—should be taken into account as well.

These theoretical considerations presuppose that there are no constraints on the supply of raw materials and intermediate goods. However, in developing countries, import-substituting industries may depend heavily on imports of raw materials. If these cannot be procured in adequate quantities at prevailing prices because of a limited foreign exchange budget, firms may not be able to utilize existing capital stock fully regardless of the incentive offered. For this reason, Pakistan’s capacity tax was applied to industries that used only indigenous raw materials. Since it appears that these industries were already operating at the highest utilization rates in the entire manufacturing sector, the efficacy of the capacity tax as a production incentive cannot be readily ascertained. From the evidence it may be surmised, however, that the removal of constraints on supply rather than the application of a production incentive would solve to a great extent the basic policy conundrum.

Of course, neither production incentives nor adequate supplies can provide a solution for excess capacity resulting from inefficiencies in policies and planning, although a capacity tax in the form of a property tax on business assets could correct a relative underpricing of capital goods and thereby inhibit further unjustified expansion.80 Such a measure is never as effective as direct recourse to the price of the scarcity factor, the exchange rate itself, because it is not as comprehensive. In Pakistan, the complex and overvalued exchange and trade system, involving multiple currency practices and widespread restrictions, was overhauled in May 1972—accompanied by a substantial devaluation of the rupee and some import liberalization.81 In the same year the tax holiday scheme was abolished, after having been suspended in the previous year. Over time these measures should redress the imbalances caused by the relative underpricing of capital and should give the economy a chance to grow up to capacity. Furthermore, liberalized imports and commodity loans should have a favorable effect on capacity utilization by increasing the supply of raw materials.

ADMINISTRATIVE FEATURES

Clearly, by its nature, the proper measurement of capacity is a very complicated exercise requiring expert accounting and engineering skills if it is not to become, like some methods of presumptive income taxation, “mainly a guessing game organized according to variable rules.”82 Even if machine ratings or peak forward projections from past and comparative production data provide useful approximations, some inherently arbitrary elements remain; this is a crucial flaw in the new scheme. Obviously, each manufacturer will have his own ideas about the capacity output of his plant; moreover, these will probably change as often as any of the other factors that determine capacity.

In Pakistan, virtually unavoidable ambiguities in capacity concepts proved difficult to reconcile with the taxpayers’ ideas of equity. This resulted in frequent recourse to courts, pressures to set up committees to review capacities, increased correspondence and file work, and delays in tax collections—thus nullifying most of the gains made at the assessment stage. This process may be repeated if capacities are reassessed. In this respect, the capacity tax resembles a real estate tax, under which a taxpayer’s sense of equity is usually not satisfied until his assessment is reduced to that of his neighbor who occupies a similar abode, with the result that tax demands are often adjusted to the level of the lowest assessment. In addition, conceptual difficulties led to increased reliance on past production data as the crucial variable in determining capacity output, thereby reintroducing the biases caused by underreporting in the excise system. Of course, if past production data are the primary determinant, the excise administration must verify their reliability, and that can only be done by thorough audit work. Under the capacity tax there really seem no shortcuts that enable the tax administration to avoid the problems, first of proper initial assessment, and second, of adequate updating and verification.

On the other hand, a capacity tax interferes less with the utilization of capacity than a conventional excise tax. This is a considerable advantage. Production controls are probably acceptable when excise taxes are a relatively minor source of revenue levied mainly on a few homogeneous sumptuary commodities, but they become an archiac tool of tax enforcement, completely disregarding the taxpayer’s convenience, when industry expands and diversifies. Of course, the thwarting restrictions thus placed on manufacturers can be removed in other ways, such as changing and expanding the tax base to a general sales tax and shifting the verification process to account books, thereby modernizing a country’s manner of conducting public business. For instance, to facilitate the transition of an extended excise system with production controls to a manufacturers’ sales tax relying on documentary verification, a capacity tax might be employed as an interim device, providing the tax department with a valuable interlude for training its personnel in audit and accounting techniques, for making the public aware of oncoming changes, and for actively educating taxpayers in proper record keeping.

A second case in which a capacity tax can be usefully employed is that of a country which has only recently turned toward industrial development. Typically, the first industrial ventures relate to products that were previously largely imported and that usually contributed heavily to customs revenue. The loss must be made up through the imposition of domestic levies. Excises are generally considered the appropriate form of taxation, but a capacity tax is a viable alternative, particularly as the country will initially be short of supervisory and audit staff required for administering the excises. Machine ratings can be established without much difficulty (and the tax liability computed accordingly), since the new industries generally engage in the manufacture of basic foodstuffs, such as sugar or margarine, beer, soft drinks, soap, tobacco products, cement, and similar products. But in the light of Pakistan’s experience, rules would have to ensure a limited application of relief provisions, inter alia putting on the taxpayer the onus of proving that actual production falls short of capacity output by a specified margin.

Of course, the capacity tax would always be useful as an administrative tool for ascertaining the excise or sales tax liability of manufacturing units that are not willing to keep adequate accounts nor to disclose to the tax department their sales or any other information required for tax purposes. However, to discourage such practices and to encourage the keeping of books of account, estimated assessments would have to be on the high side, and this practice would be resisted by the taxpayers. Finally, for taxpayers who cannot be expected to keep adequate records, such as small textile manufacturing units, the capacity tax could serve as a proxy for the excise or sales tax liability otherwise payable.83

evaluation

Pakistan’s case shows that tax reformers should consider the substantial uncertainties affecting both taxpayers and the tax administration whenever new taxes are introduced or significant changes are made in old levies. Taxpayers cannot be expected to accept willingly measures which they regard as being confiscatory. New laws and procedures are bound to lead to an increase in litigation, and legitimate pressures will strongly increase if the determination of tax liabilities is at all arbitrary. If a number of firms are successful in getting their tax liabilities reduced or in delaying their payments, others will follow to keep their competitive position. This concerns the certainty aspect of taxation.84 The public’s respect for the law is undermined by the arbitrary nature of tax provisions, and, more generally, by continuously changing tax regulations, procedures, and practices. To be respected, tax provisions must be precise and must be changed as little as possible, so that a country’s citizens may know their content and accept them either out of fear or custom.85

This point can be usefully illustrated by the Canadian production incentive tax, which had uncertain economic and operational effects, with subsequent undesirable repercussions on business decisions, departmental efficiency, and effectiveness. Cumbersome antiavoidance sections had to be devised to prevent noneligible integrated companies from splitting up. In addition, there was discrimination against unincorporated businesses excluded from the benefits, while firms with sharply fluctuating sales were unintentionally favored over firms whose sales increased at the same rate each year. Moreover, price as well as volume increases were rewarded, although the former obviously did not contribute to higher output.86

To conclude, it should probably be emphasized that if a government’s objective is to increase production, an essential requirement is the provision of a suitable macroeconomic framework in which business activities can be carried on. Disincentives and imperfections which hinder an efficient functioning of the market mechanism should be removed before incentives to increase production are added to the government’s tax policy arsenal; even then, the potentially beneficial effects of the incentives should be carefully weighed against the costs of the uncertainties and administrative complications that are inevitable. An important, if unintended, effect of the capacity tax is that it may increase capacity utilization through an upward correction of the price of capital. The tax also has some merits on administrative grounds, particularly as a transitional device. Pakistan’s experience with this novel tax device may have useful lessons for other developing countries.

*

Mr. Cnossen, a Senior Economist in the Fiscal Affairs Department, is a graduate of the Netherlands Academy of Taxation and Erasmus University, both in Rotterdam. He was formerly an income tax inspector with the Netherlands Ministry of Finance and is author of The Indonesian Sales Tax. In addition to thanking staff colleagues for their assistance, the author wishes to express his appreciation to Gordon C. Winston, Williams College, for his comments on an earlier draft of this article.

1

Gordon C. Winston, “Capital Utilisation in Economic Development,” Economic Journal, Vol. 81 (March 1971), pp. 36–60.

2

Gordon C. Winston, “Overinvoicing, Underutilization, and Distorted Industrial Growth,” Pakistan Development Review (Winter 1970), p. 416

3

For a review of the literature, see Derek T. Healey, “Development Policy : New Thinking About an Interpretation,” Journal of Economic Literature, Vol. 10 (September 1972), pp. 757–97, with a summary of capacity utilization data for 13 developing countries (p. 785).

4

Winston, “Capital Utilisation in Economic Development” (cited in footnote 1), p. 58.

5

For a summary and discussion of tax incentive studies relating to developing countries, see Jack Heller and Kenneth M. Kauffman, Tax Incentives for Industry in Less Developed Countries, Harvard Law School, International Program in Taxation (Cambridge, Massachusetts, 1963), and George E. Lent “Tax Incentives for Investment in Developing Countries,” Staff Papers, Vol. 14 (July 1967), pp. 249–323. Depreciation allowances can also be employed to increase capacity utilization by permitting higher write-offs for multiple shifts. For references to this practice, see Richard M. Bird, Taxation and Development: Lessons from Colombian Experience (Harvard University Press, 1970), p. 86; Richard Slitor, “Reform of the Business Tax Structure: Analysis of Problems and Alternative Remedial Proposals,” in Fiscal Reform for Colombia, ed. by Richard A. Musgrave and Malcolm Gillis, Harvard Law School, International Tax Program (Cambridge, Massachusetts, 1971), pp. 484–85; and Lauchlin Currie, Accelerating Development: The Necessity and the Means (New York, 1965), p. 111.

6

For a useful description of these agricultural income taxes, see George E. Lent, “Taxation of Agricultural Income in Developing Countries,” Bulletin for International Fiscal Documentation, Vol. 27 (August 1973), pp. 324–42. In the specific case of Colombia, an interesting and carefully argued study concludes: “Overall, it seems probable that a presumptive income tax on land is a relatively ‘safe bet’ policy in terms of the output and distribution goals and in terms of the low probability of its having negative results. …” (See R. Albert Berry, “Presumptive Income Tax on Agricultural Land: The Case of Colombia,” National Tax Journal, Vol. 25 (June 1972), p. 180).

7

Resolution of the Third Inter-American Conference on Taxation, Mexico City, September 3–8, 1972.

8

Joan Robinson, The Economics of Imperfect Competition (London, 1950), p. 163. See also A. C. Pigou, A Study in Public Finance (London, 1949), Chapter 8; Alfred Marshall, Principles of Economics (London, 1961), Book 5, Chapter 14; Hugh Dalton, Principles of Public Finance (London, 1954), p. 42; and Ned Shilling, Excise Taxation of Monopoly (Colombia University Press, 1969), pp. 44 and 99–101. In an analysis of Mrs. Robinson’s scheme, Higgins proposed as an alternative “a tax on profits in excess of a fair return on utilized capacity,” which in his view would approximately attain the same objective and be easier to administer than the tax-and-bounty scheme. As long as the rate of net (after tax) profits rises with output, it would be advantageous “not only to utilize existing plant and equipment more fully but to operate the part that is utilized as efficiently as possible.” See Benjamin Higgins, “Fiscal Control of Monopoly,” in Readings in the Economics of Taxation, Vol. 9, ed. by Richard A. Musgrave and Carl S. Shoup (Homewood, Illinois, 1959), pp. 312–21.

9

Klaus Knorr and William J. Baumol, eds., What Price Economic Growth? (Englewood Cliffs, New Jersey, 1961).

10

See Richard Bird, “A Tax Incentive for Sales: The Canadian Experience,” National Tax Journal, Vol. 18 (September 1965), pp. 277–85.

11

Vito Tanzi, “Theory of Tax Structure Development and the Design of Tax Structure Policy for Industrialization,” paper presented at the Conference on Fiscal Policy for Industrialization in Latin America, February 17–20, 1971, held at the Center for Latin American Studies, University of Florida (unpublished), pp. 16–17.

12

Ministry of Finance, The Budget Speech for 1966/67, pp. 16–17. (Hereinafter referred to as Budget Speech.)

13

Because of the seasonal pattern of production, sugar manufacturers were given the option of paying the total annual amount in eight equal installments beginning in November of each year.

14

Act No. XI of 1966,” Gazette of Pakistan, June 30, 1966, pp. 545–46.

15

Pakistan has excellent budget documents with details of the tax system. The Ministry of Finance issues the Budget Speech, The Budget in Brief, and The Explanatory Memorandum on the Budget annually. For an analysis of indirect taxes, including excises, for the period 1954/55–1962/63, see Stephen R. Lewis, Jr., and Sarfraz Khan Qureshi, “The Structure of Revenue from Indirect Taxes in Pakistan,” Pakistan Development Review, Vol. 4 (Autumn 1964), pp. 491–526, and Ghulam Mohammed Radhu, “The Rate Structure of Indirect Taxes in Pakistan,” ibid., pp. 527–51.

16

Thus for four products—cement, sugar, vegetable products, and textiles—the capacity tax rate was made up of the old excise duty, the sales tax (a 20 per cent levy at the manufacturing and importing stages), and two surcharges levied for defense and the rehabilitation of refugees.

17

“Central Finance Minister Defends the Budget in the National Assembly” in Budgets of Pakistan, 1966–67 (prepared by the Economic Adviser, Ministry of Finance), p. 71. It is worth mentioning that a reduction in revenue was also not envisaged. In this respect the capacity tax, like a presumptive income or land tax in the agricultural sector, differs from the usual tax incentive that sacrifices revenue.

18

Budget Speech for 1961–62, pp. 23–24.

19

Reportedly, the incentive became so strong that in 1965 the exempt, small industry sector comprised 13,000 looms as against 35,000 looms in the mill sector (Budget Speech, 1965–66, p. 26). For similar avoidance techniques in the sugar industry, see Government of Pakistan, Taxation Enquiry Committee Report, Vol. 1 (1961), pp. 180–81.

20

Budget Speech for 1959–60, p. 19. However, units with not more than four power looms remained exempt from tax. A similar administrative expedient was adopted in India in 1964 (Government of India, Budget Speech for 1964–65, part B, p. 4).

21

Budget Speech for 1963–64, p. 15.

22

Budget Speech for 1964–65, p. 27. Another interesting production incentive, similar to the Canadian experiment cited above, was proposed at budget time. It involved a rebate of the excise duty on tea for outputs above the highest level of production attained during the preceding four years. For small tea gardens with less than 500 acres under cultivation, the rebate would be 50 per cent and for larger gardens, 25 per cent (ibid., p. 26).

23

Ibid., p. 27, and Budget Speech for 1965–66, pp. 27–28.

24

Pakistan Ministry of Finance, Central Board of Revenue, Central Taxes in Pakistan: A Decade of Development and Self Effort (Islamabad, 1968), p. 16.

25

“A Talk by the Secretary, Ministry of Finance, Broadcast from Radio Pakistan, Rawalpindi, on 12th June 1966,” Pakistan Budgets 1966–67, p. 205.

26

Ministry of Finance, The Budget, 1969–70, p. 33 (hereinafter referred to as The Budget) and Ministry of Finance, The Budget in Brief, 1972–73, p. 28 (hereinafter referred to as The Budget in Brief).

27

Budget Speech for 1966–67, p. 16. The collusion argument was also voiced by the Finance Secretary on June 13, 1964 (Pakistan Budgets, 1964–65, p. 211). See also A. Farouk, “Pakistan Budget for 1965–66,” in Pakistan Budgets, 1965–66, p. 238.

28

Winston, “Capital Utilisation in Economic Development” (cited in footnote 1), p. 60.

29

Gordon C. Winston, “Capacity Taxation,” (unpublished, Pakistan Institute of Development Economics, Karachi, November 1967), p. 2.

30

Budget Speech for 1966–67, p. 20.

31

Budget Speech for 1967–68, p. 15.

32

Pakistan Ministry of Industries and Natural Resources, Report of the Textile Industry Capacity Committee (unpublished, February 1968).

33

Budget Speech for 1970–71, p. 16.

34

Ibid., p. 17.

35

Dawn, November 22, 1972 and Gazette of Pakistan, October 14, 1972.

36

Central Statistical Office, Monthly Statistical Bulletin, Vol. 20 (July 1972). (Hereinafter referred to as Monthly Statistical Bulletin.)

37

See Richard E. Gift, Estimating Economic Capacity: A Summary of Conceptual Problems (University of Kentucky Press, 1968), pp. 30–31.

38

Thus, in the United States, the McGraw-Hill Company, arguing that “… companies follow a commonsense definition of capacity, such as maximum output under normal work schedules,” accepts the aggregated composite response of what manufacturers themselves regard an appropriate level of output. The Wharton School Econometrics Unit, on the other hand, ascertains capacity from the peaks of seasonally adjusted, averaged quarterly production figures. Capacity for any period is then the straight line from one peak to another. (For forecasting purposes, the last straight line is extrapolated.) A third measure, used by the Industrial Conference Board, involves the computation of constant dollar book values of fixed capital which are in turn combined with deflated and inventory-adjusted output data into capital-output ratios. Peak ratios indicating “virtually full capacity” utilization are selected by inspection, and other output rates are measured with reference to the peak capital-capacity output ratio. Fourth, it is sometimes assumed that capital outlays themselves reflect capacity utilization. Finally, engineering or historical data are used to establish production functions—the relationship between the quantity of output and the quantities of various inputs, for example, machine ratings—that offer a basis for determining physical production capacity. See Almarin Phillips, “An Appraisal of Measures of Capacity,” American Economic Association, Papers and Proceedings of the Seventy-fifth Annual Meeting, American Economic Review, Vol. 53 (May 1963), pp. 275–92.

39

It will be noted that this closely parallels similar methods of determination of standard yields under some agricultural taxes. Fluctuations are also inevitable in the crushing season, which may last from 120 days to 200 days, but this was taken care of by an amendment of the sugar capacity rules to the effect that a day on which no crushing of sugar from cane took place in the factory was deemed to be a day of closure, making the factory eligible for abatement. See Statutory Rules and Orders (S.R.O.) 1984 (K)/68 and the section of this paper on Abatements, below.

40

United Nations Economic and Social Council, The Tax System in Pakistan: Studies in Tax Reform Planning (July 19, 1971), p. 40 (revision of a study originally prepared by Nurul Islam, Director, Pakistan Institute of Development Economics, Karachi). For the rebates, see S.R.O. No. 61 (R)/68, rule 3, Gazette of Pakistan, April 22, 1968, p. 198; the common denominator of the regional reliefs is taken as the country-wide rebate. For cotton yarn a one-time 10 per cent rebate was given for 1968/69.

41

For brevity, reference is mainly to cotton fabrics in the following paragraphs. The only difference between the capacity tax rules for cotton fabrics and yarn lies in the kind of machines (looms versus spindles) and the tax rate structure (fineness per square yard versus counts per pound).

42

S.R.O. No. 61(R)/68, rule 3(9)–(11) op. cit. Further implementing rules prescribed that each manufacturer had to maintain a daily account of each working shift, showing inter alia their time of commencement and termination, the number of looms worked, and the quantities of each category of cotton fabrics manufactured during the shift.

43

S.R.O. No. 61(R)/68, rules 4 and 5, op. cit.

44

Winston, “Capacity Taxation” (cited in footnote 29), p. 6.

45

In the case of cotton fabrics, an additional rebate is granted ranging from Rs 0.55 a pound to Rs 2.80 a pound, depending on the fineness of the fabric and reflecting the estimated incidence of the duty on yarn; see Table 3 and S.R.O. No. 61(R)/68, rule 8(3), op. cit.

46

The rebate problem is particularly difficult to solve in the case of fabrics containing mixed counts or mixed fibers. Initially, when fabrics with mixed counts were exported, a rebate was allowed on the basis of average count, but with revenue apparently suffering, the rules were revised and rebate is now granted on the basis of the actual weight of each category of yarn. To qualify for the rebate, the exporter has to keep a record of the number of ends per inch and counts of warp yarn, as well as the number of picks per inch and counts of weft yarn, even though the administration continues to draw samples from export consignments. See S.R.O. No. 3(1)/71, Gazette of Pakistan, January 4, 1972.

47

See Budget Speech for 1969–70, pp. 8–13. In the excise system a large number of other bracketed specific and ad valorem rates employ quality, price, or other differential characteristics of the taxable commodities. Products that are taxed in this way include tea, cigarettes, soap, cosmetics, knitting wool, electric batteries, bulbs, tubes, and lubricating oil.

48

Of course, the crucial underlying assumption is that the tax is actually passed on to the consumer; presumably this is largely the case for the highly competitive cotton textile products. Another point is that bracketed rate structures may be deceiving because rates which are progressive in relation to quality may only be proportional or even regressive in relation to price.

49

Similarly, small changes at the margin (for example from 34 to 35 counts of yarn) involve an increase in the export rebate from Rs 1.10 to Rs 1.75 per pound. The share of fine and medium cloth in total production increased by 7.5 per cent in West Pakistan over a four-year period ended 1971/72, while coarse cloth output declined correspondingly. Data computed from Monthly Statistical Bulletin, Vol. 20 (July 1972.)

50

The regional development incentives were in addition to the general relief mentioned above. For yarn units, a 10 per cent rebate was granted to factories in East Pakistan in 1968/69 and 1969/70. During the same period, specified units in West Pakistan received rebates of 5 and 10 per cent. Again, rated capacities could not be reduced below average actual production in the three years prior to the imposition of capacity tax. See S.R.O. No. 61(R)/68, rule 3(3), (4), and (5), op. cit.

51

Budget Speech for 1966–67, p. 18, but in particular Budget Speech for 1969–70, pp. 8–13. For textiles, see S.R.O. No. 120 (I)/69, Gazette of Pakistan, June 28, 1969.

52

Budget Speech for 1972–73, p. 46.

53

Carl S. Shoup, Public Finance (Chicago, 1969), pp. 9–10. Sugar may be an exception to the condition that there should be no untaxed substitutes. Cane is often diverted to the production of gur or khandsari, crude forms of sugar with a very low recovery rate.

54

In the following analysis, the effect of differences in assessment and collection methods is ignored. (Excises are payable at the time of sale, but capacity tax is assessed annually and payable in monthly installments.)

55

The effects are opposite to the imposition of an excise duty, the case generally examined in literature. See Richard A. Musgrave, The Theory of Public Finance: A Study in Public Economy (New York, 1959), pp. 288–89.

56

Ibid., p. 210.

57

In “Capital Utilisation in Economic Development” (cited in footnote 1), p. 57, Winston says: “In making it more expensive to let plant sit idle, the tax may increase utilisation through an income effect on entrepreneurial utility maximisation such that managerial input may increase as profits are reduced.” For supporting arguments he refers to Tibor Scitovsky, Welfare and Competition: The Economics of a Fully Employed Economy (Homewood, Illinois, 1951), pp. 142–47.

58

For the treatment in this and the following paragraph, reference may be made to Shoup, Public Finance (cited in footnote 53), pp. 273–79 and 394–99.

59

Winston, “Capacity Taxation” (cited in footnote 29), p. 11. Since both sugar and cement operated at high levels of capacity when Winston made this calculation, it was only necessary to express the tax as a percentage of sales, so that the product of this ratio and the inverse of the capital output ratio (computed in connection with another study) gave an estimate of the tax per rupee of installed capacity. In this form, the tax is expressed as a yearly charge on installed capacity—an addition to the effective interest rate.

60

Two other variables are the capital intensity of the production process and the elasticity of substitution. See Gordon C. Winston, ‘The Four Reasons for Idle Capital” (unpublished, Williams College/Nuffield College, September 1971) and his pioneering study “A Primer on Pure Flow Production Analysis” (unpublished, Williams College, January 1973).

61

For Pakistan, the prevalence of oligopolistic tendencies is argued persuasively by many authors. For example, see Mohammed Yaqub, “The Elasticity of Taxes in a Developing Country—A Case Study of Pakistan” (unpublished, Pakistan Institute of Development Economics, Karachi, 1966), as quoted in M. Z. Farrukh, “Tax Concessions to Industries—An Overview of Pakistan’s Experience” (unpublished, Harvard University Law School, International Program in Taxation), p. 26; also Lewis and Qureshi, ‘The Structure of Revenue from Indirect Taxes in Pakistan” (cited in footnote 15), p. 498.

62

Winston, “Capacity Taxation” (cited in footnote 29), p. 9.

63

See Ghulam Mohammed Radhu, “The Relation of Indirect Tax Changes to Price Changes in Pakistan,” Pakistan Development Review, Vol. 5 (Spring 1965), pp. 54–63.

64

Compare also Budget Speech for 1970–71, p. 13, in which the Government of Pakistan indicated that it expected that the duty on paper and paperboard would be paid out of profits and would not be passed on to the consumer.

65

Winston, “Capital Utilisation in Economic Development” (cited in footnote 1), p. 48. For an analysis of Pakistan’s import substitution policies, see also Stephen R. Lewis, Jr., Pakistan: Industrialization and Trade Policies (Oxford University Press, 1970), in particular Chapter 5; also Lewis and Stephen E. Guisinger, “The Structure of Protection in Pakistan,” in Bela Balassa and Associates, The Structure of Protection in Developing Countries (The Johns Hopkins Press, 1971), Chapter 10.

66

Winston notes that, since imported raw materials and competing imports are highly correlated, the problem of multicollinearity arises. However, other studies support the hypothesis regarding the negative correlation between imported raw materials and capacity utilization. See, for instance, Pakistan Central Statistical Office, Economic Affairs Division, Report of Survey on Capacity Utilization by Manufacturing Industries 1965, Annexure B, and Warren Hogan, “Capacity Creation and Utilisation in Pakistan Manufacturing Industry,” Australian Economic Papers, Vol. 7 (June 1968), pp. 36–38, who also lists the inefficiencies in policies and planning, both of government and private enterprise, as well as some forms of technological or demand determinism often associated with capacity underutilization.

67

Winston, “Capital Utilisation in Economic Development,” (cited in footnote 1), p. 43.

68

Ibid., pp. 43–49.

69

For an analysis of this phenomenon, see Winston, “Overinvoicing, Underutilization, and Distorted Industrial Growth” (cited in footnote 2). Earlier, the practice was signaled in Budget Speech, 1965–66, p. 24. It was facilitated by an overvalued exchange rate that made imports artificially cheap, import duty exemptions, and a liberal system of depreciation allowances—up to 35 per cent on plant and machinery in the first year of installation, which further reduced the cost of capital goods. See Central Board of Revenue, Tax in Pakistan: A Brief Outline With Particular Reference to Tax Concessions to New Industries and Incentives to Foreign Investment (rev. ed., Karachi, December 1965).

70

If the industry reported higher operating rates, actual operations were taken as the standard. See Winston, “Capital Utilisation in Economic Development” (cited in footnote 1), pp. 41–42.

71

For these and following developments, see Ministry of Finance, Pakistan Economic Survey (Islamabad), various issues.

72

The rate of 42.52 in Table 4, representing the whole nonmetallic minerals industry, understates the degree of capacity utilization in the cement sector itself; actually cement had the highest utilization rate of the industry. See Report of Survey on Capacity Utilization by Manufacturing Industries (cited in footnote 66).

73

See, on earlier years, A. H. M. Nuruddin Chowdhury, “The Predictability and the Flexibility of Tax Revenues in Pakistan,” Pakistan Development Review, Vol. 2 (Summer 1962), pp. 189–214.

74

No changes in rates were made before the introduction of the capacity tax. The figures have some inevitable shortcomings. Thus, actual output data could not be corrected for changes in stocks of excisable commodities. Tax collections for some years show lags or leads, if only on account of the litigation procedures which were initiated. In addition, the fact that many production figures are collated by the Central Board of Revenue may have some effect on their reliability because firms that evade tax would also be likely to suppress production figures.

75

It will be noted that this ratio differs from the elasticity concept used in the first paragraph of this section, as it relates to production volume rather than value and eliminates the effect on collections of rate increases.

76

S.R.O. No. 73(R)/69, Gazette of Pakistan, May 6, 1969.

77

Islam, The Tax System in Pakistan (cited in footnote 40), p. 40.

78

J. Russell Andrus and Azizali F. Mohammed, The Economy of Pakistan (Stanford University Press, 1958), p. 342.

79

S.R.O. No. 127(1)/70, Gazette of Pakistan, June 29, 1970, p. 847.

80

The phenomenon that capital goods may have a substantial cost advantage relative to labor is not restricted to Pakistan. An overvalued exchange rate that cheapens the cost of imported capital goods is cited as one of the main reasons, but other contributory factors are: (1) comparatively low interest rates for large foreign producers or government-owned enterprises, (2) the transplanting of capital-intensive techniques from abroad, and (3) as regards labor, the prevalence of artificially high wage levels on account of minimum wage laws, union demands, and the employment of well-paid foreign staff against which local wages are measured. See Bird, Taxation and Development (cited in footnote 5), pp. 124–25, and John F. Due, Indirect Taxation in Developing Economies (The Johns Hopkins Press, 1970), pp. 142–43.

81

For a review of these measures, see IMF Survey, November 6, 1972, pp. 103–104.

82

Amotz Morag, “Some Economic Aspects of Two Administrative Methods of Estimating Taxable Income,” National Tax Journal, Vol. 10 (June 1957), p. 180, quoted by Richard M. Bird in Taxing Agricultural Land in Developing Countries (Harvard University Press, forthcoming), Chapter 7.

83

In this respect, the capacity tax can be compared with the forfait system of assessment for the sales and business income taxes in French-speaking West Africa and other countries. See Due, Indirect Taxation in Developing Economies (cited in footnote 80), p. 169.

84

The classic reference is to Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations, edited by Edwin Cannan (London, 1950), Vol. II, pp. 310–11. In a more recent study, Harley H. Hinrichs considers “certain” tax administration preferable to appealing tax inducements of doubtful merit. See “Certainty as Criterion: Taxation of Foreign Investment in Afghanistan,” National Tax Journal, Vol. 15 (June 1962), p. 153.

85

See G. Schmolders, Der verlorene Untertan (Düsseldorf: Econ. Verlag), pp. 181–82, quoted by J.C.L. Huiskamp, “Belastingontduiking” in Economisch Statistische Berichten, November 1, 1972, p. 1048.

86

See Bird, “A Tax Incentive for Sales: The Canadian Experience” (cited in footnote 10).