Mining taxation in Fiji The case of Emperor gold mines
International Journal of Social Economics; Bradford; 1999; Roman Grynberg; Peter Fulcher; Peter Dryden;

Volume: 

26

Issue: 

1/2/3

Start Page: 

79

ISSN: 

03068293

Full Text:

Copyright MCB UP Limited (MCB) 1999

Roman Grynberg: Department of Economics and the Department of Accounting and Financial Management, University of the South Pacific, Suva, Fiji

Peter Fulcher: Department of Economics and the Department of Accounting and Financial Management, University of the South Pacific, Suva, Fiji

Peter Dryden: Department of Economics and the Department of Accounting and Financial Management, University of the South Pacific, Suva, Fiji

Introduction

This paper considers the development of the mining tax regime in Fiji in light of the development of the country's largest private sector employer, Emperor Mines Limited (EML)[1] and its principal Fiji subsidiary Emperor Gold Mining Company Limited (EGM). The paper argues that EGM and EML have been the recipients of extremely generous tax concessions and subsidies throughout the last 40 years of their existence and that the company has, since the signing of the 1983 Vatukoula tax agreement (VTA) paid negligible amounts of company tax and minimal royalties to the Government of Fiji[2]. The tax agreement is so structured that the company is unlikely to pay any corporate tax or royalties to the Government until after the year 2011[3]. By the normal standards of international tax agreements in the mining industry the VTA is exceptional and stands as virtually unique among that of small low and middle income developing countries in allowing tax free status for a gold mine.

The paper develops through several stages. In the first, we consider the historical relationship between the Government of Fiji and EGM as it has emerged through various concessions granted under the taxation system as well as through the system of royalties and subsidies. The post-war colonial and post-independence taxation regime that has emerged at EGM has meant that the company has become dependent on state subsidy for close to half a century. The second section of the paper considers the corporate structure of EML, of which the three companies (Emperor, Koula and Jubilee) at Vatukoula are the main revenue centre. In the third part of the study we consider the terms of the Vatukoula tax agreement that was negotiated in the early 1980s and the circumstances that gave rise to the arrangement. In the fourth stage there will be some consideration of whether known economic variables such as gold prices, head grades and labour productivity can be used to explain the observed profits of the company. On the basis of available data there seems to be little evidence that a normal relationship exists between the profits of EGM and standard economic variables. However, the EML group profitability as a whole does exhibit a close relationship with standard economic variables.

In the final sections we consider the past experience with EGM and consider what implications the VTA may have on the future relations that the Government of Fiji will have with other mining companies. In particular the Placer Pacific is intending to develop a very large but low grade deposit of copper in the Namosi Province on the main island of Viti Levu. The fact that the Government of Fiji has permitted what are exceptional terms and conditions for mining companies in the case of EGM would suggest that this situation may be replicated in the future. The case of the Namosi deposit will be discussed in the conclusion.

While in comparison to many of the concessions that have been granted to industry in Fiji the tax concessions that EGM have received from the government may appear unexceptional, they are very substantial when viewed over the longer historical sweep of a period of 40 years. They are even more exceptional when they are considered in light of the tax agreements that are normally negotiated between mining companies and the state in other developing countries.

A great deal has been written both internationally and domestically about the Vatukoula gold mine and Emperor. The recent work by Bain (1994) has highlighted the history of relations between the company and labour. However, the difficult industrial relations situation at Vatukoula, where there have been a number of violent confrontations between Emperor and its labour force in the post-war era, has an uncomfortable parallel in the relationship between government revenue and the profits of the Emperor group. The Government of Fiji has been preoccupied with the maintenance of employment levels at Vatukoula and while the Emperor mine has been an important source of employment for the last 50 years, the Government allowed the continuation of agreements that have permitted EGM and EML to earn profits with little direct contribution to the revenue. The one exception has been the gold export tax which was reintroduced in December 1987. This relationship between Emperor and the Government of Fiji also reflects very much the situation that prevailed during the colonial era with the Colonial Sugar Refineries, which was the most important firm in the country. With the acquisition of CSR in 1973 Emperor has taken its place as the most influential of the private sector gaining substantial tax and other concessions from the state whenever there appeared to be a serious likelihood of redundancy and unemployment.

Section I: the history of concessions to EGM

The Government of Fiji has, throughout the entire period under study, maintained a taxation regime in the mining industry that has been heavily reliant on ad valorem imposts. The system has been based on three different components that have evolved over time. First is a system of royalties that can be as much as 5 per cent of fob value. Second, there has also developed an export tax regime which is based on varying rates of export tax but which currently stands at an effective rate of 3 per cent. Finally, there has also been the usual company tax provisions under Fiji's Income Tax Act that have developed over time but that have been eroded by tax concession as well as the tax efficient practices of the company.

The taxation provisions have changed dramatically in Fiji since the time that EML was first incorporated in 1935 and began operations[4]. Almost at the same time, EGM was incorporated in Fiji[5]. The tax regime that existed in Fiji during the colonial era was viewed by mine developers as not being conducive to the development of new mines.

For the Government of Fiji, in the case of the Vatukoula goldmine, neither the end product of the gold, nor the revenue brought to the state coffers appears ever to have been as significant as the employment offered by its extraction. The concessions that have been offered by the State since 1952 to EGM to solve problems of what the company claimed were inadequate finance, rising production costs, exploration and development costs and a static gold price, has been, at least publicly, to keep, on average, 1,500 workers employed and the town of Tavua operating. The State has granted concessions to Emperor since 1952 when Port and Service Tax was waived from bullion exports. Emperor, prior to this, had publicly expressed its dissatisfaction with the State's taxation policies. The chairman of the board of directors of EGM Company, of Lomaloma Goldmining and the secretary of the Associated Mining Companies, expressed their dissatisfaction with a royalty charge "based on profit on a sliding scale as against a fixed rate of 5 per cent ad valorem"[6]. The State was to collect approximately Pounds 300,000 in taxation (Port and Service tax, royalty and income tax) from the company for the year. These taxes, not uncharacteristically, were viewed by the directors as excessive, and the example of gold mining in Australia was pointed to where no income tax on mining profits, no royalty tax and no tax on dividends from gold mining were paid[7].

In order to provide financial support to the EGM the company received tax concessions and subsidies from the state throughout the 1950s. After 1952, income tax and royalty was waived for two years from 20 June 1957 (Taylor Report, 1960). On 19 June 1958, the State granted to the company a subsidy of Pounds 2 per fine ounce for three years - 1958/1959, 1959/1960, 1960/1961, with a maximum of Pounds 150,000 per year, a total of Pounds 450,000 for three years (Taylor Report, 1960).

The terms of the subsidy were, inter alia:

the company shall carry out capital development costing an equal or greater amount on mine buildings, plants and underground workings (exclusive of staff or employees' housing) during the period the subsidy; the company will be required to refund to the Government all sums received by way of subsidy in excess of the cost of such capital works; and

in the event of production in any one year falling below 75,000 ounces, the subsidy shall be reduced below Pounds 150,000 at the rate of Pounds 2 per ounce of the shortfall.

The relationship of corporate welfare recipient and donor that evolved between EGM and the State in the 1950s has continued unabated in different forms to the present. In order to stabilise employment in the Tavua area the Government has been willing to provide ongoing subsidy to the mine. The State, whether colonial or post-independent, always provided a regime that was deemed to be necessary to generate and sustain levels of employment. What appears to have been explored with only the most moderate interest by the government was whether the mine would have been viable without ongoing subsidy.

Despite the financial assistance offered in the early 1950s, operating the mine remained uneconomical and the company requested further assistance from the State in the form of a refund of Port and Service Tax from June 1958-1961, of Pounds 75,000[8]. The company also asked for the gold subsidy to be increased from Pounds 2 per ounce to Pounds 3 per ounce from June 1961, for a further three years to apply to up to but not more than 75,000 ounces per annum, i.e. Pounds 225,000 p.a. at a total cost of Pounds 675,000. The company finally also requested an exemption from income tax and royalty for five years, for the period 1962/1966 inclusive[9]. This was to be the first of many such requests for tax concessions. Subsequent attempts at tax exemption were to prove more successful.

The granting of subsidies and requests for further financial assistance did not end in the 1950s. Another agreement was signed between EGM Company and the State where EGM was given $2.040 million over a three-year period beginning 1 April 1967 for development and capital expenditure (Siwatibau Report, 1977, p. 11). Some of the conditions for this assistance were that tunnelling and rising would proceed from 10,000ft annually to 30,000ft, the maintenance of an annual ore production of more than 285,000 tons and the State's appointment of an observer to the EGM board. This $2.04 million could be regarded as a goodwill measure to EGM by the State to keep the mine operating, as Emperor did not repay the loan and the amount was "progressively written off" (Fiji Government, 1977, p. 30).

Again in May 1971, following independence, a further agreement was signed between the State and the company (Siwatibau Report, 1977, p. 12) where, inter alia:

a government grant of $150,000 was given for the extension of Borthwick shaft from level 11 to level 16 to be completed by 30 June 1974;

$450,000 was given to the Company for liquidity purposes in an interest free loan;

an industrial estate 80 per cent owned and funded by the State and 20 per cent owned and funded by EGM Co. Ltd was established in Vatukoula.

Following the US monetary crisis of the early 1970s the price of gold rose to almost US$200 per ounce in 1974. The government, therefore, briefly reintroduced an export tax on gold at 2 per cent of its value (Siwatibau Report, 1977, p. 12). But this was short-lived as government granted more concessions to Emperor after a drastic fall in the price of gold in 1976. According to EGM's directors, the company was once again experiencing extreme liquidity difficulties. The concessions included a waiver of 1975 tax due amounting to $105,000, a waiver of the balance of the 1971 loan which amounted to $206,326 and an advance of $300,000 without interest over a three-year period (Siwatibau Report, 1977, p. 13).

Yet another concession was granted in 1977[10]. In the Parliamentary Session of 28 November 1977, a motion was moved by the Hon. C.A. Stinson, Minister for Finance in the ruling Alliance Government, for advance of a further loan of $2 million to Emperor Gold Mining Co. Ltd. This had been a recommendation of the Committee appointed by the Minister of Labour to look into the closure of the goldmine at Vatukoula.

Stinson, in his speech, outlined the assistance given by the State to the Emperor goldmine since 1968. He justified the extensive subsidies that had been offered to the company on the grounds that the "economic and social contributions" EGM was making to the Fiji economy and people more than compensated for the assistance, and justified it. Stinson stated that between 1968 and 1976 inclusive, wages and salaries paid by Emperor were almost $24 million, PAYE tax received totalled $2.156 million, FNPF $2.008 million and company tax, gold tax and royalties totalled $2.2 million (Fiji Government, 1977, p. 30)[11].

Jai Ram Reddy, the leader of the opposition, in his reply to Stinson argued that even though the company had stated that it had "liquidity problems" in 1974/1975, they paid dividends of $817,000 to their parent company in Australia and had paid out $1.571 million to subsidiary companies in Fiji (Fiji Government, 1977, p. 38). These subsidiaries, Reddy also pointed out, had made profits in 1974, 1975 and 1976 of $609,000; $475,000; and $526,000 respectively. EGM Co. Ltd was, therefore, making a profit through its subsidiaries[12]. The Siwatibau Report also pointed out that certain inter-company transactions were incorrectly charged to the Emperor Goldmine account instead of the subsidiary accounts, though found no strong evidence of transfer price manipulation.

The motion put before Parliament and approved was that EGM Co. Ltd be given the following concessions:

an advance of $2 million in total repaid at the rate of 50 per cent of pre-tax profit of EGM Ltd, repayment to Government to be exempted from tax, 50 per cent of the advance to be written off and the advance subject to interest at the rate of 3 per cent per annum;

workers were given a 5 per cent wage cut.

In January 1978, soon after these concessions were granted, EGM announced it was going to lay off 1,000 mineworkers as a cost reducing measure[13]. On 19 January 1978, the Fiji Government announced its intention to expropriate the company. This was announced personally by the then Prime Minister, Ratu Sir Kamisese Mara, who travelled especially to Vatukoula to make the announcement to striking mine workers.

But in May, negotiation between the Government and the company failed after disagreement about the price of the company's assets[14]. Government offered $6 million for Emperor's assets and operations while the company asked for $37.5 million. The Minister for Economic Planning Sir Vijay Singh, who had been given the task of negotiating Government's take-over of the mine, stated in Parliament that "there was no way in which Government could justify an expenditure of anything like $37.5 million"[15]. The negotiations over nationalisation ended without agreement and the company continued to be privately owned.

In 1982, the government of Fiji again offered financial help to Emperor after a Cabinet sub-committee study found that EGM Ltd was, in a manner consistent with its history, experiencing serious financial difficulties. Government had decided to help after a cabinet sub-committee had studied "EGM's troubles"[16]. As before, one of the reasons given by the government to justify government aid was that mine workers would continue to be employed. Emperor Goldmining Company along with a new and much larger partner Western Mining (Fiji) Corporation (WMFC) were granted 21 year goldmining leases by the Government. One of the conditions of the lease was that workers would not be laid off without prior consultation with Government. The consortium was also given a special prospecting licence covering 6,123 hectares "with an agreed three year $3.13 million exploration program in the Tavua Basin".

The historical evidence begs an important economic question. Were the concessions granted to the company actually necessary or was the Government of Fiji a poor negotiator or the unwitting victim of a company that was able to continually threaten unemployment? There exists no definitive answer to this question, and it is not the purpose of this paper to seek one, as it would necessitate knowledge of the actual cost of operations over a long period of time. No doubt such data were given to both Taylor and Siwatibau Commissions that have been referred to above. However, given the free rider problem of which economists are well aware it is simply not knowable whether these data were accurate or not. Clearly, Siwatibau was of the view that the subsidies were justified[17]. Nonetheless, a number of commentators over the years have viewed with great scepticism the decision to subsidise the company. In particular, Ward (1971) in one of his less sagacious comments argued:"

...it is pertinent to ask whether the payment of subsidy worth $750 per worker plus the probable continued subsidisation of the industry is really prudent or justified when the long run prospects of the gold mining industry in Fiji (and in the world as a whole) are relatively dismal."

It would be easy but quite unfair to criticise Ward with the benefit of hindsight given the floating of gold prices in 1973 and the continuation of production at Vatukoula for at least another 30 years. However, given what was in place at the time, i.e. the Breton Woods system with a fixed gold price of US$35 per ounce and a rising wage rate, a low ore grade mine, one could well understand the subsidy decisions prior to 1973. Moreover, there appears to be evidence to suggest that in the 1960s and early 1970s Emperor was indeed caught in the classic position of being a marginal mine facing a fixed price of gold and rising costs of production (Atu Emberson-Bain, 1994, pp. 161, 238, 247).

All this is not to suggest that the data necessarily vindicate the decision of government. The foregoing was intended to demonstrate only that the Government of Fiji had maintained a regime of subsidy for 30 years prior to the introduction of the Vatukoula tax agreement. However, the taxation regimes that were set in place, which is the main focus of this paper, were in place in 1982, well after the end of the fixed gold price system and are hence far more questionable. Some commentators have generally supported the system of subsidy and tax exemptions as being necessary to transform EGM. Kasper (1988, p. 117) has argued that the joint venture agreement transformed Vatukoula:"

... from a rather conventionally run mining operation that experienced difficulties in continuing operations into an efficient and profitable enterprise."

It is not the purpose of this paper to establish the justification or otherwise of the decisions of the Government of Fiji to offer subsidies and a taxation regime that is exceptionally generous. The purpose is to document the concessions, demonstrate how exceptional they are and to show the consequences in terms of corporate structure and operations.

What is particularly significant is the timing and nature of the concessions, subsidies and soft loans that have been offered to Emperor over the years. The concessions, as can be seen in Figure 1, have almost invariably followed periods of decline in output and employment. But what is more significant is that the concessions have been followed by substantial increases in gold output. This very much falls in line with the arguments that the subsidies have been the basis of the continuation of the development of new shafts and mines. What is evident is that Vatukoula and the Tavua Basin have substantial deposits of gold that will continue to sustain production well into the twenty-first century. At a recent business meeting in Fiji senior Emperor officials said they intend to develop two new mines by the year 2000 with output of 200,000oz per annum at a cost of US$300 per oz[18]. The Government of Fiji subsidies have assisted the company in subsidising its expansion.

However, all this is not to argue that the concessions to Emperor were in effect a conditio sine qua non, for the continued existence of the Vatukoula operations and that production may well have continued without them. Certainly from the magnitude of the reserves, the very high grades of the reserves and the apparent willingness to continue to invest, it is by no means obvious that production would have ceased had the Government failed to continue the subsidies. At the very minimum, what is needed is some analysis of profitability, but even this is inadequate for those who wish to know whether the company would have invested without the subsidies.

Figure 2 considers the average operating during the 1950-1970 period in comparison to the price of gold on the London market in Fiji dollars. The data are derived from published cost figures[19]. The figures indicate that there certainly was a decrease in profitability in the late 1950s which accompanies the early subsidies of the colonial government and that were discussed in the Taylor Report. No cost data were available after 1970 and hence the arguments for subsidy have less supporting evidence.

None of the data presented answer the really significant counter-factual question of whether Emperor and its related companies would have undertaken the investment in the expansion of mining capacity in the 1960s and 1980s without the concessions or with much less generous concessions. To this answer this question with resort to ex-post data is simply not possible.

In the final analysis no judgment can or perhaps should be made of colonial or post-independence government policy towards Emperor. It is certainly not the intention of this paper to either justify or criticise the Government of Fiji's decisions over a period of 30 years. Certainly the published data, assuming they are accurate, support the arguments that profits declined in the 1950s and when new investments were undertaken and new mines developed production and profitability rose again. Whether the company judiciously managed its relations with the colonial government of Fiji to obtain subsidies to fund its expansion or whether it was genuinely on the verge of bankruptcy and the loans were absolutely essential remains largely unanswered and unanswerable. What is certain is that the Vatukoula Tax Agreement, as the most recent chapter of relations between the Government and the corporation, was a generous and expensive subsidy, probably far greater than was necessary to induce an investment in a such high grade deposits.

Section II: The structure and finances of EML-1992

Corporate structure

EML is a listed company in Australia. Apart from the subsidiaries in Fiji, it has a number of subsidiaries incorporated variously in Fiji, New Zealand, Hong Kong, USA and Vanuatu. A number of these companies are dormant. The corporate structure, as of 1992, is diagrammatically depicted in Figure 3.

Financial analysis

The financial performance of the group has fluctuated widely over the period 1980-1992. In general, while EML has made profits, the main operating subsidiary, EGM, has recorded substantial losses and has effectively paid no income tax. From the limited information available, it appears that EGM has borrowed heavily from its parent company (inter-company payables to EML totalled $47 million in 1992) and that interest payable on these advances has run at approximately $4 million per annum in recent years.

During the period under review, EGM appears to have paid little or no taxes. This is during a period when gold production averaged 80,000 ounces per year. At prices approaching US$400 per ounce, the revenue earned over the period would be in the order of US$400 million.

Emperor Mines Ltd

Since 1986 EML has been domiciled in the tax haven of the Isle of Man, although its shares are listed in the Australian Stock Exchange. A summary of the indicators of the company are presented below in Table I.

Excluding the effect of associate company losses and abnormal and extraordinary items, EML is a consistently highly profitable company which appears to pay no tax. Whether this is due in part or in whole to its domicile in the Isle of Man is unknown.

Associate company losses refer to investment in Odin Mining. This investment was sold at a loss of $A4.1 million in 1993. Extraordinary and abnormal items relate in part to operations outside the Fiji operation and thus their exclusion leads to the conclusion that EML's Fiji operations are extremely profitable. An analysis of the profitability of the Fiji operation appears in Table II.

It therefore appears that EML has reported profits of approximately $50 million for the period 1986-1992 from its Fiji goldmining operations. By way of contrast, the following analysis shows reported profits by the major Fiji operating companies and attempts to reconcile them to the parent company figures (Table III).

What is relatively clear from the above is that the operations in Fiji were profitable. This can be demonstrated in two ways. If one relates profits to capital investment over the period 1986-1992 Emperor's return on assets averaged 12 per cent. This is most likely to be an underestimate as assets of EML include assets outside Fiji. The second way is to compare return on assets with other mining operations. In the 1980s prior to its shutdown due to insurrection, Bougainville Copper Limited averaged return on assets of 25 per cent. However, BCL was considerably larger than Emperor. One Australian firm which had revenues similar to that of Emperor was Central Norseman Gold Corporation Ltd, which averaged 8.5 per cent return on assets over the period 1986-1990. However, the profitability of EML is not the issue - what remains of concern, in light of its apparent profitability, is the absence of direct company tax payments.

From the above analysis, incomplete thought it is, two questions arise. First, why is the parent company reporting much higher profit over the period than the operating subsidiaries? Part of the answer may be due to inter-company indebtedness and interest thereon. The second question is: given that the operating subsidiaries of EML have made substantial profits, why have they not paid tax in Fiji since the early 1980s? The Vatukoula Tax Agreement provides a reasonable explanation.

Section III: The Vatukoula tax agreement

In 1983 EGM and WMFC established joint ventures[20] to conduct gold operations at Vatukoula. The Vatukoula join venture (interests apportioned initially 90/10 and subsequently 80/20 between EGM and WMFC) provided for the working and development of the existing Emperor mine. The Tavua Basin joint venture (equal interests to EGM and WMFC) provided for exploration over adjacent areas of the Tavua Basin. WMFC acted as manager under each joint venture. At the establishment of the ventures, WMFC acquired interests in EGM's mining and prospecting assets in proportion to the company's interests under the ventures[21].

Exploration activity established the feasibility of further mining operations in a portion of the Tavua Basin joint venture tenements adjacent to the existing Emperor mine. In 1985, a third joint venture, the Tavua Basin Mining joint venture, was established to develop and work the Philips shaft. As with the existing Tavua venture EGM and WMFC held equal interests. In 1987 EGM conveyed its interest in the new venture to Jubilee Mining Limited (JML), a then dormant subsidiary of EML.

In 1991 EML acquired WMFC from its Australian parent. The acquisition was in part funded by the issue of fully paid-up shares in EML to the Australian parent which thus remained indirectly interested in the Vatukoula operations (EML, 1991). Tailings from the Emperor mine were excluded from the Vatukoula joint venture and related asset sale. During the 1980s EGM attempted to interest a number of foreign parties in working these tailings[22]. Eventually a 50/50 joint venture was established in 1991 between Vatukoula Tailings Co. Ltd (a wholly owned subsidiary of EML) and Ranger Fiji Ltd to rework tailings from the Emperor mine. The venture was unsuccessful and tailings retreatment ceased at 1992 year end. EGM subsequently acquired all assets of the two joint venturers and mothballed the retreatment facilities.

Exploration and mining activities under the Vatukoula joint venture (VJV), the Tavua Basin joint venture (TBJV) and the Tavua Basin Mining joint venture (TBMJV) benefit from a three-party agreement between the two original joint venturers (and their assigns) and the Fiji government concerning Fiji tax liabilities[23]. The VTA operates from March 1983 through to March 2004, and in the case of any new mine, to March 2011. Concessions granted by the agreement concern income tax, stamp duty, royalties and customs and fiscal duty[24].

The VTA is not supported by any special implementing legislation. It purports to rely on provisions in existing tax legislation authorising the Minister of Finance to grant specific tax concessions[25]. Three general topics are covered in the VTA: tax consequences of establishing the joint ventures; operations under the joint venture; and withholding taxes on payments to non-residents.

Establishing the joint ventures

On establishing the VJV and TBJV, WMFC acquired existing assets of EGM in proportion to its interests under the two joint ventures. Fiji tax rules would subject the asset sale to stamp duty[26], require EGM to recognise for income tax purposes any recouped depreciation[27], and forbid WMFC deducting for income tax purposes that portion of the sale price concerning capital assets other than mining tenements[28]. The VTA annuls or even reverses each of these tax consequences of the asset sale. Stamp duties are waived[29], EGM is relieved from recognising any of the sale moneys for income tax purposes[30] and WMFC is granted an immediate write-off for income tax purposes of its acquisition expenses[31].

Operations

Under Fiji law, mining operations are subject to income tax, royalties, customs and fiscal duty on imported plant chemicals and explosives etc. and export tax. The VTA provides for concessions concerning all these taxes except export tax.

The operations of the joint venture benefit from two independent income tax concessions. The first and more significant is the grant of a seven year tax holiday for any new mine[32]. New mine is defined as any mine established on the TBJV tenements[33]. The Philips shaft established under the TBMJV with ore grades considerably richer than the very old Emperor mine operated under the VJV benefits from this exemption.

A second concession concerns the write-off of exploration expenditure and capital expended in mining operations. Regular tax rules provide for such expenses to be written off by five equal deductions spread at the election of the taxpayer over an eight year period[34]. The VTA permits these expenses to be wholly written off in the year incurred[35]. Alternatively, deduction may be postponed for up to eight years, thereby preserving the benefit of deduction until after any tax holiday.

Fiji law reserves precious metals to the state[36]. Royalties are the price paid to the state for exploitation of minerals. The Mining Act requires the miner to pay the royalty "prescribed"[37]. Regulations to the Act provide for a royalty on gold of 5 per cent of the value of ore extracted. Alternatively, royalties may be calculated, subject to the Minister's consent, on a yield basis[38]. In essence the yield formula provides for a royalty equal to 10 per cent of profits. Where profits exceed 20 per cent in value of sales the royalty rate increases by increments. The royalty formula in the Mining Regulations may be supplanted by royalty terms in a mining lease. The new joint mining leases issued to EGM and WMFC at the commencement of their joint ventures in 1983 set royalty at 2.5 per cent of taxable income. An additional royalty due under the VTA (labelled special royalty) ensures the total royalty for a year equals 2.5 per cent of taxable income before the deduction of any loss carried forward from earlier years. As is usual, royalty payments are deductible in calculating taxable income for income tax purposes.

Fiji's customs legislation provides for the imposition of a customs duty, fiscal duty and export duty[39]. A general goods and services tax introduced in 1992 now replaces fiscal duty[40]. The VTA exempts from fiscal duty all plant and equipment imported exclusively for use in mining operations[41]. Additionally, specified chemicals are exempt from customs duty. Customs duty on all other items is capped at 5 per cent[42].

No concessionary treatment is granted for export duty on gold. At the commencement of the VTA in 1983 export duty on gold was 0 per cent. Various rates of duty have applied in later years, ranging from 2 per cent to 5 per cent.

Withholding taxes

Any consideration of the joint venturers' Fiji tax liabilities must take account of the fact that these companies are operating subsidiaries of foreign parents and take account of the tax position of the whole group. Concessions granted under the VTA regarding withholding taxes are not tax neutral and go a long way towards explaining why the parent EML derives much of its income by way of interest and know-how payments.

Fiji's Income Tax Act imposes withholding taxes on, inter alia, dividends, interest and know-how payments made to non-residents[43]. In the period under consideration, withholding tax on interest payments was 10 per cent until 1984 and thereafter 15 per cent[44]. The tax on dividend and know-how payments was 15 per cent until 1990 and 30 per cent thereafter[45]. The withholding tax on know-how payments applies only to the profit component of a payment. Rates stipulated by the Income Tax Act may be reduced by Double Taxation Treaties. For example, Fiji's 1990 treaty with Australia limits withholding taxes on interest to 10 per cent and on dividends to 20 per cent. The VTA excludes withholding taxes on interest[46] and know-how payments[47] but leaves in place the withholding tax on dividends.

The concessions on withholding taxes are the most interesting component of the VTA. They prompt several questions. Why should the VTA deal with withholding taxes at all? Related to the first question, who is the intended beneficiary of the concessions? Finally, is the intended beneficiary the actual beneficiary or has a distortion arisen?

From a purely legal perspective the withholding tax arrangements have no place in an agreement with Fiji joint venturers. Conceptually, withholding taxes are taxes levied on the recipient of income and merely collected by the party making the income payment. The substantive tax obligation waived is that of a third party who is not party to the agreement. The situation is different from a commercial viewpoint. Commonly, the benefit of the waiver of a withholding tax is passed along to the party making the income payment. Abolition of withholding tax on interest income, for example, is passed on to the borrower as cheaper money. The commercial practice justifies the inclusion in the VTA of a concession which from the legal perspective concerns only a third party.

A waiver of withholding tax may alternatively benefit the income recipient or party making the income payment. The obvious question arises: which party was the Government of Fiji intending to benefit with these particular withholding tax concessions? Here the pattern of concessions is revealing. Fiji's Income Tax Act has three withholding taxes relevant to the Emperor Group. Only two are waived. The two concern interest and know-how payments that represent expenses to the Fiji joint venturer. The withholding tax retained concerns payments - dividends - on which no "commercial" benefit of waiver can be passed along to the party making the payment. The pattern of concession suggests the agreement's objective is to waive withholding taxes for the benefit of the joint venturer's themselves rather than any third party.

Whether the intended beneficiary of the concession is the actual beneficiary will depend on the commercial practice of passing along the benefit. The practice presumes a competitive market with parties, for example borrower and lender, dealing at arm's length. Where borrower and lender are related there is no assurance benefits of waivers will be passed along.

The absence of arm's length dealing does not of itself necessarily mean benefits will not be passed on. However, what is ultimately perverse about the withholding tax concessions is that given the structure of the EML group the overall pattern of concession in the VTA almost guarantees such an outcome. The parent, EML, as both lender, supplier of know-how and owner may extract money from the joint venturers as interest income, know-how payments or dividend income. Retention of the withholding tax on dividends pushes the parent to ensure sums as large as plausibly possible are extracted as interest or know-how payments. Under such circumstances no benefits will be passed on to the borrower or know-how user, the Fiji joint venturer, and thus the intended beneficiary will not benefit. The actual beneficiary becomes the parent company, EML.

The tendency of high interest and know-how payments charged by the parent to the joint venturers has a further consequence. Should the waiver produce the usual commercial outcome, cheaper loans and know-how, tax revenue foregone from the loan or know-how supplier would be balanced by increased tax revenues from the increased profitability of the joint venturers. However, the pattern of concession found in the VTA encouraging more expensive loans and know-how ensures lower profitability of the joint venturers. Rather than there being any balancing, further tax revenues are foregone.

If, as suggested by the pattern of concession under the VTA, concessions on withholding taxes are meant to benefit the joint venturers and not third parties, the concessions should be limited to loans and know-how from independent third parties. Alternatively, if loan capital and know-how is to come from the parent the pattern of concession should be to waive the withholding tax on dividends alone. Loans and know-how will then be supplied at market or better rates. The balance that exists under the Income Tax Act in a wholly domestic situation between debt and equity funding will be restored. Tax forgone on dividend payments to the parent will be balanced by increased taxable profits of the joint venturer.

The comparative tax position

How ordinary or extraordinary are the tax concession found in the VTA outlined above? From a purely formal perspective the concessions granted are not exceptional. All, with one minor exception, have a statutory basis in existing Fiji legislation. The joint venturers have only received tax advantages already envisaged and provided for in Fiji law.

Equally, from a comparative Fiji perspective, the concessions are not exceptional. Fiji's present Income Tax Act provides for a broad range of concessions in a variety of industries. For example, manufacturers for export receive 13 years tax holidays, exemption from customs and fiscal duty and VAT and tax free repatriation of profits to foreign owners. Farmers may receive, and have received, tax holidays, most recently from 1986-1991. Parties investing in hotels and tourist watercraft benefit from accelerated depreciation and the waiver of interest withholding tax. Seen against this range of concessions the VTA appears unexceptional.

A quite different conclusion is reached if comparison is made with fiscal regimes imposed on mine operations in other jurisdictions. This is the more appropriate comparative perspective for at least two reasons. First, goldminers operate in a world market and not a Fiji market. Second, fiscal regimes for mines and other industries are not readily comparable. A tax holiday, for example, has quite different implications for a mine, where operating profits will typically peak after capital cost recovery in early years, and manufacturing, where there may be indefinite steady growth with a corresponding profit profile.

There is in economics a long history of analysis of what constitutes an appropriate taxation regime where the resource is non-renewable in nature and where economic rents are generated. This paper will not consider these theoretical arguments, however; these have been made in mainstream economics with sufficient regularity; so that taxation regimes based on resource rent taxation are now considered to be part of the standard taxation rules employed in such situations[48].

What is unique about the VTA is that among developing countries it is the only taxation regime that has effectively exempted the developer from any income tax for a period of some 27 years. It is possible to compare the VTA with either legislative or negotiated agreements. However, as the VTA so adequately demonstrates, it is pointless to use these statutory taxation regimes for a basis of comparison. These statutory comparisons of jurisdictions have been made in the past and their fundamental weakness is well understood (Grynberg and Powell, 1992, pp. 215-20). Table IV considers the taxation agreements in various countries that has been employed in a recent CFTC report on mining regimes commissioned by the government of PNG (Commonwealth Fund for Technical Co-operation, 1990, pp. 69-73). The regimes, except in one case, have no name attached as they are actual contract terms and are hence not normally available.

What is evident from Table IV is that VTA is the most generous of the actual contract terms that can be found among developing countries. The Fiji regime has no state participation element, let alone one where the cost of the state participation is borne by the developer as the cases above. There is, moreover, no additional profits tax[49] in the event of the firms' earning economic rents. The ad valorem taxes employed are almost invariably higher in all other agreements. Simply, the VTA has no international parallel except for the case of Australia, which has had no gold mining taxes up until 1990 when the exemption was finally repealed (Otto, 1995, p. 304). As a mining tax agreement the VTA is unprecedented, especially given the relatively high ore grades in some of the mines that were developed by the joint ventures (Grynberg and Powell, 1992).

Section IV: The implications of the Vatukoula tax agreement

This section considers some of the implications of the tax agreement on the operations of EGM and EML in Fiji and abroad. It is clear that the agreement effectively exempted the "new mines" owned by EML from taxation. In the years immediately following the agreement, two companies, Jubilee and Koula, both products of the joint venture agreement between Western Mining and EML, became active and both began to earn substantial profits as is evident from Table II in Section II.

However, at the same time, EGM, which was not operating any "new mines" as defined under the agreement and hence paid taxes, began to earn decreased profits. This observation is entirely consistent with a rational tax planning approach to its operations. EGM's earnings dropped and the company appears to have become a significant liability. In part the disparities between the performance of the companies can be easily explained by the much higher recovered gold rates of Jubilee and Koula[50]. However, there are a number of other factors which might offer a reasonable explanation of the apparent profitability during this period. The ore from the high grade Tavua Basin operations was treated at Vatukoula JV treatment plant[51]. In other words the main operating cost of the facility was shared with the low profit portion of the firm's Fiji operations. The method of cost apportionment is not known for this cost, nor is the method of apportionment employed for other overhead costs which would almost certainly be shared between the "old mines" and tax free "new mines".

The VTA has been a source of controversy within Fiji for a number of years. The Auditor-General, in the 1989 annual reports, had cited several aspects of the agreement that were, in his opinion, questionable, the first being that the agreement appears never to have been gazetted[52]. In 1989 the Auditor-General noted that VTA meant the government had lost $13 million in royalties and the report recommended that the agreement be reviewed[53].

The most obvious question that arises is whether in fact the VTA agreement has caused or been associated with a fundamental change in the way in which mining operations have been conducted and whether this has meant that the government of Fiji has been receiving tax and royalty returns from the mine that reflect the real financial situation or whether there have been economic distortions caused to the efficient operations resulting from the VTA. The hypothesis that is examined below is that if the firm operates in a normal manner, along the line of most gold mining companies, its profits will be positively associated with gold prices, recovery grades, output and productivity. An ordinary least squares analysis was undertaken employing variations of the equations below:

[Pi] = [alpha][sub]1X[sub]1 + [alpha][sub]2X[sub]2 + [alpha][sub]3X[sub]3 + [alpha][sub]4X[sub]4 + [alpha][sub]5X[sub]5 + [alpha][sub]6X[sub]6(1)

where

X[sub]1 = intercept

X[sub]2 = price of gold

X[sub]3 = output

X[sub]4 = productivity

X[sub]5 = recovery or head grade

X[sub]6 = proxy for industrial conflict (dummy variable)

Equation (1) is assumed to have the following relationships:

[delta][Pi]/[delta]X[sub]1 > 0, [delta][Pi]/[delta]X[sub]2 > 0, [delta][Pi]/[delta]X[sub]3 > 0, [delta][Pi]/[delta]X[sub]4 > 0, [delta][Pi]/[delta]X[sub]5 > 0, [delta][Pi]/[delta]X[sub]6 0

Data sources constitute a perpetual problem in any statistical analysis and proxies had to be employed in several cases. The measure of profits was operating profit of the individual firms in the EML Group[54]. The data for output and recovery grades came from reports of The Australian Stock Exchange to which EML reports[55]. However, this was only useful for the period from 1986-1992. Prior to this date data are less reliable and come from several sources which may not be entirely consistent. Gold prices came from the annual averages published internationally (IMF, 1994). The employment figures are those provided by the Mineral Resources Department for employment in the mining and quarrying sector of Fiji. No exact employment figures for EML or its subsidiaries were available but, as EML constitutes between 80-90 per cent of value added in the sector, total sector employment is an adequate though imperfect proxy[56]. Because of the way in which tax liabilities are presented in EML accounts, taxation was based on receipts of tax revenue by the government of Fiji, unless otherwise stated. Given the obvious importance of the long strike that occurred in 1991/1992 at Vatukoula, a proxy was also employed for the existence of industrial conflict.

The results of the OLS analysis are provided in Table V.

The number of degrees of freedom in equations 1-3 in Table V is very low. However, it is not the intention of the authors to draw conclusions regarding the accuracy of the independent variables in explaining profitability. The intention is simply to demonstrate that there is a substantial difference between the results regarding the significance of the independent variables, which are not biased by a low degree of freedom. The results indicate that the standard economic variables of price and industrial relations factors are significant explanatory variables of the profits of the EML group in the period under study (see equations 1 and 2)[57]. However, there is no evidence that the same variables explain the observed total operating profit of the three revenue centres (see equation 3 in Table V) in Fiji, namely EGM Jubilee and Koula[58]. This is a curious result and cannot easily be dismissed on the basis of poor data or a short time series. The three revenue centres are together responsible for 90 per cent of the profits of the group. It would appear to indicate that there is a divergence between the observed profitability of the revenue centres and that of the EML company. This could be explained by transfer price manipulation and/or tax efficient strategies that result directly from the VTA.

The EML holding company in 1992 earned revenues of AUD 6.4 million, of which $3.36 million came from interest charged to subsidiaries (EML, 1993, p. 27). The firm also earned AUD 1.2 million in dividends and $1.34 in "other income". While there is no clear indication of the source of this income it no doubt comprises management fees from its subsidiaries. In 1991 EML wrote off $2.3 million in unrecoverable management fees charged to subsidiaries (EML, 1993, p. 29). Thus it would appear that the main source of revenue of the holding company is the charging of interest and management fees to its subsidiaries. Both these are exempt from the 15 per cent withholding tax under the terms of the VTA and these activities of the holding company, whose revenues are 12 per cent of 1992 EML revenues, may well account for the reason why variables such as production and gold price appear to exhibit no significant relationship with the profits of the three main revenue centres of the company as shown in the results above.

What is also of more than passing interest is the failure of the model to predict profits of EGM prior to effective operation of the Western Mining joint venture[59]. Equations 4-6 in Table V indicate that none of the independent variables are significant in explaining EGM profits and, moreover, the regression equation itself is insignificant. However, even eliminating the weakest of the variables, namely productivity, does not tend to increase the predictive power of the model. While it remains possible to explain the failure of these variables to predict profitability on the data, an alternative explanation is, once again, that the recorded profits of EGM could over this period reflect optimal tax avoidance practices[60].

Tax revenue from the Vatukoula operations

There are three tax sources of revenue from the Vatukoula mine for the government of Fiji. These include:

- (1) Company tax.

- (2) Royalties.

- (3) Export tax.

The taxation implications of the VTA are of crucial interest in this study. In the period 1987-1992 the Emperor group of companies appear to have paid almost no corporate tax in Fiji. Given the way in which reports are presented to the Registrar of Companies it is impossible to determine conclusively whether there was any tax paid[61]. However, the EML report to the Australian Stock Exchange indicates that over the period 1986-1992 there were income tax expenses only in 1986, 1988 and 1989[62] but the company last made provisions for tax in 1987 when it paid $3,000[63] on a turnover of almost AU$72 million and an operating profit for that year of AU$22.4 million.

The royalties paid by the mining industry are presented in Table VI. It is interesting to note that not only are these minuscule amounts, they are often paid two years in arrears[64].

In large measure in response to the situation described above the Government of Fiji has imposed export taxes of varying rates on the exports of gold from Fiji (Table VII)[65]. The VTA does not prohibit export taxes and as a result this has become the preferred device for dealing with the obvious loss of revenue from the mining operations. Export taxes, while revenue efficient, are the least economically efficient taxation instrument in mining operations as they raise costs and shorten mine life.

Table VIII attempts to answer the question of what the forgone earnings are to the Government of Fiji from the VTA. The estimates are based on the assumption that Emperor should, if confronted with a reasonable tax regime, pay a similar tax liability to the one existing in PNG which by the standards of developing country mining terms is widely considered to be quite reasonable[66]. The two estimates provided are based on the assumption that EML should pay a rate of tax as a percentage of exports equivalent to the average paid to the PNG government, both national and provincial, in the period 1986-1991[67]. The two estimates are based on the ratio of the contributions by the mining companies to the PNG Mineral Resource Stabilisation Fund (MRSF) divided by exports. The lower estimate of 8.55 per cent is based on the period 1986-1991 and the latter on a shorter period 1986-1989. The reason for the second estimate is that the Bougainville Copper Mine was closed as a result of the insurrection on Bougainville; as a result contributions to the MRSF were negligible after 1990. In the period 1986-1989 the average government revenue as a percentage of exports in PNG was 11.69 per cent[68]. This contrasts very sharply with the average take of the Fiji Government of 1.87 per cent of total exports between 1986-1992. Had the returns been the same as in PNG, the very minimum that the Fiji government should have taken in revenue was $37.07 million and had the higher estimate been used the Government should have received $50.67 million. Thus the cost of the VTA is a subsidy of between $4 and F$6 million per annum to the company over the period. Assuming that the objective of the Government of Fiji is, as it has been throughout the last 50 years of policy to the mine - namely the maintenance of employment at Vatukoula, the cost of the agreement is high as the Government is paying a subsidy over the period under study of between 42 per cent and 62 per cent of the gross salary of every employees of Emperor[69].

Several attempts were made to explain the level of taxes derived from the mine from available data. Using estimated government tax payments (corporate tax and royalties) there seems to be an inverse relationship between tax revenue and the gold price, which means that the Fiji taxation system appears to have reversed the normal mining taxation practice, which is the capture of a greater portion of the economic rents as mineral prices rise. Moreover, as the OLS results below indicate, there is no significant relationship between either the corporate taxes or the total tax paid and the profits of EGM over the 1970-1985 period, which was prior to the effective operation of the VTA (Table IX). This would support the hypothesis that, even prior to the advent of the VTA, the Emperor group of companies' taxes paid did not vary in accordance with the situation normally associated with the mining industry.

Conclusion

Since the creation of EML in 1983 there seems to be little doubt that it has been profitable. It has operated as a tax free mine in Fiji under the terms of the VTA. The loss of revenue, as calculated above, is significant. However, the real effect of the VTA is not simply that it has provided a mining company with an exceptional tax agreement that allowed it to arrange its affairs to avoid taxes, not only for new mines, but for all its operations. What the Government of Fiji has established is a precedent that may well prove inescapable. The Government of Fiji is presently entertaining the possibility of granting a development permit to Placer Pacific to develop the huge but low grade Namosi copper deposit. This project is being negotiated at the very time when investment in Fiji is at a low ebb. Investment as a percentage of GDP has fallen to 15 per cent according to recent estimates, down from 24 per cent at the beginning of the 1980s.

The Government of Fiji wishes to promote investment so as to create employment. In many ways the situation that existed in the mid-1930s, when investment dried up because of the depression, is being replicated but for quite different reasons. The failure of private investment in Fiji is very much a result of factors particular to the Fiji economy, whether they be political or economic, rather than being a result of global recession. Since the 1930s, when Emperor began its operations in Vatukoula, it has throughout most of its life received tax rebates, subsidies, export bounties, interest free loans because the state has been willing to accept Emperor's arguments that if it failed to receive what it claimed were needed subsidies and concessions it would be forced to retrench large parts of its predominantly Fijian labour force.

The Namosi copper deposit is, in terms of employment creation and investment, the largest project in the history of Fiji. However, Placer Pacific claims that the deposits are very low grade ore and hence it will not be difficult for the developer to pursue precisely the same form of pressure on the Government of Fiji as EGM has done. There is every likelihood that the Government of Fiji will find it very difficult to escape the precedent created by the VTA. In fact Placer has applied precisely the same sort of pressure that was used by Emperor. In June of 1995 it ceased further exploration and development activities at Namosi pending a successful outcome of tax negotiations for the development of the mine. In light of the experience of EML and EGM, there is little reason for optimism that an appropriate tax regime will evolve[70].

Given the situation currently existing in Fiji, one is at least obliged to offer some form of alternative to the current situation in Fiji. The case of PNG, which has been a source of uncomfortable and invidious comparison to the mining tax situation which prevails in Fiji, is a basis from which to begin. The best aspect of the PNG system is its transparency and clarity, which contrast sharply with the situation of Fiji. A mine developer in PNG knows exactly the tax regime he will confront because the terms are published and available in legislation. It is not possible to negotiate concessions from the tax regime without an act of parliament. This puts considerable pressure on policy makers not to depart from the published terms. Thus the very first measure needed to improve the mining tax situation in Fiji is a clear and published draft tax agreement. While a change in legislation may provide a proper de jure framework, the de facto situation will not change until the discretionary power available to ministers and Cabinet in Fiji is severely curtailed or eliminated. Companies, in light of the Vatukoula situation, and the Namosi tax agreement, will feel, irrespective of whatever legislation or draft agreement is available, that they are in a position to negotiate such arrangements until the power of the minister and Cabinet is tempered by the authority of Parliament to review such arrangements in open and transparent discussion.

While the form of the PNG mining tax legislation would constitute a desirable transplant in Fiji it is by no means evident that the content and substance of various parts of the PNG tax regime is in itself desirable. In particular, the much vaunted additional profits tax that has been seen as the cornerstone of the PNG taxation regime should not be introduced into Fiji. With rates of return that averaged 25 per cent during the 1980s BCL managed to pay additional profits tax on one occasion[71]. In large part this was possible because the additional profits tax regime negotiated with BCL was based on percentage rate of return, but even the cash flow systems make possible distortions in behaviour that will result in avoidance of the additional profits tax even in years when returns are very high[72]. What is needed in the case of Fiji is an additional profits tax system based on the price of the resource which is known and transparent and cannot result in the distortions of mining activity in order to avoid the payment of tax. Moreover, these systems of taxation are easier to administer in countries where the tax administration is in need of strengthening. In light of the Fiji government's commitment to the tax regime at Vatukoula it is highly unlikely that the Government would seek or the company would agree to a renegotiation of so exceptional and generous a taxation regime. The Government of Fiji is perfectly within its legal rights under the VTA to induce the company to negotiate a new agreement by raising the export tax on gold, which it has done on a number of occasions. Given that such a measure would be perceived by the private sector in Fiji as a government attempt to renege on an important agreement and hence a further erosion of its credibility, it is unlikely that the Government of Fiji would pursue such a course of action.

Notes

1. Total mining employment at Vatukoula has varied between 650 and 1,800 since World War II.

2. The paper reviews EGM and EML up until the end of 1992. In 1993 there was a significant corporate restructuring.

3. While the agreement expires in 2004 it is possible that any mine established in that year would enjoy a seven year tax holiday.

4. The company was initially incorporated in 1935 in Victoria, Australia. Then in 1986, when there was a change in Australian tax law, the company was incorporated in the Isle of Man.

5. EGM was incorporated in Fiji in September 1934.

6. Report on the Economics of the Gold Mining Industry at Vatukoula by John Taylor and Sons, London: Consultants to the colonial government, June 1960. It was presented to the Legislative Council of Fiji (Council Paper No. 26. of 1960) Government Press, Suva, Fiji. p. 15 (hereinafter referred to as the Taylor Report, 1960).

7. Fiji Times, 26 July 1950, p. 6 and 2 December 1950. EGM has consistently used the comparison with the Australian taxation regime to argue for cuts.

8. This was to be refunded by June 1961 (Taylor Report, 1960, p. 16). This had been paid from the government subsidy of 1958-1961, and by asking for a refund, the company was in fact asking for the waiving of Port and Service Tax for the three years which in the first place had been paid for by the State.

9. What remains unclear from the historical records is whether the state agreed to further subsidy requests by EGM. Neither the Siwatibau Report nor the parliamentary debates nor secondary sources appear to have suggested that EGM received the requested tax exemption.

10. In 1977 the company announced its intention to close its operations at Vatukoula on 14 February when negotiations with the Mine Workers Union broke down.

11. Stinson played a key role in the defence of EGM and its interests in Fiji.

12. The opposition argued for the nationalisation of the company.

13. Fiji Times, 4 January 1978, p. 1. It was subsequently announced that the company would lay off only 770 workers. Fiji Times, 17 January 1978.

14. Fiji Times, 11 May 1978, p. 1.

15. Fiji Times, 30 May 1978, p. 4.

16. Fiji Times, 18 March 1983, p. 1. No details of the exact financial help offered to Emperor were made public. The Government of Fiji attempted to keep agreement between the State and EGM and Western Mining a secret. However, the terms of the agreement have been published. Unlike all previous tax arrangements between the Government of Fiji and Emperor, this was never to be debated on the floor of Parliament.

17. Siwatibau Report, 1977, p. 37. Recommendation 3: "We do not believe that Government assistance to the Gold Mine over the past ten years has been excessive in light of all the economic and social benefits which the country would otherwise have lost if mining operations were to close down".

18. Address by the CEO Emperor Gold Mines to the Fiji Australia Business Council, Nadi, 15 September 1996.

19. The cost of production figures are derived by multiplying average cost per tonne mined of EGM by total tonnes of ore treated by Emperor Loloma and Dolpin (the Emperor group) and dividing by total Fiji exports of gold. The estimate is imperfect because tonnes treated does not necessarily equal tonnes mined and because the cost data are on a financial year basis with output and exports on calendar year basis. The results may overestimate operating profit/sales as it assumes that all gold exported from Fiji was by Emperor and associated companies. However, as a counterweight the figures do not include silver exports. The significance of the results lies in the trend rather than in the value of the actual profitability estimates.

20. Details of the joint venture appear in annual reports of the Fiji Mineral Resources Department and annual returns of EGM, EML, WMFC and Jubilee Mines Limited (JML) for the years 1983 to 1993.

21. The asset sale as it concerned exploration and mining tenements was implemented by the issue of new tenements to EGM and WMFC jointly covering areas previously held by EGM. Initial grants were Special Mining Lease (SML) S.54, Special Site Right (SSR) No. 4, SSR No. 5, to EGM and WMFC 90/10 and subsequently 80/20. Special Prospecting Licence (SPL) S.1201; SML S.55; to EGM and WMFC 50/50. Subsequently SSR No. 4 was converted to SML No. 56. A further prospecting interest, SPL S1271 was issued to EGM and EMFC 50/50 in 1985 and renewed in part in 1989. In 1989 SSR No. 5 was replaced by SSR No. 6 and SPL 1208 was converted to SML 1344.

22. Details of attempts to exploit tailings appear in annual reports of the Fiji Mineral Resources Department and annual reports of EML for years 1985-1993.

23. The VTA, unlike the tax free factory system which is widely publicised, has never been published by the Fiji government.

24. The relevant legislation for the period under review includes the Income Tax Act (1974) as amended, Stamp Duties Act (1971) as amended, Mining Act (1966) as amended, Customs Act (1986) and the Customs Tariff Act (1986), as amended.

25. The Fiji Government is of course as much subject to the laws of Fiji as anyone else and may not contract to vary tax obligations imposed by law. Each of the concessions granted or to be granted under the VTA must be referable to the discretionary powers of the Minister of Finance under existing legislation and the Minister's discretionary decision must be expressed in the form required by law. This paper neither debates nor considers the legal efficacy of the VTA.

26. Stamp duty is tax imposed on documentation effecting a transaction. Duty is a small percentage of the transaction price with different transactions attracting differing rates of duty. The total price of the asset sale was $4,200,000.

27. Income Tax Act, s. 11(k).

28. Income Tax Act s. 19(1) and 23(2)(b).

29. Stamp Duty Act s. 4(a) empowers the Minister of Finance to waive duty payable on any instrument.

30. Fiji does not tax capital gains. Asset sale moneys would be a capital receipt for EGM and thus excluded from taxable income on general principles. Hence this provision only benefits EGM with respect to recouped depreciation included in taxable income under the Income Tax Act, s. 11(k). There is no statutory foundation in the Act for this concession.

31. Income Tax Act, s. 23(4) empowers the Minister of Finance to authorise an immediate write-off of acquisition costs.

32. Income Tax Act, s. 16.2(a) empowers the Minister of Finance to grant tax holidays to mining companies. From 1980 income tax rates for resident companies were 37.5 per cent and from June 1992 35 per cent (ITA, Fourth Schedule).

33. The VTA does not offer a geological or mining engineering definition of a new mine that would allow the Fiji Commissioner of Inland Revenue to make a distinction between a new mine and new shafts.

34. Income Tax Act, s. 23(1).

35. Relying on Income Tax Act, s. 23(4), see Income Tax Act, s. 11(k).

36. Mining Act, s. 3.

37. Mining Act, s. 54.

38. Mining Regulations, ss. 6 and 7.

39. Customs Act, s. 6.

40. Under the Terms of the Value Added Tax Decree, 1992 export sectors are exempt from VAT.

41. Relying on the Customs Tariff Act, 1986, s. 11.

42. Concessionary treatment on customs and fiscal duty is conditional on plant, equipment and services being purchased locally where local terms equal or better terms obtainable elsewhere. This standard form conditionality is essentially to promote local industry but incidentally also discourages transfer pricing with regard to items available locally.

43. The tax on dividends is known as non-resident dividend withholding tax, Income Tax Act, s. 8. This is a final tax, Income Tax Act, s. 17(39). The tax on interest is known as interest withholding tax, Income Tax Act, s. 9 and is not technically a final tax, Income Tax Act,s. 104(2). Know-how payments are subject to the dividend withholding tax, Income Tax Act, s. 8(2)(c). As regards know-how payments this is not technically a final tax, Income Tax Act, s. 104(1).

44. Act 25 of 1983.

45. Decree No. 10, 1990.

46. Relying on the Income Tax Act, s. 14(2)(e).

47. Relying on the Income Tax Act, s. 8(6)(c).

48. Garnaut and Clunies (1983), see also Otto (1995) for a contemporary treatment of the subject.

49. The mining companies in PNG have, until 1992, only paid additional profits tax on one occasion in 1980 pers comm, Commissioner of Revenue, 1992.

50. The new mines of the Tavua Basin Mine which began operation in 1988 had recovery grades of between 36.13 g/t in 1988 falling to 20.36 g/t in 1991. In May 1991 the mine was closed after a fire was set to the surface infrastructure. See "Emperor Mines Ltd", Australian Stock Exchange 1993, p. 2.

51. See "Emperor Mines Ltd", Australian Stock Exchange 1993, p. 3.

52. "Report of the Auditor-General on the accounts of the Government of Fiji for the year 1989", Paper No 3 of 1991, Vol II p. 23. The Auditor-General had asked the Permanent Secretary of Finance where the agreement had been gazetted. The authors have also been unable to find the gazettal notice.

53. In its reply to the Auditor-General's statement the Mineral Resources Department defended the decision to grant the VTA on the grounds that the development would not have occurred without the concessions. With recovery grades at 36g/t it is difficult to imagine that tax concessions of such an order of magnitude were absolutely necessary. (The response of the Mineral Resources Department was provided by the Fiji Auditor-General Office).

54. This is operating profit including extraordinary items, unless otherwise stated.

55. If it were not for the relatively rigorous requirements of the Australian Stock exchange no reliable statistical research on Emperor would be possible. The Fiji government reporting requirements leave companies with very little reporting responsibility.

56. Total employment for Emperor was available for certain years from various government reports but it was by no means complete. As a result, for the years 1970-78 employment at Emperor was employed and for the remaining years total employment in mining and quarrying was used as a proxy.

57. It should be noted that other variables such as output and head grades were also insignificant even for EML.

58. If TPM activities are occurring or the company is simply making efficient use of its various tax havens the company's profits should reflect the fundamental variables that determine the profitability of any mine, i.e. output levels, price etc.

59. The period 1970-1986 was chosen because it was prior to the operation of WMFC joint venture and the operations of Jubilee and Koula.

60. The Siwatibau Report which covers this period does not indicate the existence of significant TPM. However, the report suggests that Emperor (NZ) was charging EGM 5-7 per cent of input charges as a procurement commission for inputs purchased in New Zealand. See Siwatibau Report, 1977, p. 15.

61. In its full page advertisement in which Emperor discusses at length its tax payments to the government of Fiji it does not suggest that in the 1987-1991 period it has paid any corporate tax (The Daily Post, 5 September 1992, p. 38).

62. Australian Stock Exchange, "Emperor Mines Limited", Company review service, 1993, p. 11.

63. By way of facile comparison, Emperor's corporate tax liability in 1987, the last year in which it made provisions, was less than half the liability paid by a senior lecturer at the University of the South Pacific.

64. This includes royalties paid to the Mineral Resources Department under the terms of the Mining Act. While no royalty is prescribed under the terms of the Income Tax Act the royalty was paid to the Department of Finance.

65. In December of 1987 the Government reintroduced gold export taxes at 2 per cent. In November 1989 it was raised to 5 per cent and then lowered again to 2 per cent in November 1991. This was raised once again to 3 per cent in 1992.

66. Under the PNG mining tax regime existing in the late 1980s and early 1990s companies pay royalties of 1.25 per cent of fob to landowners and provincial government. They paid 35 per cent corporate tax, and 70 per cent of all positive net cash flow for rates of return above 20 per cent or US prime plus 12 per cent. Companies also have to offer 20 per cent equity to the government at historical cost.

67. This period was chosen because it corresponds to the period under study of the Vatukoula Tax Agreement. It does not include 1992 because in that year PNG began to export oil and this sector, while contributing to the MRSF, is subject to a different tax regime to that of the mining companies.

68. There remains one possible criticism of the approach taken above and that is that comparing EML and Bougainville is unfair as BCL is five times larger than Emperor, is open cast, capital intensive and very profitable and as a result the returns that the Government of PNG obtained during this period may be a result of the economies of scale of the operation rather than the mining taxation regime. There are two possible replies to this line of argument. Our estimate is based on MRSF returns divided by exports, which includes both the very large and profitable BCL operation and the equally large but totally unprofitable Ok Tedi mine. Thus the PNG government take of mineral exports is a reasonable reflection of what government can expect from both profitable and unprofitable mines. Moreover, the lower estimate made above of 8.55 per cent of export revenue is downward biased by the fact that it includes two years, 1990 and 1991, when BCL made no profits because its operations were closed. Throughout the period Ok Tedi paid nothing to the Government except the 1.25 per cent royalty. Thus I would contend that despite the effect of economies of scale on government revenue take the 8.55 per cent figure does not constitute an unreasonable figure for a reasonably profitable mine operating under a normal taxation regime in a developing country. It should be kept in mind that BCL, during the 1980s, on average paid the Government of PNG some 20 per cent of export earnings.

69. This is based on the assumption that average gross annual wages in the period were F$8,000 per miner and that there was average employment of 1,200 over the period (see 'Atu Emberson-Bain, 1994, p. 211).

70. Following the completion of this paper in November 1995, the Minister of Trade and Commerce of Fiji, the Hon. Jim Ah Koy, announced during the 1996 budget session in Parliament that he had come to an agreement with Placer over the tax agreement for the Namosi mine. While he did not clarify the exact nature of the tax agreement he did say that the tax agreement Placer had been offered "... slightly more than what is offered to the Tax Free Factories and would encourage the company to continue further development work and progress towards a full feasibility study on proposed mine", Fiji Times, 4 December,p. 7. If this is in fact the case then Placer will be offered a minimum 13 year tax holiday, duty free imports and exemption from Fiji's value added tax.

71. Personal Communication, 1992, PNG Commissioner of Taxation.

72. The BCL additional profits tax was not based on a percentage of cash flow but rather used a 15 per cent rate of return on investment as the threshold where the additional profits tax became operative. This created enormous potential for stockpiling and other activities to raise inventory investment and thereby decreasing apparent rates of return on investment.

References

1. 'Atu Emberson Bain (1994, Labour and Gold in Fiji, Cambridge University Press, Cambridge.

2. Commonwealth Fund for Technical Co-operation - Technical Assistance Group (1990, Financial Terms of Mineral Exploitation in Papua New Guinea - A Comparison with International Terms, London, pp. 69-73.

3. Department of Inland Revenue, Annual Report for the Year 1979, Paper No. 45 of 1981, Fiji Government.

4. Department of Inland Revenue, Annual Report for the Year 1985, Paper No. 20 of 1986, Fiji Government.

5. Department of Inland Revenue, Annual Report for the Year 1986, Paper No. 16 of 1987, Fiji Government.

6. Emerson, C., "Mining in Asean, Australia and PNG", Asean - Australia Economic Papers No. 14, Kuala Lumpur Asean Australia Joint Research Project..

7. EML (1991, Annual Report 1991, Emperor Mines Limited, Fiji.

8. EML (1993, Annual Report 1993, Emperor Mines Limited, Fiji.

9. Fiji Government (1977, Parliamentary Debates, House of Representatives, Vol. II of 1977, October-November.

10. Garnaut, R. and Ross, A.C. (1983, Taxation of Mineral Rents, Clarendon Press, Oxford.

11. Grynberg, R. and Powell, M. (1992, "Taxation and development in forum island countries", Vol. I, Forum Secretariat, Suva, Fiji.

12. IMF (1994, "International monetary fund", International Financial Statistics Yearbook - 1994, Washington, DC.

13. Johnson, C. and Pintz, S."Minerals and developing countries" in Vogley, W.A. (Ed.), Chapter 12 inEconomics of Mineral Industries, 4th ed., Society of Mining Engineers of AIME, New York, NY.

14. Kasper, W. (1988, Fiji: Opportunity from Adversity?, Centre for Independent Studies..

15. Mineral Resources Department, Report for the Year 1980, Fiji Government..

16. Mineral Resources Department, Report for the Year 1984, Fiji Government..

17. Mineral Resources Department, Report for the year 1989, Paper No 30 of 1991, Fiji Government.

18. Otto, J.M. (1995, Taxation of Mineral Enterprises, Graham & Trotman, London.

19. Pacific Islands Development Programme (1990, "Analysis and review of gold mining taxation, Papua New Guinea, Solomon Islands and Fiji", University of Hawaii, Honolulu.

20. Palmer, K. (1980, "Mineral taxation policies in developing countries", IMF Staff Papers, September, pp. 517-92.

21. Siwatibau Report (1977, Report of the Committee Appointed to Examine the Economics of the Gold Mining Industry at Vatukoula, S. Siwatibau (Chairman), 18 April (committee appointed by the Minister of Labour).

22. Taylor Report (1960, Report on the Economics of the Gold Mining Industry at Vatukoula, by Consultants John Taylor and Sons, London, Legislative Council Paper 26 of 1960, Government Press, Suva, Fiji.

23. Tilton, J., Milet, J. and Ward, R. (1986, Mineral and Mining Policy in PNG, Port Moresby, PNG, Institute of National Affairs.

24. Walrond, G. and Kumar, R. (1986, Options for Developing Countries in Mining Development, St Martin's Press..

25. Ward, M. (1971, The Role of Investment in the Development of Fiji, Cambridge University Press, Cambridge.

[Illustration]
Caption: Figure 1; Fiji exports of gold and employment in the gold mining industry; Figure 2; Estimated average cost, revenue and rate of operating profit for Emperor (1950-1970); Figure 3; Emperor Mines Limited corporate structure (1992); Table I; EML financial performance 1986-1992 (AUD 000s); Table II; EML profits attributable to Fiji operations (AUD 000s); Table III; Operating profits of EML subsidiaries in Fiji (AUD 000s); Table IV; Comparison of various international mining tax contract terms with the Vatukoula tax agreements; Table V; Determinants of EML and EMG profitability; Table VI; Royalties paid to the Government of Fiji; Table VII; Nominal levels and effective rate of export tax on gold; Table VIII; Estimated and actual government revenues under Fiji and PNG tax regimes; Table IX; Determinants of EGM tax liability (1970-1985)