Report No. 9620-TUN Republic of Tunisia Changing the Structure of Incentives November 191l Countr Operations Division Country Department 11 Europe, Middle East and North Africa Region FOR OFFICIAL USE ONLY 4> -~~~ 7>, Ai ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~.. 4444 St7. .N'.~~~~~~~~~~, -e &~~~~~~44 we f m* fiW df U Ns . '4 V~~~~~ '~~~*~~Vbd FOR OFFICIAL USE ONLY REPUBC OF TUIISI CHANGING THE STRUCTURE OF INCENTIS TABLE OF CONTENTS Page No. EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . I. TRADE POLICIES AND INCENTIVES . . . . . . . . . . . . . . .1 A. INTRODUCTION . . . . . . . . . . . . . . . . . . . .1 B. TRADE LIBERALIZATION AND THE STRUCTURE OF INCENTIVES 1 C. LIBERALIZATION AND EXTERNAL TRADE . . . . . . . . . . 7 Trade Reform and Import Behavior . . . . . . . 7 Export Growth and Exchange Rate Policies . . . iO Export Behavior . . . . . . . . . . . . . . . . 13 Implications of Accelerating the Trade Reform . 15 D. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . 15 II. FISCAL INCENTIVES FOR INVESTMENTS AND EXPORTS . . . . . . . 19 A. INTRODUCTION . . . . . . . . . . . . . . . . . . . . 19 B. RECENT TAX REFORM . . . . . . . . . . . . . . . . . . 19 Calculating Tax Wedges . . . . . . . . . . . . 20 C. INTERNATIONAL COMPARISONS . . . . . . . . . . . . . . 24 D. INVESTMENT INCENTIVES .26 Incentives to Foreign Investors . . . . . . . . 28 E. CONCLUSIONS .31 III. FINANCIAL INCENTIVES TO INVESTMENT AND EXPORTS 34 A. INTRODUCTION . . . . . . . . . . . . . . . . . . . . 34 B. THE FINANCING OF CORPORATE INVESTMENT . . . . . . . . 35 C. REGUIATORY CONTROLS ON THE BANKING SYSTEMl 37 D. EXTERNAL FINANCING .41 This report ws prepared by a mission which visited Tunis in November 1990o Th mission conpift d of Miria Figato (Task Manager), Azita Amjadi, Riccardo Faini. Olivier Godron and Henri Lamotte. Michael Devereux and Mark Pearson prepared a background paper for Chapter II. At the time this report was prepared Kemal Dervis was the Director and Mahmood Ayub was the Division Chief. H614ne Talon typ the report. I This, do-urr-nt h '°roetrirts-r dliStrihUtinn and ry V he ite,*d hv roarinienv onlv in the perform nce l Page No. E. THE INSTRUMENTS OF SELECTIVE CREDIT . . . . . . . . . 43 Interest Rate Subsidies . . . . . . . . . . . . 43 Special Purpose Government Funds . . . . . . . 46 Shortcomings of the Current Selective Credit System . . . . . . . . . . . . . . . 48 F. FINANCIAL MARKETS . . . . . . . . . . . . . . . . . . 52 Institutional and Regulatory Framework . . . . 53 Feat"res and Tax Treatment of Financial Ins trtunents . . . . . . . . . . . . . . . . 54 Tern. Structure of Interest Rates . . . . . . . 58 G. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . 61 Foreign Exchange Risk . . . . . . . . . . . . . 61 Selective Credit Policies . . . . . . . . . . . 62 Financial Markets . . . . . . . . . . . . . . . 63 ANNEXES ANNEX I STATISTICS ANNEX II THE MODEL OF THE TRADE SECTOR ANNEX III THE TUNISIAN TAX SYSTEM ANNEX IV METHODOLOGY OF CONSTRUCTING TAX WEDGES AND EFFECTIVE TAX RATES REPUBUC OF TUNISIA CHANGING THE STRUCTURE OF INCENTIVES EXECUTIVE SUMMARY 1. In the context of the preparation of the Eighth Development Plan (1992-96), the Tunisian authorities are taking stock of the results of the reform program latvnched in 1986 and discussing the economic strategy to be adopted in the future. This program, aimed a- transforming the country into an efficient and diversified outward-oriented economy, included: a reform of the trade incentives through the rationalization of tariff rates, the liberalization of domestic prices, and the reform of direct and indirect taxation; the promotion of domestic competition, through the elimination of direct regulatory impediments to entry and a reform of investment incentives; and the e,icouragement of import competition through the gradual elimination of quantitative restrictions. Measures to improve competition among banks and to reform the public enterprise system were also undertaken. 2. Few would deny the importance of the changes in the structure of the economic incentives generated by the program and the resulting strong recovery during 1989-1990. GDP increased by 7.1% in 1990 compared to 3.7% in 1989, and 1.1% in 1988 when it was influenced by a severe drought. Real investment increased by 35% ir. 1989 and by 28.1% in 1990 after much stagnation. Manufactured exports have grown, on average, by 17.3% in real terms since 1986. In 1990, however, the upsurge in imports, mainly of capital goods reflecting the boom in investments, resulted in a current account deficit of 5.3% of GDP. The political instability in the region, following the Gulf crisis, has created uncertainty in the economic prospects for the near future. The sharp drop in tourism revenues, already manifested in early 1991, is expected to be above 34% by the end of the year (or 13% of total export earnings). This comes at a time of low international reserves, estimated at 1.7 months of imports at the end of 1990. The authorities have already taken stringent fiscal measures to reduce domestic absorption and to prevent a further deterioration of the current account. Balance of payments support has also been sought from the IMF and the World Bank. 3. This report reviews the results of the reforms initiated in 1986. It focusses on those aspects of the reforms where more policy actions will be needed to produce a structure of economic incentives that better fits the outward looking strategy of the Government and its objective to increase the efficiency - ii - of resource allocation. Building on previous Bank reports,' it asiesses the results of the trade reform (Chapter I); the effect on investment decisions of fiscal incentives (Chapter II); and the financial incentives (Chapter III). The underlying conclusion is that, although the structure of incentives has changed significantly in the last few years, the need to persevere with structural reforms remains strong, given the country's continued vulnerability to exogenous shocks. As discussed in other Bank reports, sustained economic growth needs to be supported by more efficient investment as well as by a sready increase in exports. A major constraint to achieving these objectives is the deep-rooted dualism within the industrial sector between wholly exporttng firms and firms producing for the domestic market. On the one hand, firms producing for the domestic market are sheltered from international competition and enjoy substantial tax exemptions and access to preferential credit. On the other hand, exporters are insulated from the negative effects of the policies aimed at protecting domestic markets through free access to imported inputs, generous fiscal and financial incentives, and the active management of the exchange rate conducted by the authorities. The continuation of reforms in the trade, fiscal and financial sectors is crucial for unifying the industrial sector, thus allowing the spill-over of efficiency gains from export industries to the whole economy. Such a continuation represents a major challenge for the Tunisian Government soon. More generally, a time-bound, non-discretionary incentive framework capable of correcting market failures without creating distortions and rents is needed. The Reforms of the Eighties: Remaining Issues 4. Trade liberalization: More efficient resource allocation and providing equal incentives to domestic sales and exports were at the heart of the 1986 reform program. The program included the decontrol of domestic prices and distribution margins, the rationalization of the trade taxes, reductions in the tariff rates, the devaluation of the dinar, and the gradual elimination of most quantitative restrictions (QRs). The initial emphasis of trade liberalization was on the rationalization nf the tariff structure: the range of customs duties was compressed from 5%-235X in 1986 to 15%-43% in 1990. Capital goods were most affected, vith the average rate o. nominal protection decreasing from 26.1% in 1987 to 21.2% in 1990. However, for the economy as a whole, the import-weighted tariff only fell to 22.3% in 1990 from 24.9% in 1987. 5. The early reduction in tariff rates in 1987-88 made the elimination of QRs more difficult later. Thus, the revision of non-tariff barriers proceeded cautiously. From its inception, the liberalization program involved mainly intermediate and capital goods, i.e., commodities that were not domestically produced and that represented essential inputs for the exporting industries. j/ See in particular: "Tunisia: Country Economic Memorandum: The Road to an Outward-Oriented Economy", World Bank (1990), Report No. 8044-TUN; and "Tunisia: Industrial Sector Note", World Bank (1990), Report No. 8277- TUN. - iii - Indeed, since 1987, as an incentive for exports, firms exporting a minimum of 15% of their turnover are allowed to freely import all inputs needed, irrespective of their general policy regime. The importation of products subject to non-tariff barriers is permitted by means of an import authorization or a licence, which is liberally granted. However, the share of freely importable goods, measured in terms of domestic output, was only 28% at the end of 1990. Thus, overall, the liberalization program did not go far enough to expose domestic producers to import competition. Moreover, rhe earlier liberalization of intermediate inputs, without the liberalization of the related finished products increased the rate of effective protection in many activities in recent years. For example, the rate of effective protection of the manufacturing industries producing for the domestic market increased f:om 78% in 1988 to 87% in 1989. Protection, by raising the profitability of domestic sales and by failing to expose firms to international competit-lon, is also responsible for the lack of integration between exporting and non-exporting firms. An example of this weak integration is that, in spite of the numerous incentives granted to indirect exporters, more than 90% of the inputs purchased by exporting firms are from sources outside Tunisia. 6. Trade liberalization and current account gerformance: Evidence from both developing and developed countries suggests that trade reforms are likely to produce current account imbalances in the short run because import flows respond more rapidly than exports to liberalization measures.2 Thus, the prediction of the response to reform measures of imports and exports is of paramount importance in the implementation of the liberalization program. The view taken by the Tunisian authorities in the adjustment program was that export growth, as opposed to import compression, would have to be responsible for achieving external balance. In this respect, exchange rate policies played a major role in raising the competitiveness of Tunisian goods in recent years. The response of manufactured exports to the significant real depreciation of the exchange rate in the mid and late eighties was substantial.3 Remarkably, the improvement in competitiveness of Tunisian goods was noticed not only for industrial countries but also for competing countries such as Morocco, Turkey, and Greece. As to import demand, the econometric estimates presented in this report indicate that imports of consumer goods are very responsive to changes in trade policy, especially in quantitative restrictions. However, imports of both consumer and capital goods, as well as export flows, also respond strongly to changes in the exchange rate. This would suggest that policies to increase 2, See for example Khan and Zahler, "Trade and Financial Liberalization given External Shocks and Inconsistent Domestic Policies", IMF staff papers, 1985. 2/ The real exchango rate (based on the wholesale price index) depreciated about 11% from 1984 to 1986 and a further 1.5% in 1988. After registering a growth rate of just above 6% during 1980-85, the manufactured exports recorded (see paragraph 2) an average annual rate of 17.3% since 1986. iv - export growth, particularly careful management of the exchange rate, may be sufficient to avoid a large deterioration in the external balance as quantitative restrictions are eliminated. 7. Overhaulin: the tax system: the 1988-1990 Kreform: The stated objectives of the comprehensive tax reform enacted in 1988-90 were to increase fiscal neutrality and equity and to reduce distortions in the incentives to save and invest. The reform package, expected to be revenue neutral, Included the introduction of a Value Added Tax (VAT) on production in 1988, its extension to wholesale trading in 1989, and the adoption of new personal and corporate income tax laws in 1990. The implementation of the VAT has proceeded smoothly, and the revenue performance has been better than expected. International comparisons of marginal t"x rates indicate that the new direct tax laws respond better than those in most OECD countries to the objective of neutrality (e.g., towards debt/equity choices, pay-out ratios, the incorporation decision, anmong others). For example, (a) the alignment of the maximum corporate tax rate with the top rate of the personal income tax reduces the incentive to convert labor income into the (f;-rly) less taxed corporate income; (b) the simplification of the tax rate structure reduces opportunities for tax avoidance while tne reductic- in the top rates may weaken the incentive for tax evasion; and (c) the double taxation of companies' income, at the corporate and personal level, has been eliminated. Dividends are exempt from personal tax. However, as in most OECD countries, the corporate tax laws present a bias toward debt financing as interest payments are tax deductible. 8. The Investment Codes: a legacy from the past. Undermining the simplicity, neutrality and revenue potential of the new tax law is a system of incentives contained in several Investment Codes. They are primarily aimed at encouraging investments in the industrial, tourism, agricultural, fishing and -trading sectors although they also promote "horizontal", non-sectoral objectives such as export growth and regionally br-lanced developmer.t. The extensive sectoral targeting is difficult to justify. It may have been necessary in the past to offset the side effects of distortive trade and fiscal policies, administered prices and underdeveloped capital markets. As these distortions are being addressed, however, there is less need for industry-specific State intervention. 9. The incentives contained in the Codes include partial or permanent exemption from payment of registration, income and trade taxes; tax deductions on reinvested profits; interest rate rebates; and capital subsidies. There is no apparent rationale for the use of so many different instruments. They contribute to an increased rate of dispersion of the cost of capital among different investment projects and sources of finance and may result in arbitrary and unintended distortions. For example, interest rate subsidies reduce the cost of capital only to debt financed projects, which already benefit from the incentive of deductible interest payment. On the other hand, some instruments, in particular tax holidays, are poor and expensive means of attracting investments. This is so because firms must write off depreciation allowances during the tax holiday. This implies that when the tax holiday is over, firms may still pay taxes on income generated by investments tindertaken during the holiday which, in most cases, have already been depreciated. Clearly, for a tax holiday to be truly effective, firms should be allowed to defer depreciation until after the holiday. Ti,s provision cannot be recommended, however, as the foregone tax revenue of this incentive is relatively high. 10. Financing the corporate sector: the bias toward debt financing: The financing patterns of corporate investment in Tunisia reveal a very low degree of self-financing. During 1979-1987, internal sources of funds only accounted for 20% of investment financing and new share issues for an additional 19%. A great part of the financing was provided by debt in the form of short-term loans (22%), medium- and long-term loans (26%), and by government subsidies (13%). The heavy reliance on debt has largely depended on the easy credit and on its low real cost. Both elements have been the result of Government intervention aimed at influencing the allocation of financial resources. This has included: selective policies to channel credit to priority sectors of the economy; preferential rediscount at the Central Bank; interest rate subsidies; repayable advances;; investment grants; regulatory controls to influence the lending and pricing decisions of commercial and development banks; and recourse to external funds at: a cost lower than that of domestic funds since the former excluded the exchange risk. 11, During the eighties, the percentage of selective credit, relative to total credit to the economy, was arou.d 25% (declining to 20% in 1990). Financial incentives have created a strong bias toward debt financing by encouraging an artificial demand for loans from firms unable to obtain credit at market rates. The minimum proportion of self-financing being required on the projects that benefit from selective credit is extremely low, ranging between 10% to 40% of the project cost. Repayment of government loans granted through public funds is generally rare, and loan collect-on is poor. According to a recent study, 40% of the enterprises financed by a *,overnment fund, the FOPRODI, went bankrupt or faced serious difficulties. Moreover, the absence of a centralized monitoring system makes the overall assessment of the cost of the selective credit schemes particularly difficult. Estimates provided in this report indicate that in 1990 the budget cost was equivalent to 24% of the budget deficit, or 0.8% of GDP. 12. Regulatorv controls. During 1987-88, many reforms were implemented by the Central Bank, including the elimination of the Central Bank's prior approval system applied to most lending; the freeing of most interest rates; and the development of a money market. Despite these reforms, regulatory controls on the banking system, resulting in the segmentation and fragmentation of credit markets, are still important in Tunisia. Remaining regulatory controls on commercial banks include a cap of 300 basis points above the money market rate (TMH) imposed on lending rates, and a ceiling of 200 basis points below the TMM for the rate on special savings accounts. The lending rate cap was established in 1988 to prevent a general rise in interest rates that could have been engineered by banks inexperienced with competitive loan pricing. As price competition, within the cap, has indeed developed, continuing the ceiling may hamper the ability of banks to further differentiate their rates according to the quality and creditworthiness of borrowers. On the other hand, the ceiling on the rate of special savings accounts appears to be excessively high relative to the yield of longer term instruments. Special savings accounts compete - vi - unfairly with other types of financial instruments; being an expensive resource, they also reduce the banks' operating margins. 13. A further constraint on the activities of commercial banks is the prohibition to offer long-term loans. However, even if this constraint were relaxed, commercial banks would probably be prevented from granting long-term loans by their weak capital structure, relative to that of the development banks. The discrepancy in the capital structure of commercial and development banks, if not addressed, is likely to perpetuate the segmentation of the credit market and eo inhibit the emergence of market-based pricing policies. Finally, commercial banks are required to comply with a detailed system of credit guidelines defining the purpose, maturity and amounts of the loans they can grant. Because of a lack of audited accounts and difficulties in screening viable projects, the credit guideline system is used as a poor substitute for a more comprehensive system of supervisory and prudential regulations. For example, it does not preclude a bank portfolio from being excessively exposed to certain economic sectors or poorly matched to resources in terms of maturities. It induces banks to provide, and borrowers to request, loans simply because they are included in the guidelines, irrespective of real needs. It dilutes and discourages the responsibility of banks to screen and monitor their loans. 14. External financing. During the eighties, the complete coverage by the state of the foreign exchange risk on external loans raised by financial institutions encouraged the financial institutions to borrow abroad in currencies carrying a low interest rate. At the end of 1989, funds originating from foreign credit lines represented, respectively, 8% and 30% of commercial and development banks borrowing. Until 1988, the foreign exchange losses arising from external loans were covered by a guaranteed fund, the "Fonds de Perequation des Changes et Taux d'Int6ret", which was greatly underfunded. As it became clear that the fund could not break even unless the dinar appreciated substantially, the authorities have looked for new solutions since 1989. Hedging instruments, such as a forward cover and options have been introduced to cover the foreign exchange risk but only on short-term loans. "Ad hoc" mechanisms were agreed to between the Central Bank and financial institutions for the few concessional credit lines undertaken in 1989 and 1990. 15. Financial markets development. Historically, the predominance of debt as a source of finance for investments has been mirrored in the underdevelopment of financial and capital markets. The development of these markets has been a central objective of the authorities since 1987. Among the measures taken is the establishment of a new comprehensive set of market regulations by the Stock Exchange authorities. The tax treatment of financial instruments, of equity and bonds in particular, has been streamlined and simplified under the recent direct tax reform. However, both the stock and the bond markets are still in their infancy. The Treasury's issuing policies have not contributed to creating a strong bond market, as they have emphasized compulsory placements of low yielding "equipment bonds" by the banks (currently amounting to 20% of their deposits) and insurance companies (50% of their technical reserves). A major reform of the money market was implemented in 1988-89. Competition has been encouraged by the introduction of new financial - vii instruments, such as certificates of deposit, commercial paper and short-term Treasury bills. Commercial and development banks, as well as non-financial enterprises have been allowed to participate in this market, thereby reducing the segmentation in the market for resource funds. More recently, to encourage investment in securities, two types of investment funds, namely closed-end (SICAFs) and open-end funds (SICAVs), have been introduced. 16. Although the tax treatment of financial instruments appears to be satisfactory, some tax incentives granted to purchases of securities in the primary market are detrimental to the development of the secondary market. Under the Law 62-75 of December 1962, purchases of shares and bonds, previously approved by the Minister of Finance, are deductible up to 35% of taxable income. However, to enjoy this tax advantage, the securities must be held for two years (for stocks) and five years (for bonds), thus discouraging secondary trading. Contradictions in the existing tax treatment for SICAVs are also delaying their creation. For example, on the one hand, the interest income, which is normally taxed if securities are held directly, is not if distributed as a dividend by a SICAV. On the other hand, the 1991 Finance Law has excluded SICAVs (but not SICAFs) from the benefit of the tax deductibility provided by the Law 62-75 to holders of securities. POLICY RECOMMENDATIONS 17. The challenge faced by the Government of Tunisia in the coming years is to consolidate the economic recovery initiated in the late eighties and to achieve a sustained growth in the nineties. In line with the objectives of making the economy more outward-oriented and increasing the role of the private sector, policy intervention should be focussed on improving market efficiency. To this end, the Government should continue the structural reform program initiated in the mid-eighties to improve the system of economic incentives in three main areas: trade liberalization, rationalization of fiscal incentives, and financial sector reform. 4 Trade Liberalization! 18. In view of the delays incurred in the last two years, largely due to unfavorable exogenous factors, the acceleration of the import liberalization program is a priority. This would stimulate productivity and improve the quality of domestically produced goods. It would enhance consumers' welfare, as domestic prices approach international prices. It would eliminate the indirect costs i/ Complementary issues relating to industrial restructuring, administrative reforms, labor markets and human capital deve7opment have already been discussed in recent Bank documents (for example, in the recent "Employment and Training Fund Project", Report No. 8630-TUN and in the "Industrial Sector Note", Report No. 8277-TUN) and are the subject of ongoing discussions between the Bank and the Tunisian authorities. - viii l ansoclated with the management of import controls and special exemption schemes O.lowed to exporters. The program for the elimination of quantitative re)strictions should, therefore, be continued. It should be accompanied by ,wtausitional measures designed to ease the adjustment costs of previously pr)otectod enterprises. These measures could take the form of temporary tariff ,to.:ixeurg1ayt whiLch would maintain oxisting protection levels in the short run. *,aduuLily, these surcharges would then be lowered to allow imports to be ()#8pnded. The lntrodu='tion of tariff surcharges would increase the credibility oi: the. liberalizatlon program and reduoe existing political pressures against it. At the same time, an active exchange rate policy, aimed at increasing the (ornpetitiveness of exports, would be required to encourage export growth and to prevent a large deterioration in the current account. In particular, the otnirho,ities shoutld aim at maintaining che competitiveness of Tunisian goods with respeCt to their competitors in the developing countries. Finally, decontrol of: producer prices and of distribution margins should be continued in combination JiJ Lh t.hr removal of import restrictions to promote competition within the l al sector. The decision to continue import liberalization would enhance the process of structural integration within the industrial sector between fully exporting firms and firms producing for the domestic market. The Government could take a step in this direction by encouraging exporting firms to sell part of? their output in the domestic market. This measure would introduce some competition between goods produced for the domestic market and those for export, which are typically of greater quality. Some immediate measures could also be teker- to improve administrative and customs procedures for exports. In particular, partly exporting firms should be exempted from the obligation to post a bond as a guarantee for unpaid customs duties on imported inputs for their Cxport activities. This required bond, now equal to 5S of the value of imported goods, represents a significant financial burden for these firms. In addition, ghe simplified temporary admission procedure should be extended to all indirect exporters. At present, exporting firms acquiring inputs from domestic firms rely on a duty drawback procedure, where a refund is claimed for the customs duties paid by indirect exports. This procedure is cumbersome and discourages exporting firms from buying intermediate inputs from domestic firms. Fiscal Incentives 20. In line with the greater transparency and economic efficiency introduced by the tax reform of 1989-90, the incentive structure contained in the investment Codes should be rationalized. Government objectives should shift from being industry and sectoi specific to being "functionally" oriented (e.g., to promote infant industries, technology transfer, the development of disadvantaged areas, among others). Promoting exports, although outside the domain of market failures should still be pursued until trade restrictions and price controls have been eliminated. The existing Investment Codes, which target sectoral activities, should be replaced by a single Code. This Code would harmonize the principal incentives and eliminate those which address sectoral issues only. Fewer instruments would be less expensive to administer and their e£fects would be more easily understood by both Government and industry. They - ix - would also reduce the risk of complex interactions that could produce unintended distortions. The incentives should be chosen according to the criteria of coss- effectiveness and efficiency. Thus, they should be offered for a limited time to minimize their budgetary cost and avoid the creation of enterprises surviving only because of the concessions. Moreover, only those instruments which coul.d affect the targeted activities in a direct and unique way should be preferred. Examples of such incentives are: (a) Accelerated depreciation allowances, which allow firms to increase the rate at which capital can be written off against tax. They reduce the cost of capital and, by benefitting only companies with sufficient investment, imply lower foregone revenues to the budget, relative to tax exemptions and holidays; (b) Investment allowances benefit firms not in proportion to their profits but to the size of their investments. They are simple to monitor and their cost can be easily calculated; and (c) PJIx.est investment grant from budget resources. They constitute an effective way of reducing the cost of capital and, if properly accounted for in the goverinent's budget, they constitute a transparent instrument of subsidization. Financial Sector Reform 21. To consolidate the liberalization process begun in 1987, financial sector reforms would need to aim at three objectives: (a) increasing competition in the banking system and eliminating segmentation in the credit market; (b) eliminating selective credit policies; and (c) developing active securities markets. 22. Comnetition between financial institutions could be increased by eventually eliminating the controls on the lending and deposit rates of commercial banks. Some transitory measures could be taken immediately. The cap on the lending rates could be applied on the average of the total portfolio of each bank, as opposed to a loan by loan basis. This measure would increase the flexibility of the banks' pricing policy while still ensuring the control of lending rates on an adminietered basis. The current ceilf g on special savings accounts should also be relaxed from the current level of money market rate minus 2% to money market rate minus 3% or 4%. This measure would still grant small savitngs a positive interest rate, while it would address larger deposits towards more remunerative and less liquid instruments. A further measure to increase competition among financial inst'.tutions would be to gradually allow both commercial and development banks to lend short and long term. The quality of bank portfolios and capital bases are not sufficiently known at this stage. However, the issues relating to provisioning for bad loans and capital adequacy are being addressed by the Central Bank in the proposed reform of the prudential and supervisory framework. 23. A further reform would concern loans derived from funds in foreign currencies. Clearly, a market solution to the foreign exchange risk will be feasible only with the elimination of capital controls and the convertibility of the currency. However, in the meantime, the distortions involved in the administrative management of the foreign exchange risk should be minimized. Thus, financial intermediaries should fully transfer the foreign exchange risk to the final borrowers willing to carry it. This implies the further development - x - of hedging instruments and of "natural hedging" practices. To this end, it might be desirable to increase the proportion of foreign exchange receipts that exporters are allowed to retain above its current 20% limit. As to the medium- and long-term loans, a centralized scheme should be adopted. To ensure the efficiency of such a scheme, the rate charged by banks to the final borrowers : should be linked to, and vary with, a reference rate that reflects the foreign interest rate and the expected evolution of the currency. In Tunisia, the best approximation to such a rate is the TMM. It can be influenced by the authorities through their operations in the money market; it is highly visible and outside the control of any single financial institution. Hence a mechanism that could be considered, for the time being, would be for the borrowing banks to pay to the Central Bank the difference between the interest rate on the borrowing abroad and the money market rate, possibly adjusted periodically (for example, every six months) on the basis of the weighted average prevailing during the previous semester. 24. Selective credit policies and financial incentives have been shown to lack focus and monitoring, to provide a bias in favor of debt financing, and to entail a high budgetary cost. They should be reviewed in the context of the reform of fiscal incentives as they respond to the same preoccupation by the Government of encouraging priority sectors. Both the channels and the instruments through which financial incentives are delivered need to be re- examined. To increase the transparency of the budgetary cost of the incentive mechanism, all subsidies should be paid directly through budget resources. Thus, the preferential rediscounting system at the Central Bank should be phased out completely. The criteria for granting and monitoring direct loans and repayable advances should be reconsidered in view of the past unsatisfactory loan recovery. Finally, to avoid the excessive bias toward debt financing, the self-financing requirements imposed on projects that receive subsidized credit should be significantly increased. 25. The development of an active securities market depends on a stable political and economic environment, and on the existence of an adequate fiscal, regulatory and supervisory framework.5 Efforts in this direction should include: (a) Removing the distortions in the tax treatment of financial instruments in order to encourage the emergence of a market yield curve. Thus, the Law 62-75 should be modified to make the yearly net increase in a securities portfolio deductible from taxable income (up to a specified percentage) irrespective of the origin of the purchase (primary or secondary market), or of the nature of the transaction itself (sale or purchase), or of the identity of the investor. Decreases in portfolio would give rise to pro-rated repayments of tax benefits. This scheme would simplify and remove ./ The promulgation of accounting standards and the development of the auditing profession would also greatly encourage the development of a viable equity market. These important elements have been discussed in the World Bank's "Country Economic Memorandum", op. cit. Volume III, Annex 2. - xi - tax distortions in both the primary and secondary markets and in th( treatment of SICAVs. (b) Deepening the bond market. As far as the Treasury is concerned this would entail: (i) harmonizing the characteristics of the instruments issued; (ii) ensuring adequate market timing of issues; (iii) diversifying the maturities of bonds; (iv) allowing banks to underwrite Treasury and other bond issues; and (v) placing meditmi. and long-term bonds through public offerings as opposed to bilateral placements with institutional investors. In addition, the compulsory holdings of "equipment bonds" by financial intermediaries should be phased out, and the investment guidelines of institutional investors (e.g., insurance companies) should be loosened so as to include all instruments issued by the Treasury. CONCLUSIONS 26. The major challenge faced by Tunisia in the medium term is to achieve a rate of economic growth sufficiently high to reduce unemployment while maintaining macroeconomic stability. The strength of the economy in 1990 is the result of a marked recovery in investment and exports. Part of this recovery can be ascribed to the changes in the economic incentives brought about by the policy reforms implemented in the late eighties. As the economy becomes more sophisticated and distortions are addressed, the rationale for public intervention changes. Government intervention in the economy should cease to be industry and sector specific. Instead, it should encourage the private sector to be flexible and adjust to shocks; to attack remaining market rigidities; and to support the institutions aimed at improving the functioning of these markets. Some recommendations along these lines on how to improve the structure of incentives notably in the trade, fiscal and financial sectors have been provided in this Report. The implementation of the suggested measures would bring about a better allocation of resources. It would also contribute to the achievement of a sustained private sector-based growth, in line with the economic strategy adopted by the authorities in the Eighth Development Plan. I. TRADE POLICIES AND INCENTIVES A. INTRODUCTION 1.01. The trade and indtustrial policies in Tunisia have beeni shepcp' lov variety of considerations. The desire to provide infant-Andil2ty.y typp i- protection to the nascent lndustrial sector, and the need to de%.: se rc irl sources of finance for the Treasury motivated the introductlon of a systeit f high and variable customs duties during the seventies. Tariff protectio5.0, turn, was supplemented by non-tariff barriers which insulated domestic prodtvoevv from fluctuations in international prices. The high level of protection accordci. to domestic industry increased the profitability of impor t-substitvy inq industrialization, inducing a resource shift toward production for tlle dumeA market. At the same time, a system of investmient licensing was introduced ' limit the inefficiencies associated with excess entry into the relativeJy xy domestic market of Tunisia. In turn, the recognition that attempting to coHv-o entry would breed oligopolistic practices prompted Tunisian policy-makers so impose a pervasive system of price controls. Overall, the combination of trede protection, investment licensing and price controls produced a structute as' industry that was inefficient and contained a bias against exports. 1.02. Falling revenues from oil exports in the early eighties brougbi-i the forefront the need for a radical revision of trade, industrial and pA;A ; policies. This need was recognized in the Sixth Development Plan which c&l. le for a major overhaul in the structure of economic incentives, with the objerti.ve of opening the economy to foreign competition and fostering efficiency. '.,>1 implementation of the necessary reforms was fairly slow; indeed, it was onLyv L-rn 1986, when a deteriorating current account brought Tunisia close to a 'hal rnir of payments crisis, that a major reform aimed at providing a more neutral si-t of incentives to exports and domestic sales was initiated. The liberalixatioir program consisted of a combination of freeing of prices, simplification of impm't procedures, reduction in tariffs, and elimination of quantitative restrictions The impact of this policy reform on the structure of trade incentives is anaivFeO in Section B. The response of exports and imports of both consumption ai,C capital goods to variations in trade policy, the exchange rate and exogeno'u' external factors is analyzed in Section C. Finally, Section D reviews the tiS issues involved in the continuatior of the trade liberalization program and dr.s some recommendations. B. TRADE LIBERALIZATION AND THE STRUCTURE OF INCENTIVES 1.03. Prior to the 1986 liberalization program, customs duties .crc.- generally high and subject to wide variations. with rates ranging from r 236X. Typically, consumption goods enjoyed much higher levels of nonE protection than other commodities. The cascading structure of ctstome cwt:W translated into very high levels of effective protection for finai ,n.r 2- According to a study prepared by the Institut d'Economie Quantitative (IEQ)1, in 1985 the rate of effective protection (REP) for activities producing for the domestic market was 84%, as compared to a REP of -7% for export activities. The REP of the textiles and the food processing sectors was 203% and 553%, respectively. Agriculture, mining and transport activities were discriminated against, with a REP equal to 49%, 20% and -7% respectively. 1.04. Overall, the trade protection system generated a strong anti-export bias. However, already in 1972, the recognition of such anti-export bias and the need to redress it prompted Tunisian authorities to grant significant concessions to exporters. With the Law 1972-56, fully exporting firms, i.e., firms which catered only to foreign markets, were granted full exemption from corporate taxation and unrestricted access to imported inputs. Further attempts to promote and diversify exports were made at the beginning of the eighties, with the establishment in 1982 of the Tunisian Export Insurance Agency (COTUNACE), the introduction of simpler customs procedures for exports and the adoption of a law governing the establishment of export companies. The benefits previously granted only to fully exporting firms were then extended (Law 1981-56), in proportion to their export sales, to firms exporting part of their production. Significant changes were introduced in 1985 in the temporary admission procedures, i.e., in the import regime for intermediate inputs for exports. Until then, the importer was still forced to seek an import license and foreign exchange authorization from the Central Bank. The need for an import license was eliminated by the new regulation. In 1986 a deteriorating current account brought Tunisia close to a payment crisis. The short-run response was an intensification of import controls, but, at the same time, a set of far reaching reforms was also prepared. 1.05. The reforms introduced in 1986 brought major changes into the structure of tariff protection. Import tariffs and duties were compressed to a range of 15X-50% in 1987 and 15%-43% in 1990, compared to 5%-236% in the early eighties. Table 1.1. shows that the mean unweighted tariff fell from 32.5% in 19872 to 28.5% in 1990 and that the simplification in the tariff structure led to a reduction in the variation of the customs duties. i~/ See Institut d'Economie Quantitative (1988) and Table 6 in Annex I. The rate of effective protection measures the protection to net value added in an activity and thus the artificial pull of that activity for non- traded resources in production. It is calculated as REP - (PD-1D)/ (Pw-Iw), where P, I - output and inputs; D, W - domestic and world price. tJ The 1987 tariff rates shown in the tables include the customs duty rate and a customs elearance fee (equal to 5% for all tariff positions). This custom fee, together with other minor taxes, was integrated in 1988 into the custom duties. The finance law of 1988 also introduced a new tax, "la redevance des prestations douani&res", equal to 1.5% of the custom duty rate. This tax has been included in the tariff rates. - 3 - TIa o 1.1: RATE OF CUSTOM DUTIE (percmtages) 1987 1989 1990 Whole Products Without Whole Products Without Whole Products Without Economy with QRs QRs Economy with QRs QRs Economy with QRs QRs Mean 32.5 35.0 22.0 30.2 34.9 24.1 28.5 34.5 23.4 Coeff. of variation 44.3 40.9 38.9 37.1 30.4 36.1 41.6 32.2 42.5 Sourco: IEQ tariff files and mission estimates. 1.06. Table 1.2 shows that consumer goods and capital goods were most affected by the reduction in the tariffs, with the average rate of nominal protection falling from 41.4% in 1987 to 36.6% in 1990 for the former and from 26.1% in 1987 to 21.2% in 1990 for the latter. However, the increase in the minimum tariff rate raised the rate of nominal protection for some intermediate goods, offsetting the effect of the reduction in the top rates. As a result, intermediate goods showed a virtually unchanged average level of nominal protection since 1986. The import-weighted tariff of the manufactured sector as a whole3 declined from 26.5% in 1987 to 23.2% in 1990. J/ These values are in line with those prevailing in most developing countries. According to Erzah et al ("The Profile of Protection in Developing Countries" - Unctad, discussion paper no. 21, 1987), at the end of 1985 the weighted average tariff rate for 50 developing countries was 26%. Hovever, similar estimates for OECD countries (Finger and Laird, "Protection in Developed and Developing Countries - An Overview", Journal of World Trade Law 2, 6, December 1987) revealed instead that average tariffs on industrial goods were about 5%. W ~~~~~~~~~- 4 able 1,2: THE SThETURE OF F8TTICI, RAM OF CUSTO; DUTIES (p.rcmtaaea) 1987 1989 1990 Weighted Unweighted As S of Weighted Unweighted As S of Weighted Unwoighted As S of Average Average total Averagoe Average total Average Average total imports imports imports Agriculture 17.6 31.7 7.10 18.8 29.8 9.60 18.5 29.9 6.4 Mining 13.9 18.3 6.86 17.6 19.6 5.70 11.7 18.4 4.5 Manufacturing 26.5 32.9 86.04 26.2 30.5 84.70 23.2 28.6 89.1 - Consumer goods 39.8 41.4 (32.10) 34.4 36.8 (36.01) 32.4 36.6 (32.4) - Interned, goods 19.5 25.3 (43.70) 21.1 25.6 (42.30) 18.8 24.3 (37.2) - Capital goods 24.1 26.1 (24.20) 24.0 24.5 (21.60) 19.7 21.2 (30.4) All comcodities 24.9 32.5 100.00 25.0 30.2 100.00 22.3 28.5 100.00 Source: IEQ Tariff files. Weighting factors are based on 1987-1989 values of imports. 1.07. The reduction in non-tariff barriers, by contrast, proceeded more cautiously. This may be attributed, partly, to the fact that products subject to QRs are not prohibited, but restricted. Imports of these products require an import licence, an annual import authorization or an import card generally liberally granted.4 The percentage of tariff lines corresponding to restricted iv Tunisia operates a fairly complex system of import regulations (published in the "Notice to Importers and Exporters, Official Gazette No. 75, November 27, 1981, amended subsequently). Two broad import categories can be distinguished. The first category consists of products not subject to administrative controls or those whose imports are available to authorized operators, such as exporting firms. Imports of these products are permitted by an import certificate, obtained after presenting a commercial contract, domiciled with a commercial bank and copied to the Central Bank. A simplified procedure (admission temDoraire) may be used by industries that export 15% or more of their output for imports of prohibited products (raw materials and semi-finished products used in their production). This procedure can also be used by enterprises exporting less than 15% of their output for imports not exceeding the value of their exports. All other imports, constituting the second category, require the authorization of the Ministry of Economv. Restricted imports, in turn, fall into three different categories: (i) annual imoort authorization, where import licenses for raw materials and semi-finished goods are issued to industrial enterprises at the beginning of the year for a global limit, expressed in dinars. Since 1984, firms can carry-over to the following year any unused portion of their import authorization. Swapping between items is also permitted. (ii) import cards, mostly for small operators to cover urgent - 5 - goods fell quite rapidly from 80% in 1987 to 48.3% in 1990. However, the inadequacy of this indicator as a monitoring tool of the liberalization process is well-known. First, a rise in the import coverage of non-tariff barriers may reflect a tightening rather than a relaxing of the import regime to the extent that import licenses are less generously granted. Second, and more crucially, the liberalization program involved mostly intermediate and capital goods, i.e., commodities with little or no domestically produced substitutes. Imports competing with domestic production were virtually excluded from the liberalization process. Quantitative restrictions, therefore, continued to offer substantial protection to domestic production. Table 1.3: QUANTItATIVE RESTRICTIONS ON IMFPCTS (Whole econny) Number of tarxff positions Year Total Free Controlled X Controlled 1987 8376 1678 6698.0 80.0 1988 8376 2328 6048.0 72.2 1989 8376 3629 4747.0 56.7 1990 8376 4331 4045.0 48.3 1.08. As can be seen in Table 1.4, non-tariff barriers in 1987 covered 97.0% of total domestic output. In 1988, a large group of capital goods and some consumer goods were liberalized, reducing the percentage of production protected by QRs to 87.5%. A second phase of the adjustment program was initiated, which emphasized the elimination of Qas on import substitutes. Under this program, the percentage of freely importable goods, measured in production terms, was expected to be 75% by the end of 1991 (subsequently revised to end of 1992). The program, however, proceeded much more cautiously: at end-1990 the share of freely importable goods was only 28.1%, as compared to a target of 48.5% at that date.5 1.09. Several reasons explain the delay in the trade reform program. First, the reduction in the tariff rates implemented in 1987-88 in advance of needs amount to a limited sum each year; and (iii) general import licenses which are valid for six months. 5/ Comparisons with other developing countries reveal the importance of protection in Tunisia. For example, in Morocco the percentage of domestic production covered by quantitative restrictions fell from 21.5% in 1989 to 16.7% in 1990; in Indonesia, from 43% in 1986 to 29% in 1989; in Mexico, from 100% in 1984 to about 15% in 1990. Practically all QRs have been eliminated in Turkey and Poland. the removal of quantitative restrictions made this removal more difficult. Second, there was strong opposition from enterprises in the highly protected import substituting industries. Third, a severe drought in 1988 and 1989, combined with a fall in private investment and rising unemployment led the authorities to slow down the process of liberalization for fear of aggravating these trends. This had a perverse effect on protection. The earlier liberalization of intermediate inputs, without the liberalization of the related finished products increased effective protection in many sectors of the economy. For example, the rate of effective protection of the manufacturing industries producing for the domestic market increased from 78% in 1988 to 87% in 1989.6 Despite the Government's commitment to the continuation of the trade reform, industrialists have not yet been presented with a revised medium-term program of import liberalization. Without a credible program, which should include transitional measures to help viable firms to compete internationally, it is doubtful that significant trade liberalization would be achieved in the forseeable future. Table 1.4: PFRCENTAGE OF DaHESTIC PRODUCTION COVERED BY QRs By Sector 1987 1988 1989 1990 Total Manufacturing 96.6 94.0 85.6 68.4 - Food processing & Agro-industries 99.8 99.6 95.6 83.9 - Chemicals 94.4 93.7 90.2 50.9 - Textiles 98.5 97.8 93.6 90.5 - Electrical & Machin. 90.3 88.0 61.3 53.8 Agriculture 97.1 95.5 87.3 77.8 Mining 99.3 98.4 98.3 85.5 Total Economy 97.0 94.7 87.5 71.9 Source: Tariff positions included in the free lists are published in the Journal Officiel de la Republique Tunisienne. Weighting factors are based on 1984 domestic output data provided by the Institut National de la Statistique. hi See: IEQ: "La protect4on effective en 1989" and Annex I, Table 6. - 7 - C. LIBERALIZATION AND EXTERNAL TRADE 1.10. Because of the limited size of its domestic market, it is imperative for Tunisia to rely on foreign trade. By most standards, Tunisia is indeed a fairly open economy. In 1989, the ratio of the sum of imports and exports to GDP, a basic indicator of openness, stood at 93.3, compared with 48.9 for Morocco, and 46.0 for Turkey. In 1970, the same indicator for Tunisia was equal to 48.5. However, the shift toward greater openness of the Tunisian economy has shown considerable fluctuations, with the openness indicator displaying a steady upward trend until 19807, then declining until 1985 and exhibiting a strong recovery afterward. 1.11. The composition of Tunisia's trade, in particular of export flows, has undergone a radical change in the last few years and has mirrored, to some extent, the shift in the orientation of trade policies. Two main trends are noteworthy on the export side. First, because of falling oil prices and dwindling reserves, the share of oil in merchandise exports dropped to 20% in 1989, compared to almost 50% in the early eighties.8 Second, within the non- oil exports subsector, the share of manufactured exports in non-oil merchandise exports increased from 73.4% in 1973 to 90.8% in 1989. By contrast, the share of agricultural exports fell from 17.1% to 8.1% over the same period. 1.12. On the import side, the structural changes were more modest. The dependence of Tunisia on both capital and intermediate goods imports remains high.9 In fact, the share of imported inputs in total merchandise imports exhibited a weak upward trend during 1973-89. Capital goods imports fluctuated in a more pronounced way, mostly in response to domestic investment trends and the availability of foreign exchange. Finally, the share of non-food consumer imports showed a remarkable increase during the last few years. After oscillating around 17% until 1984, imports of other consumer goods appear to take an increasing share, almost 26% of total merchandise imports. Trade Reform and ImDort Behavior 1.13. The experience of other countries demonstrates that import liberalization brings substantial microeconomic benefits to the economy by allowing a more efficient allocation of productive resources and a fuller Z/ See Table 8 in Annex I. t/ See Table 9 in Annex I. j/ See Table 10 in Annex I. - 8 - exploitation of economies of scale.10 The increased pressure of foreign competition helps limiting monopolistic practices and increasing productive efficiency of domestic producers. At the same time, the process of import liberalization may generate some short-term macroeconomic costs insofar as it may free pent-up demand for foreign goods and may lead to a deterioration in the current account. To assess the likelihood of such an outcome, it is essential to predict the evolution of import demand in the wake of the liberalization process, which, in turn, requires an estimate of the price and income elasticities of imports. However, to derive these estimates is a fairly complex task as the observed price and income elasticities are likely to have been distorted by the presence of QRs. In order to circumvent this obstacle, it is necessary to derive empirical estimates of price and income elasticities of imports using a model for which the impact of QRs is explicitly allowed.1" 1.14. Import of consumer goods. Drawing on recent applications of rationing theory to import behavior'2, a model of the import demand for consumer goods, described in Annex II, was estimated for Tunisia. It was found that expenditure levels, relative prices and quantitative restrictions were significant in explaining the evolution of import demand for consumer goods. In particular, the size of the import response to price and total expenditure appeared to be significantly affected by the existence of non-tariff barriers. In fact, it was estimated that a 1% increase in expenditure brings about a 2.7% increase in imports without QRs but only a 1.8% increase when QRs are present.13 Similarly, a 1% decline in domestic prices (or, equivalently, a 1% depreciation LOJ See, for example, Vittorio Corbo, Jaime de Melo and Jim Tybout, "The Effects of Trade Policy on Scale and Technical Efficiency: New Evidence from Chile", PRE Working Paper; and Ann Harrison, "Productivity, Imperfect Competition and Trade Liberalization in Cote d'Ivoire", PRE Working Papers n. 451. lj/ Note that an estimate of the underlying price elasticity of imports is also required, together with an estimate of the price elasticity of exports to evaluate the extent to which real exchange rate devaluation may offset the effect on the current account deficit of a removal of QRs. 1/ *See Giuseppe Bertola and Riccardo Faini, "Import Demand and Non-Tariff Barriers: The Impact of Trade Liberalization. An application to Morocco", Journal of Development Economics, November 1990. JW/ These figures refer to 1989 and therefore reflect the degree of tightness of QRs in that year. of the dinar) determines a decrease in imports of 0.52% and 0.62% respectively with or without QRs.14 1.15. These estimates can be used to simulate the impact of past trade policies, for example, the tightening of quantitative import restrictions in 1986 in response to the incipient balance of payments crisis. Had quantitative restrictions been left unchanged at their 1984 levels, imports of consumer goods would have increased by TD 50 million in 1985 and TD 61 million in 1986 with respect to their historical levels. The current account would have consequently deteriorated by .08% and 1% of GDP in 1985 and 1986 respectively. The impact of future policy options can also be simulated. It can be shown that, following the repeal of all quantitative controls on imported consumer goods, imports would increase on impact by TD 178.2 million, and in the long run, when all dynamic adjustments are completed, by TD 525.4 million. These figures represent respectively 16.7% and 49.3% of actual imports of consumer goods in 1989 or 2.1% and 6.2% of GDP in the short and long run respectively in that year. Thus, the macroeconomic repercussions of import liberalization would not be negligible. However, as discussed later, this could be dealt with by the appropriate management of exchange rate and tariff policies. 1.16. Imorts of caRital goods. While QRs are crucial in explaining the demand for consumer goods, their impact on the demand for capital goods is likely to be minor because a share of these is produced domestically (e.g., construction and some electrical machinery). In addition, a much lower percentage of capital goods, relative to the consumer goods, is subject to quantitative restrictions. Therefore, as detailed in Annex II, the demand for imports of capital goods was analyzed without taking into explicit consideration the effect of QRs. The empirical estimates suggest that capital goods are very sensitive to the behavior of relative prices and to investment demand. In he short run, an increase in domestic investment has a more than proportion.1 impact on the demand for imported capital goods (the so-called accelerator effect), presumably because of short-run supply constraints by domestic producers which are not reflected into higher prices. However, in the long run the elasticity of capital goods with respect to domestic investment is equal to one. The above estimates can be used to interpret the evolution of capital goods imports between 1979 and 1989 (Table 1.5). ]AJ It is worth recalling that in a linear expenditure system, the size of the elasticities depends on the levels of income and prices at which they are being measured. 10 - Table 1I5 DKTEEIREAOFS OIJYDvESTMT WM DVMPFTS Years Import Investment Accelerator Relative Rugression growth growth effects price residual 1979-82 4687 25.0 1.0 4.1 18.5 1982-85 -45.9 -8.8 -17.5 -5.2 -14.3 1985-88 -37.4 -36.4 3.6 -12.4 7.7 1988-89 32.2 8.1 14.2 -0.8 10.8 Notes: The sum of the various determinants contribution plus the regression residual may not add up to actual import growth because of rounding errors. Source: Mission estimates. 1.17. Throughout the period, the main factor affecting the demand for capital goods was the evolution of total investment demand, both in its permanent component and in its transitory component. Indeed, the swings in total investment demand explain from 55% to almost 90% (according to the sub-period considered) of the changes in imports of capital goods. The contribution of relative prices was more limited but not irrelevant, with the exception of the years 1985-87 when the fall in the domestic price of capital goods, relative to that of foreign goods, was responsible for one-third of the overall decline in imports of capital goods. ExRort Growth and Exchange Rate Policies 1.18. Adjustment to external shocks can occur either through import compression or export expansion or a combination of both. Faced with declining oil revenues and stagnant export growth (less than 1.4% on an average annual basis from 1980 to 1986), Tunisia was forced to severely curtail import growth at the beginning of the eighties. In fact, real imports in 1986 were virtually at the level reached six years earlier. Over the same period, GDP growth declined to a mere 2.8%, barely sufficient to compensate for population growth. The period after 1986 witnessed a remarkable turnaround, as indicated by the performance of exports, which in the following three years achieved an average annual growth of 13.7%. The behavior of exports was mirrored in the evolution of both imports and GDP, which grew from 1986 to 1989 at average annual rates of 8.9% and 3.5% respectively.15 I2/ During 1987-89, the GDP growth was negatively affected by a severe drought that curtailed agricultural output. 1.19. Reforms in the trade regime contributed significantly to the promotion of export growth. It was not until the mid-eighties, however, that exchange rate policies played an important role in raising the competitiveness of Tunisian goods. The real effective exchange rate showed a weak appreciating trend until 1984 (Figure 1 and Table 12 in Annex I). Specifically, from 1980 to 1984, the real exchange rate appreciated by 4.3%, as measured by relative wholesale prices (or by 7.5%, as measured by relative GDP deflators). Ejg.ure l The Real Exchange Rates (1980 = 100) 120 60 .... .. .. . . 20Dg 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 year sRelative-WPl based MReIative-CGP basea 1.20. Starting in 1986, the Government adopted a more flexible exchange rate policy. The nominal exchange rate was devalued by 10% in August 1986, which, together with the devaluation in 1985, translated into a real depreciation of almost 11% from 1984 to 1986 (with the GDP deflator-based index showing instead a 24.4% real depreciation). The real exchange rate registered a further real depreciation in 1987 and remained virtually unchanged until 1989. 1.21. The impact of real exchange rate variations on exports depends on many factors. In particular, manufactured exports appear to be much more responsive than primary goods to shifts in relative competitiveness levels. In Tunisia, with its structure of exports increasingly dominated by manufactured goods, export performance since 1985 has been remarkable. Manufactured exports, after recording a meager 6.4% growth rate during 1980-85, grew at an average annual rate of 17.3% from 1985 until 1989. However, the task of identifying the effect of real exchange rate variations is rendered difficult because other 12 - factors, in particular the evolution of international demand for Tunisian exports, also contributed to determining the behavior of exports. 1.22. To disentangle the effects of the real. exchange rate from the impact of fluctuations in world demand, it is useful to examine the changes in Tunisia's market share abroad (Table 1.6). In view of the geographical and commodity composition of Tunisia's exports, the analysis focusses on Tunisia's share in the EEC, the OECD, and in the French and German markets for agriculture, manufacturing and clothing products. It difficult to discern a clear pattern for agricultural exports, whose performance is strongly affected by weather and other exogenous factors. By contrast, the pattern of manufactured exports, particularly of clothing, reveals a close correlation to the behavior of the real exchange rate. In most cases, Tunisia's market share shows a declining trend until 1985, followed by a sustained recovery in 1986, after the exchange rate policy made exporting more profitable. Table 1.6: SHARES OF TINISIAN EXPORTS IN 1MPETS OF MAJoR TRADING PARTNES 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 EEC 10 0.43 0.45 0.37 0.40 0.36 0.35 0.37 0.41 0.38 0.42 Agriculture 0.27 0.35 0.25 0.32 0.34 0.33 0.35 0.43 0.46 0.44 Manufactur. 0.39 0.41 0.45 0.44 0.35 0.34 0.36 0.37 0.37 0.41 Clothing (84) 3.31 3.24 3.63 3.74 3.13 2.99 3.11 2.85 3.34 3.81 FRANCE 0.26 0.37 0.33 0.43 0.36 0.40 0.32 0.30 0.35 0.38 Agriculturo 0.38 0.52 0.46 0.50 0.45 0.40 0.38 0.32 0.30 0.27 Manufactur. 0.27 0.30 0.33 0.34 0.29 0.30 0.32 0.31 0.32 0.38 Clothing 3.50 3.72 3.85 4.01 3.58 3.37 3.55 3.28 3.61 4.34 GERMANY 0.15 0.13 0.13 0.14 0.11 0.11 0.14 0.19 0.14 0.14 Agriculture 0.00 0.00 0.00 0.00 0.01 0.01 0.01 0.03 0.02 0.02 Manufactur. 0.18 0.18 0.19 0.21 0.18 0.14 0.16 0.17 0.18 0.17 Clothing 1.57 1.58 1.76 1.95 1.65 1.38 1.33 1.31 1.52 1.67 OECD 0.14 0.16 0.14 0.13 0.11 0.10 0.09 0.10 0.F9 0.10 Agriculture 0.09 0.11 0.08 0.09 0.09 0.09 0.09 0.12 0.12 0.11 Manufactur. 0.09 0.08 0.09 0.09 0.07 0.06 0.07 0.08 0.0w 0.93 Clothing 0.96 0.92 0.97 0.95 0.72 0.66 0.75 0.75 0.85 0.91 Source: Comtrade data. 1.23. The pattern of Tunisia's penetration in foreign markets can be compared to that of other developing countries. It is often argued that the competitiveness of Tunisian goods with respect to other developing countries represents an important factor in the country's export performance. This effect, however, is not well captured in traditional measures of the real exchange rate which tend to give a disproportionate weight to industrial countries. Turkey - 13 - and Morocco (and, presumably to a lesser extent, Greece) are traditional competitors for Tunisia. Table 1.7: BILATERAL REAL EXCHAMG RATES (relative wholesale prices) Year Turkey Morocco Greece Average 1979 68.53 85.00 101.53 83.77 1980 88.81 88.21 101.29 92.54 1981 87.71 90.97 96.34 91.56 1982 100.13 92.36 97.95 96.79 1983 99.72 94.27 99.99 97.97 1984 100.91 96.18 98.70 98.60 1985 100.00 100.00 100.00 100.00 1986 110.83 92.87 95.55 99.56 1987 104.89 83.01 83.26 89.97 1988 106.29 80.39 82.04 88.99 1989 90.76 77.66 80.37 82.82 (a) An increase indicates an appreciation of the TD. Source: Mission calculations. 1.24. As indicated in Tablc 1 7, the competitiveness of Tunisian goods relative to Turkey and Morocco fe2L quite markedly from 1980 to 1985. Both these countries undertook significaEnt adjustment measures, including a pronounced real depreciation, much earlier thln Tunisia. The falling competitiveness with respect to Turkey and Morocco provice3s a plausible explanation for the loss in market shares that Tunisia suffered dtring 1980-85. The latter could hardly be attributed exclusively to the behavic of competitiveness relative to industrial countries which, as indicated by the traditional measure of the real exchange rate (Figure 1), registered only modesW changes during those years. Finally, since at least 1986, Tunisia has .aintained and somewhat improved its competitiveness with respect also to its competitors in the developing countries, particularly Morocco and Greece. This is reflected in the recovery of its market shares abroad. The paramount role of real exchange considerations is confirmed by examining the evolution of market shares of Turkey and Morocco. Both countries re4iitered significant gains in their market penetration abroad, with Turkey showing both a larger depreciation and significant gains.16 Export Behavior 1.25. Econometric analysis was used to disentangle the effects on export performance of world demand and real exchange rate variations (with respect to competitors in both developing and developed countries). The empirical analysis i/ See "Morocco 2000. An open and competitive economy", Trade Expansion Program, UNDP - World Bank, 1990. - 14 - focussed on manufactured exports It was noted that the evolution of international demand has exerted an important effect on the behavior of Tunisia's manufactured exports; on average a 1% increase in international demand brought about, ceteris paribus, a 6.1% increase in exports. Relative prices have also played an important role in affecting export performances. Exports appear to have been very sensitive to the level of competitiveness with respect to other developing countries (the corresponding price elasticity was estimated to be equal to 1.71). More limited, and indeed statistically insignificant was the effect of price competitiveness with respect to industrial countries. Overall, these estimates support the idea that the real exchange rate had a significant influence on manufactured exports. They also infer that exports compete quite fiercely with products from other developing countries. Therefore, in monitoring the evolution of the real exchange rate, particular attention should be devoted to the behavior of competitiveness in developing countries. 1.26, The contribution of the different demand factors to the growth of manufactured exports is shown in Table 1.8. Table 1.8: DETElhANS OF MANUFACTURED EXPORTS (c-mlative growt rates) Years Export World Relative Regression growth demand price residual 1980-82 13.2 15.4 -7.7 5.5 1982-8' 5.4 11.0 -3.2 -2.4 1984-86 23.1 25.7 -1.6 0.9 1986-89 63.1 49.5 31.5 -18.0 Notes: The sum of the various determinants contribution plus the regression residual may not add up to actual export growth because of rounding errors. Source: Mission calculations. The first column reports the actual rate of growth of manufactured exports over the relevant period. The next two columns indicate by how much foreign demand growth and changes in competitiveness with respect to other developing countries contributed to export growth. The last column shows the regression residual. The results underscore the prominent role of competitiveness effects. From 1980 to 1984, declining competitiveness had a stifling effect on export growth. If competitiveness with other developing countries had remained unchanged, the growth rate of exports would have been 7.7% higher between 1980 and 1982 and 3.2% higher in the following two years. The loss of competitiveness also explains why, after 1984, Tunisian exports were only partly able to benefit from the strong recovery in world demand. Conversely, since 1986, more than 50% of export growth is explained bv the behavior of the real exchange rate, highlighting the crucial role that more flexible exchange rate policies played in promoting export growth. - 15 - Imi2licAtions of AccelerAtingt the Trade Reform 1.27. To evaluate the macroeconomic repercussions ot hmpoy liberalization, a small econometric model of Tunisia's external trade sec:tor wa!: constructed. The model consisted of three behavioral equations which describeJ the determinants of exports of manufactures and imports of capital and consume-- goods, and one trade balance identity. As a first approximation, it is assumed that all other items in the current account do not respond to price variations. This assumption, albeit unrealistic, allows one to put an upper limit on the need for a compensatory depreciation in the wake of a further liberalization of imports. Clearly, if other items, such as tourist receipts or intermediate imports, also respond to changes in relative prices, the required real depreciation will be smaller. In all the simulations, exogenous variables in the import equations, such as domestic investment and private consumption, were assumed to grow at an average annual rate of 5% until 1992. International demand was posited to grow at an annual rate of 1%. 1.28. The impact of the liberalization of consumer goods imports Wa8S examined earlier. Under the assumption of a quasi complete repeal bf quantitative restrictions in 1990, with only 5% of tariff lines still subject to QRs, the model predicts that imports would have increased by TD 221.3 million.17 After two -ears, and with unchanged trade policies, imports would grow by TD 497.3 million with respect to the base scenario. Compensatory exchange rate policies could be relied upon to achieve the objective of an unchanged resource balance.18 Indicatively, if QRs had been repealed, the real exchange rate would have needed to be depreciated by 7.2% in 1990 and by a further 4X in the following year to keep the current account from deteriorating. Import tariffs could also be increased to offset the current account impact of import liberalization. If the tariff rate on imports of both consumer and investment goods were increased by 10%, the required real depreciation (for current account purposes) would have declined to 3.2% in 1990 and to 3.8% in 1991. Overall, whereas the current account impact of trade liberalization is likely to be substantial, strong responses of both import and export flows to changes in relative prices imply that only relatively modest adjustments in the exchange rate would be required to avoid a deterioration in the current account. D. CONCLUSIONS 1.29. Trade policies have undergone a significant shift during the eighties. The process has accelerated since 1986. The objectives of the new 1j1 See Annex I, Table 13. I/ In a more complete model the real exchange rate should be an endogenous variable. The following results, therefore, shou:.d be considered as mostly indicative. They, nonetheless, sugg-st some order of magnitude for compensatory policies. - 16 - policy reforms have been to reduce the export bias and enhance the efficiency of firms catering to the domestic market. A crucial role in this context has been the progressive opening up of domestic markets to foreign imports. The recent performance of the Tunisian economy, in particular the substantial export response to changing incentives, underscores its long-run growth potential and indicates the need to persevere on the road ti structural reforms. ).. 30. A snapshot of Tunisia's trade policy today reveals two main features: on the one hand, domestic producers still enjoy a significant level of protection. The rate of nominal protection remains high and its dispersion across sectors is considerable. Similarly, despite the import liberalization measures adopted by the authorities since 1986, more than 70% of domestic production continues to enjoy quantitative protection from foreign competition. 1.31. On the other hand, exporters have been insulated from the negative impact of policies aimed at protecting domestic markets by receiving generous fiscal and financial incentives and free access to imported inputs. These policies have generated a deep-rooted dualism within Tunisia's industrial sector between firms competing in foreign markets, forced to achieve adequate levels of efficiency, and firms catering to the domestic market which are still substantially shielded from the pressure of international competition. The lack of integration between exporting and other firms is particularly revealing. According to a recent study'9, more than 90% of the inputs purchased by exporting firms are from sources outside Tunisia, despite the existence of several provisions giving incentives to indirect exporters (i.e., domestic firms supplying exporting firms). For example, indirect exporters benefit, on a pro-rata basis, both from a corporate tax exemption and a duty drawback procedure on inputs purchased abroad. Under these circumstances, the lack of integration between exporting and other firms must be predicated on the sizeable level of protection which is still granted to firms producing for the domestic market. By increasing the profitability of domestic sales, trade policy limits the incentive for domestic producers to supply exporting firms. Furthermore, lasting protection is bound to have an adverse impact on the quality of domestic inputs, therefore encouraging exporting firms to locate their input sources abroad. 1.32. In light of the delays experienced during the last two years, it is crucial that the Government accelerate the import liberalization program. This would improve the allocation of resources in production: enhance the welfare of consumers, as domestic prices on consumer goods become closer to international prices; eliminate the indirect costs associated with the management of import controls and the special exemption schemes allowed to exporters; and accelerate the process of structural integration within the industrial sector. It would also allow, in a medium-term perspective when the anti-export bias is eliminated, a reduction in the size of the fiscal incentives granted to exporters. This change would have a beneficial effect on the Treasury accounts and, consequently, also on macroeconomic stability. The experience of other countries, like Turkey U/ See World Bank (1990) "Tunisia - Industrial sector note", February 1990. - 17 - and Morocco, suggests indeed that the range and size of fiscal exemptions can be restricted after the new trade regime has been consolidated. 1.33. The elimination of quantitative restrictions would entail some costs. First, as suggested by our econometric estimates, a deterioration of the current account should be expected. However, the high sensitivity of both imports and exports to changes in relative prices also indicates that a careful management of the exchange rate should be sufficient to avoid large negative effects on trade flows. If, in addition, import tariffs on the liberalized goods were raised, the total effect of the elimination of the QRs on the current account could be virtually offset. Second, adjustment costs, especially in terms oi unemployment are likely to be felt by domestic firms. Whereas export growth would eventually provide a substantial contribution to employment absorption, the need remains, in the short-run, to devise accompanying measures to cushion domestic agents from the costs of adjusting to the new trade regime. These measures should be of a temporary nature, to allow the elimination of quantitative restrictions, and be designed with a view to enhancing the long-run credibility of the reform process. A popular method of liberalizing quantitative restrictions2 consists in converting the licensing system to a tariff system that initially is equally restrictive and then gradually lowering the tariff rate to allow an expansion of imports. Practically this would entail calculating the difference between domestic and world prices before the removal of quantitative restrictions. Temporary tariff surcharges could then be imposed on products for which domestic prices exceed international prices by more than the amount of the existing tariff. The surcharges could subsequently be eliminated according to a well-prepared and publicized time schedule. Following the elimination of quantitative restrictions, the Government should continue the tariff reform, by making tariff rates lower and less dispersed. By reducing distorsions in relative prices, a lower and more uniform nominal protection will allow the whole economy to fully benefit from the effects of a more open and competitive environment. 1.34. In the near future, trade policy in Tunisia should focus on import liberalization. This would be in contrast to the past tendency of increasing the range and the size of incentives to exporters to counterbalance the protection granted to producers for the domestic market. However, some measures could be taken to improve administrative and customs procedures for exports. In particular, it would be advisable: (a) To eliminate the obligation by partly exporting firms to post a bond as a guarantee for unpaid customs duties on imported inputs for their export activities. The required bond is now equal to 5X of the value of imported inputs. It represents a non-negligible burden on the financial operations of exporting firms. ZQ/ See Takacs W. (1990) "Options for Dismantling Trade Restrictions in Developing Countries", The World Bank Research Observer, Vol. 5, No. 1. - 18 - the value of imported inputs. It represents a non-negligible burden on the financial operations of exporting firms. (b) To extend temporary admission procedures to indirect exporters. Present regulations allow exporting firms to rely on a duty drawback procedure, where a refund is claimed for the customs duties paid by indirect exporters. Complex documentation is required to support this claim. This cumbersome procedure discourages exporting firms from relying on domestic firms as a source of intermediate inputs. (c) To encourage fully exporting firms, established under Law 72-56, to sell part of their output on the domestic market. This measure would introduce some competition between goods produced for the domestic market and those for exports, typically of higher quality. Thus, it would promote the integration within the industrial sector. - 1.9 - II. FISCAL INCENTIVES FOR INVESTMENTS AND EXPORTS A. INTRODUCTION 2.01. The guiding principle of Tunisian industrial policy since Independence has been the promotion of economic growth and private initiative. The first legislative Act granting tax incentives to emerging industries dates back to the "Lettre d'Etablissement" of 1946. The introduction of the Industrial Investment Code in 1969, was followed by the promulgation of several other codes and promotional laws. Government support of economic activities was justified on both economic and political grounds to: promote export firms; compete better in international markets; achieve self-sufficiency in certain sectors, notably agriculture; achieve regional balance and assist employment growth; and attract foreign investors by offering at least the same incentives as those of competiting countries. Fiscal and financial benefits were progressively extended to most activities of the economy, including agriculture, industry, tourism and services. 2.02. The recent reform of the personal and corporate income tax systems, which aimed at introducing greater administrative simplicity, economic neutrality and equity, did not include a revision of the existing incentive legislation. Moreover, by creating a more neutral incentive structurc, the reform questions whether the selective and sectoral interventions provided in the Investment Codes should be maintained. The analysis of this incentive structure is the focus of this chapter. Section B discusses aspects of the recent tax reform, in particular the new corporate tax law, and provides quantitative estimates of the net impact of taxation on marginal decisions to undertake investment. Section C compares the Tunisian basic tax structure (excluding the incentives contained in the Investment Codes) with that of the OECD countries. Section D discusses the provisions of the Investment Codes and assesses if they can improve the incentive to invest in the targeted activities and regions. Finally, Section E summarizes the main findings and gives recommendations. B. RECENT TAX REFORM 2.03. The tax reform program, initiated recently by the Tunisian Government, included the introduction of a value added tax (VAT) in 1988 and the reform of both the personal and the corporate tax systems in 1990.21 The VAT, which replaced three turnover taxes and various excise taxes, introduced an average rate of 17X, a reduced rate on basic consumption goods of 6X, and a higher rate of 29% on luxuries. The unified personal income tax has replaced jV The Tunisian tax system is described in Annex III. - 20 - various existing taxes. It is based on a broader definition of income, most indemnities and exemptions being eliminated, and it has reduced the previous 68% maximum rate and its 18 tax brackets to a top rate of 35% and 6 tax brackets. The new corporate income tax has streamlined the previous 6 rates and a top rate of 44% into two rates: 10% for agricultural, fishing and handicraft sectors, and 35% for all other sectors. 2.04. Interest in tax reform has stemmed not only from a desire for fiscal neutrality and administrative simplicity but also from a growing concern over the distortions in the allocation of resources produced by the previous tax system. Thus, the reform has paid considerable attention to the tax-induced incentives to save and invest. It has aimed at achieving a greater neutrality among different sources of corporate finance, and it has eliminated the double taxation of dividends that previously existed. Corporate income is now taxed once at source, while dividends are tax-free for shareholders. Calculating Tax Wedges 2.05. Several questions arise in considering the system of corporate taxation and fiscal incentives in Tunisia. Does the tax system encourage or discourage investments? How does it compare with the tax and incentive framework of other countries? Are investment incentives successful in allocating aggregate investment among the priority activities and sectors of the economy? The answers to these questions require an analysis of the effects of taxation on the decision to undertake new investment projects. Easily computable and frequently used measures of tax pressure, such as the average tax rate (the ratio of tax paid to profit) are unsatisfactory for this purpose. In fact, average tax rates crucially depend on the history of the company and on the age of the capital stock and, therefore, give a misleading indication of the effects of taxation on new investment decisions. 2.06. Standard investment theory predicts that a company would undertake all investment projects which break even; that is, all projects offering a rate of return which, after payment of taxes, is at least equal to the market real interest rate (the rate that can be earned by potential suppliers of finance on alternative assets, for example Treasury bills). The difference between that rate of return and the after-tax real market rate, namely the "tax wedge", is therefore a measure of the incentive to undertake an additional investment at the margin. The framework adopted here for measuring the tax wedge is based on the commonly used King-Fullerton model, adapted to include the particular features of the Tunisian tax system.3 a/ A formal treatment of the King-Fullerton model is given in Annex IV which also describes the manner in which the various provisions of the investment codes were modelled. The King-Fullerton methodology has the advantage of summarizing and quantifying the essential features of the tax system in a relatively simple manner. Internationally, it is also the most familiar framework for comparing tax systems. It has, however, very important limitations. For example, it considers only a limited number of assets - 21 - 2.07. For purposes of practical calculations, the King-Fullerton model assumes a particular value of the real market interest rate.23 We chose 5%; then, ye calculated the pre-tax rate of return which a firm must earn on an investment project for savers to be paid at least a 5% real market rate. This pre-tax rate is conventionally called the cost of capital. The market interest rate will also represent the post-personal tax return for those domestic savers who are exempt from personal tax, and for foreign savers. These two rates will generally differ for those savers liable to pay personal taxes. Thus, estimates of a total tax wedge have been provided, calculated as the difference between the cost of capital and the post personal tax rate received by savers. 2.08. Estimates of the cost of capital required to earn a 5% return are shown in Table 2.1. Only the basic features of the tax system (a 35% corporate tax rate, and no investment incentives24) and the 7.4% inflation rate prevailing in 1989 have been taken into account. Investment projects in buildings, machinery and inventories were considered as financed by retained earnings, equity and debt. Averages were calculated by applying weights which reflected and finance types. More fundamentally, current investment decisions are assumed to be made on the basis of the current tax system and the current inflation rate, while they depend on future returns, inflation rates and tax parameters. Thus, the King-Fullerton methodology correctly represents business investment decisions only if the current values of inflation and tax parameters are an unbiased indicator of the future values. 2/ The methodology is a partial equilibrium analysis so one rate of return must be taken as fixed. In a general equilibrium framework, adjustments to tax rate differentials would determine all rates of return endogenously. ji/ The basic provisions modelled in Table 2.1 are a corporate tax rate of 35%, and no personal taxation. Stocks are valued under the "First-In/First- Out" (FIFO) system. Depreciation is based on the straight line method. The rates of depreciation are those allowed by the law, 15% on machinery, 5% on buildings, as compared to "real" economic depreciation rates conventionally thought to be 12.25% for machinery and 3.6% for industrial buildings (See Hulten, C.R. and Wykoff, F.C. 1981 "Tne Measurement of Economic Depreciation" in C.R. Hulten (eds), Depreciation, Inflation and the Taxation of Income from Capital, Washington, D.C., Urban Institute). It is assumed that all inputs are locally produced and no investment incentives are given. - 22 - current investment and financing patterns.25 The overall marginal tax rate represents an equi-proportional increase in all investments and savings. The standard deviation (in parenthesis) indicates the degree of dispersion among the returns on different assets and types of finance. Table 2. 1 PRE-TAX RAMS OF RETURI WITH 7.4Z INFLATICX AND ZERO PERSONAL TAX Buildings Machinery Stocks Average* Retained earnings/ new oquity 7.92 7.97 11.68 8.69 Debt 1.82 2.23 5.00 2.57 Average 3.89 4.18 7.26 5.67 (3,35) * Overall unweighted standard deviation of the cost of capital in parenthesis. 2.09. A neutral tax system woula require the rates in Table 2.1 to be equal to 5%, the real interest rate, for all assets and finance types. Neutrality implies that the tax system does not distort the choice and the ranking of investment projects and does not discriminate between sources of finance. The overall average (5.67%) suggests that the Tunisian tax system is fairly neutral and does not discourage investment. The tax wedge, indicating whether taxation creates a disincentive to invest, is only 0.67 percentage points. The tax wedge can be expressed as a percentage of the before-tax rate-of-return, giving a 'marginal effective tax rate" which, in this case, is about 12Z. Thus, the combination of tax parameters, inflation rate, and depreciation rules existing at the end of the 80s has resulted in an effective rate of taxation on the 2./ The weights used for the different assets are: stocks 20%, building 52% and equipment 28%; and for the different finance types, debt 66.1%, new equity 5.6% and retentions 28.3%. The weights for the sources of finance have been derived from flow of funds data (World Bank, Country Economic Memorandum, Report No. 8044-TUN, Volume IV, Annex 3, 1990, pp. 66-67). As far as the weights for the different assets are concerned the Annexe Statistique au Rapport sur le Budget Economique 1990 reports that the ratio of construction investment to investment in equipment is nearly 2:1 and the National Accounts (1988) shows that fixed capital formation accounts for around 80% of total investment. The proportions imply that 20% of investment are in stocks, 52% in buildings, and the remainder in equipment. - 23 - marginal investment of 12X, as compared to a statutory rate of corporate taxation of 35X. 2.10. Overall, however, the high standard deviation indicates that there is a wide divergence between required rates of return on projects in different assets and finance types. For example, investment in buildings is favored over investment in stocks or in machinery4, and investment in stocks financed by retained earnings and equity is discouraged. Debt is favored over other forms of finance because interest is deductible from the corporate tax base. This tax benefit is of particular importance during inflation periods27 when it may drastically reduce the cost of capital financed with debt. Tnflation, however, results in the creation of inventory profits and, therefore, increases the tax burden of firms. This is because in the calculation of the taxable income, raw materials, work in progress, and stocks are evaluated at the original purchase cost. Thus, when prices rise, recorded profits will be higher since the acquisition inventory cost is charged against the current value of sales.28 Finally, inflation has the effect of reducing the real value of depreciation allowances. A neutral depreciation system would allow assets to be depreciated, for tax purposes, at rates as close as possible to the true rates of economic depreciation. However, economic depreciation is very difficult to measure accurately. Thus, most countries, including Tunisia, rely on a fictitious "tax life" of the asset and on mechanical rules of depreciation. The so-called "straight line" method used in Tunisia, whereby equal depreciation allowances are given over a number of years,9 is vulnerable to inflation. An alternative "declining balance" method, where the allowance is larger in the initial year ZAI The estimates of the assets are subject to caution as they depend on the true economic depreciation rates which cannot be known for any particular investment. 7iJ As nominal interest rates rise with in'lation, the tax burden to the firms is decreased. The effect of inflation on the cost of capital can be seen by comparing Table 2.1 with Table 14 in Annex I, calculated assuming a zero inflation rate. t/ Alternatively, some countries allow firms to evaluate stocks at replacement cost, thus eliminating the creation of inventory profits during inflation. However, in order to make up for past inflation, the Tunisian tax law allows enterprises to periodically reevaluate certain qualifying assets. After revaluation, depreciation is computed based on the revalued amount. The revaluation gain is treated as a special component of capital reserve. 2J~ Exceptions are allowed. The new income tax law permits accelerated depreciation for some equipment goods, computers and agricultural assets. Depreciation rates higher than normal are allowed to firms operating on a 16 or 24 hours per day cycle. - 24 and gradually diminishes in subsequent years, may be preferable. It offsets the effects of inflation (when the asset starts its economic life), and it is generally preferred by risk-averse companies in cases where the economic life of the asset may be difficult to ascertain. C. INTERNATIONAL COMPARISONS 2.11. Tables 2.2 and 2.3 provide estimates of the tax wedge, relative to a 5X real interest rate, in Tunisia, Morocco, and selected OECD countries. The figures in Table 2.2 are calculated using the average weights throughout the O1CDt, the average inflation rate of 5.2% prevailing in 1989, and zero personal taxation. In this table, therefore, it is possible to compare the structure of the corporate tax system among the different countries eliminating all non corporate tax effects on the required returns, such as personal taxes and differences in the inflation rate. Table 2.2: TAX HEDGES PELSONU1 TAX RATES, 5.2X INFLATION, AVEAGE WEIGHTS Country Buildings Machinery Stocks Retain. Earn. Equity Debt Average France -0.5 -1.2 2.6 3.4 -1.4 -2.2 0.1 Germany -0.8 -1.3 6.5 9.3 -2.7 -3.8 1.0 Greece -0.6 -0 7 -0.1 2.4 2.4 -3.3 -0.5 Italy 0.2 -0.4 -1.1 4.4 -3.4 -3.4 -0.4 Morocco 0.05 -1.3 3.8 3.9 3.9 -2.6 0.5 Portugal 0.0 -0.6 -0.2 2.7 2.7 -3.2 -0.3 Spain 0.4 0.6 2.8 3.9 3.9 -1.5 1.2 Tunisia 0.05 0.1 2.6 3,4 3.4 -1.6 0.8 Turkey 0.1 -0.3 4.8 5.4 5.4 -2.7 1.3 UK -0.1 -0.6 2.3 3,2 -0.3 -1.7 0.4 USA 0.54 -0.9 -0.1 2.5 2.5 -2.9 -0.2 Table 2.3 introduces the effects of personal taxation. The top rate of personal tax in each country is applied. Moreover, the weights given to each asset and finance type and the inflation rate are those of each country. This Table, therefore, gives the total tax wedge for projects financed by investors that pay 0Q/ Average OECD weights are as follows: machinery, 0.41; buildings, 0.31; stocks, 0.28; debt, 0.52; new equity, 0.10; retained earnings, 0.38. - 25 - the highest rate of personal tax and in circumstances which reflect local investment conditions. Table 2.3: TAX VIEXG TOP RATE OF PERSOAL TAXES, OXOUNTY SPECIPIC INFLATION, COUNTRY SPEIFIC WEIGBTS Country Buildings Machinery Stocks Retain. Earn. Equity Debt Average France 1.8 1.0 4.1 3.3 7.2 0.8 2.4 Germany 1.1 1.0 5.3 2.7 2.7 1.8 2.2 Greece 1 9 2.8 -0.2 -0.5 8.9 2.0 1.7 Italy 1.2 1.3 -0.5 2.3 3.3 -0.8 0.8 Morocco 1,7 1.0 3.9 1.1 6.9 1.9 1.7 Portugal 0.2 0.3 -1.3 2.1 6.9 -3.3 -0.2 Spain 4.0 4.6 7.0 4.3 9.2 4.9 5.1 Tunisia 1.1 1.6 4.3 1.9 1.9 1.9 1.9 Turkey -19.5 -11.7 45.2 35.9 35.0 -30.7 1.7 UK 2.8 2.7 6.0 3.7 4.1 2.5 3.4 USA 2.5 1.4 1.8 3.7 5.9 1.0 1.8 2.12. Taken individually and jointly, the fi¢ ares in the two tables suggest that tax parameters in Tunisia are in line with ;nose of most OECD countries. Indeed, the total tax wedge in Table 2.3 is closer to zero, which indicates that in Tunisia the tax system is more neutral than that in Germany, the UK or France. Table 2.2 shows that in all countries, the deductibility of nominal interest payments from corporate tax reduces the marginal tax rate on investments financed with debt. Thus, in the absence of personal taxes, debt is the most tax efficient source of finance. This is especially so in countries with high inflation rates, such as Turkey (73.3X) and Portugal (12.7X). As nominal interest rates rise with inflation, interest payment deductibility increases. In these circumstances, the tax wedge on investments financed by debt may even become negative indicating that these investments are not taxed but subsidized. Inflation, however, as suggested by the values of the total tax wedge in Turkey, exacerbates the distortions in asset and finance types. Stocks are discriminated against in Tunisia and Germany because they are evaluated, for tax purposes, at acquisition cost. For example, in Germany the tax wedge is 5.3 (see Table 2.3) in spite of an inflation rate of 3.3%. In comparison, the relatively high inflation rate in Greece (13.8X) does not affect the tax wedge because inventories are evaluated at replacement cost (which reduces current taxes by eliminating inventory profits). 2.13. Most countries discriminate among different sources of finance. The tendency for debt to be preferred (see Table 2.2) is reversed in those countries - 26 - (see Table 2.3) where the taxation of dividends and capital gains is lower than the taxation on interest income. Table 2.3 also suggests that when the marginal finance for an investment project is provided by a top rate taxpayer, the tax system in Tunisia is neutral regarding the source of finance. This occurs because interest payments are assumed to be taxed at the personal level with a 35% rate: the same rate at which dividends are taxed at the corporate level. Thus, all returns to the suppliers of capital, whether in the form of dividends, capital gains or interest, are taxed in the same way. D. INVESTMENT INCENTIVES 2.14. The present Tunisian incentive framework includes several Investment Codes.3' They are primarily aimed at encouraging investments in the industrial, tourism, agricultural, fishing, and trading sectors although they also promote "horizontal", non-sectoral objectives, such as export growth and regional balance. The revision in 1987 of the most important Code, the Industrial Investment Code, represented an initial step towards the gradual disengagement of the State in the economy. The Code has abolished the requirement of prior approval from the Investment Promotion Agency (API), previously needed for all investments. Prior approval has been maintained only for those investments seeking incentives, and the procedures for obtaining these incentives have been simplified. During 1988-90, some of the Codes (for example, agriculture and tourism) were revised along the lines of the Industrial Code. In addition, the Code for Service Activities was introduced and the Artisan Code was prepared. The Codes contain numerous provisions which can be broadly classified into two categories: (a) fiscal incentives: partial or permanent exemption from payment of registration duties and income tax; preferential tax treatment of income from exports; partial or total exemption from tax of reinvested profits; suspension or repayment of VAT and customs duties; and (b) financial incentives: interest rate subsidies, direct capital grants, assumption by the State of expenditure on preliminary studies, infrastructure costs and, in some disadvantaged areas, of up to 100% of the investment costs. 2.15. As in the previous sections, we are concerned with the relationship between tax provisions and firms' investment decisions.3 Specifically, the jj./ In this Chapter the five major Investment Codes are considered: the Industrial Code (Law No. a7-51), the Agriculture and Fishery Code (Law No. 88-18), t'lie Tourism Code (Law No. 90-21), and the Code for Service Activities (Law No. 89-100) and for investments made by International Trading Companies (Law No. 88-110). it/ A further question, not investigated in this report, concerns the effect of incentives on the demand and supply of labor. Labor taxation includes the income tax, employers and employees social security contributions and payroll taxes (see Annex III). Preliminary research indicates that for the economy as a whole the marginal effective tax wedge, i.e., the difference between the gross labor costs to the employer and the after tax - 27 - King-Fullerton methodology was applied to investigate whether the incentives of the Codes resulted in a lower cost of capital for the targeted investments. Given the number and complexity of these incentives, only the most important were analyzed: exemptions from payment of income tax and import duties; tax exemption on reinvested profits; and interest rate subsidies. Empirical results are reported in Table 2.4 and in Tables 14-17 of Annex I. Table 2.4 shows the rate of return (or cost of capital) required for a firm to provide a potential saver with a return of 5%. A cost of capital of 5% would indicate perfect neutrality of the tax system. A value below 5% would indicate that the investment is being subsidized, rather than taxed, through the incentive system. 2.16. The following implications can be drawn from the estimates in Table 2.4. The tax system, by definition, is neutral for wholly export-oriented industries under both the Industrial and the Services Codes, because investors are exempt from paying taxes. The most generous Code is the Agriculture and Fishery Code. The combination of a reduced corporate tax rate (10%), tax holidays, interest rate rebates, and grants can lower the cost of capital for all combinations of finance and asset types. Indeed, the "redundancy" of incentives produces a negative cost of capital for marginal investments financed with debt (the same happens for investments in decentralized areas under the Tourism Code). The incentives given in the Services Investment Code are the same as those given under the Industrial Code. 2.17. Debt is highly favored over all other sources of finance. In addition to the tax deductibility of interest payments, debt financed investments enjoy financial incentives (in the form of interest rate subsidies) under both the Agriculture and the Tourism Codes. Financial incentives significantly reduce the cost of using debt finance and increase the variation in the rates of return wage available for consumption to the employee is 50%; this implies that 50% of the cost to the employer of taking on a worker is represented by payments to the Government in the form of taxes. The Investment Codes may reduce this wedge. Exemption from the payroll taxes (granted to wholly- exporting firms) reduces the tax wedge to 0.46. If the investment were located in the decentralized zone, the employer would need to pay no social security contributions and the resultant wedge would be 0.37. The marginal tax wedge is clearly important as a determinant of labor supply. A high wedge will discourage work effort and employment, ceteris paribus. Compared to the results reported by McKee, Visser and Saunders ("Marginal tax rates on the use of labor and capital in OECD countries", 1987), Tunisia does not impose a particularly heavy tax burden on labor. For example, the marginal wedge in Sweden and the Netherlands in 1983 exceeded 70%; and figures in excess of those for Tunisia were recorded in two thirds of the countries covered in the study. - 28 - between different assets.> A further incentive given to debt finance consists of personal taxes being exempt from the interest income received from investments in wholly exporting industries (in the Industrial and Services Codes). This benefit, which is clearly relevant to suppliers of capital facing a positive rate of personal tax, contributes to increasing the bias towards debt financing. 2.18. Exemption from payment of corporate tax and import duties is granted in all codes to exporting firms, pro-rated to the output exported. In some cases (i.e., investments in decentralized areas in the Industrial Code, in agriculture and fishing activities, hotel construction and investments in the Sahara in the Tourism Code), these exemptions are given in the form of a tax holiday for 5 to 10-year periods. Tax holidays are controversial incentives and, as shown in Table 2.4, they may even raise the cost of capital. This is because Tunisian firms are required to depreciate their assets for tax purposes during the same period as the tax holiday. When the holiday is over, firms have few undepreciated assets left, and they must pay taxes on income generated by investments undertaken during that period. For example, investments made by firms in decentralized areas (benefitting from the Industrial Code) enjoy a reduced corporate tax rate of 10% for seven years. Thus, depreciation allowances must be made against the 10% corporate tax rate, instead of the 35% rate the firm would normally face. However, if the assets continue to generate a return after seven years, that return would be taxed at the full rate of 35%. Thus, tax holidays worsen the relative position of assets which can be depreciated, for tax purposes, over a short time. Incentives to Foreign Investors 2.19. Government intervention through tax and financial benefits has often been justified in Tunisia by the need to compensate for and compete with the incentives offered to foreign investors by other countries. In fact, incentives to foreign investors are among the most attractive in the world. According to & recent studyX comparing the policies adopted by 53 developing and developed countries to attract foreign capital, Tunisia emerged as one of the countries which had the most fiscal and financial incentives and assistance programs to encourage direct foreign investment. 2.20. In the manufacturing sector, 100% foreign ownership is allowed and government permission is not required by a foreign company for starting a IV Table 2.3 shows that the tax wedges for projects financed with debt are closer to zero (except for Turkey) than those financed with equity or retained earnings. IV See "International Investment Incentives" in Site Selection, October 1989. - 29 - business.5 According to the sector involved, foreign investors have access to all the benefits provided for by the Investment Codes. In addition to the fiscal and financial incentives discussed in the previous sections, incentives in the Codes include the Government's assumption of social security contributions, government supplied infrastructure for new projects in priority areas, investment grants to cover the start-up cost of a project, and subsidies for feasibility studies. Foreign investors can repatriate 100% of the invested capital in foreign exchange and the proceeds from its liquidation even if the initial foreign exchange investments are exceeded. The transfer of profits and dividends is subject to authorization, which is granted automatically by the Central Bank when the enterprise is at least 65% foreign owned. Moreover, non-resident manufacturing and international trading companies have no obligation to repatriate export profits to Tunisia providing that resident suppliers have been paid in foreign currency or convertible dinars. 35/ Some limitations on foreign ownership exist in the agriculture and fishing sectors. - 30 - hle4 PRZYTAX RATES OF REThRN ITEN Z$ PEREXOWL TAX AND 7.4Z XnPLATIOH Buildings Machinery Stocks Retain. Earn./ Debt Average New equity 1 UST)IAL IVSTME CODE 1n, Totally export. 5.0 5.0 5.0 5.0 5.0 5.0 (0.00) lb. Partly export. 4.38 5.56 6.07 7.08 4.01 5.05 (1.63) Ic. Decentr. zones 3.64 6.77 6.37 7,84 3.65 5.06 (2.34) 2. AGRICULTURE CODE 2a. Basic 0.49 0.10 1.82 3.54 -0.84 0.64 (2.17) 2b. Priority areas 0.63 -0.98 1.64 3.10 -1.01 0.38 (2.22) 3. TRADING COMPANIES INVESTgENT CODE 3&. Non-resident trading companies 4.47 4.54 5,89 6,45 3.92 4.78 (1.37) 4. TOURISM INVESTMENT CODE 4M. Hotel Constr. 2.01 4.87 4.94 8.12 0.98 3.40 (3.59) 4b. Invest, in Sahara 0,78 3.46 4.02 7.13 -0.36 2.18 (3.61) 5. S. VICES INSTM CODE 5n. Exporters 5.0 5.0 5.0 5.0 5.0 5.0 (0) 5b. Partly export. 4.38 5.56 6.07 7,08 4.01 5.05 (1.63) ' Overall unweighted standard deviation of cost of capital in parenthesis. Noto: Incentives considered in the calculation of the required pre-corporate tax rates of return (inflation - 0.074 in all cases) la. Exemption from corporate tax, no import duties. lb. Corporate tax reduced on the basis of 35Z of the firm's output being exported; tax exemption on profits from domestic sales up to 20S of profits on exports; exemption from 351 of import duties. ic. Tax holiday at 10S of the corporate tax rate for 7 years. 2a. Corporate tax at 101, reduced to 0X for 10 years and then 5S for five additional years; exemption from import duties; a SZ soft loan rebate; a grant equal to 201 of the investment cost. 2b. On the assumption that 505 of turnover is exported, the corporate tax rate is 0.05; tax holiday at 0X for 10 years; exemption from import duties; a 5t soft loan rebate; and a 20S grant. 3a. On the assumption that 50S of output is exported, the tax rate is 17.51; exemption from import duties. 4a. Tax holiday at 101 for five years, and a 3S soft loan rebate. 4b. Tax holiday at 01 for 10 years plus a 5X soft loan rebate. 5a. Exomption from corporate tax, no import duties. 5b. Corporata tax reduced on the basis of 355 of the firm's output being exported; tax exemption on profits from domestic oalcn up to 20S of profits on exports; exemption from 35S of import duties. - 31 E. CONCLUSIONS 2.21. The recent reform of direct and i.ndirect taxation has put vthe Tunisian tax system very much in line with those of the OECD countries. Indeed, in some respect, the new system for taxing income from capital is better than in most of these countries. In particular, profits are taxed only at the corporate level, and there is no attempt to tax dividends at the pexsonal lev-il. 2.22. Undermining the simplicity, neutrality and revenue poteixtiAl of 'rhe new direct tax law is a system of exemptions and incentives contained in sciveral Investment Codes. These are designed to promote investments in part:icrnlar activities e.g., exports; areas of the country (especially the Sahara); and sectors (industry, agriculture and fisheries, tourism and services). Two mu-n questions arise in considering the Codes. First, why should such activities and sectors be favored by the tax system? Second, does the present system of incentives actually succeed in reducing the cost of capital for investmenlt in these activities and sectors? 2.23. The first question is at the heart of the current discussion on the objectives of industrial policy. As indicated in this chapter, the current targeting of many of the activities in agriculture, fishing, industry, serv Ies, tourism and trading is difficult to justify. If these activi-ies are inherenitly more useful to the Tunisian economy than others, this would probably be refl^ted in their relative profitability. If so, there is no need for them to recaive special tax concessions. If they are not profitable, the argument for giv..ng tax incentives must rest on the grounds that investment in these activities produces external benefits to the economy which cannot be captured by the compe.ny undertaking the projects. But it is difficult to see why investment in so many activities should contribute such external benefits to the economy. Moreover, sectoral targeting may have been necessary in the past to offset the side effects of distortive trade and fiscal policies, administered prices and underdeveloped capital markets. As these distortions are being addressed, there is less need for continuing industry-specific Government intervention. 2.24. To answer the second question, the effect of incentives on firL;s' decisions to undertake investment have been investigated. Several observations can be made about the effectiveness of the incentives contained in the Investment Codes. (a) The proliferation of incentives in any single Code is difficult to justify. For example, investment in agriculture can currently benefit from a reduction in corporation tax, a tax holiday, exempt: oYa from import duties, and interest rate rebates. The simultaneous use of different instruments results in arbitrary and unintended distortions that could easily be avoided if the authorities used fewer but more focussed incentives. For example, interest ratQ rebates are very effective in reducing the cost of investment capital. However, they benefit only debt-financed projects, which already benefit from the tax deductibility of interest payments. As Will be discussed in Chapter III, the preference of debt over all - 32 - other sources of finance aggravates the incentives for maintaining an "overdraft economy". (b) Some incentives, notably tax holidays, are a poor and expensive means of attracting investments. Firms must write off tax depreciation during the holiday, when the tax rate is low. Such tax depreciation is therefore worth less than normal, which increases the cost of undertaking investments to the company. Clearly, for tax holidays to be truly effective, firms should be allowed to defer depreciation until after the holiday. However, given that the cost of tax holidays in terms of foregone tax revenue is relatively high, this kind of incentive cannot be recommended. 2.25. The shortcomings in the Investment Codes raise the question of which incentive structure would better fit the newly reformed corporate tax structure. Designing a new system entails a choice of objectives and instruments. Government objectives should shift from being industry and sector specific to being "functionally" oriented: for example, incentives to promote technology- intensive investments may be justified in terms of their spill-over effects and in support of industrial transformation. The promotion of infant industries, an important policy goal during the last three decades, may still be a valuable objective although it should be reviewed in the context of the results achieved and of the overall development strategy. Promoting exports should continue to be a priority, at least in the short run. Although this is outside the strict domain of market failures, it is still needed to compensate the existing anti- export bias in the economy until trade restrictions and price controls have been eliminated. Thus, the existing Investment Codes should be replaced by a single .ode which would harmonize the principal incentives and eliminate those which are addressing only sectoral issues. 2.26. The incentives currently included in the Investment Codes should be re-examined in the light of the objectives chosen, and possibly reduced. Exemption from Ra3ment -f import duties on equipment and capital goods is currently granted on a permanent basis to exporting. By reducing their fixed costs, this incentive contributes to making enterprises competitive in international markets. It is possible that enterprises relying heavily on imported inputs would not survive if the exemption was terminated. Tariff exemptions should continue to be granted, but only on a temporary basis. In any case, the exemption should be terminated as soon as such inputs can be produced at competitive prices and with adequate quality standards, by the domestic iniustries. Deduction from taxable incomew of earnings reinvested, currently available under most Codes, is a difficult instrument to administer: in fact, not only has it to be verified that earnings (as distinguished from depreciation allowances or external loans) have been retained, but also that they have been invested in certain sectors or activities. In addition, there is currently no 16/ The advantage of tax deduction of earnings invested in shares and bonds, up to 35% of taxable income, is included in law 62-75. In the various Codes the percentage is increased up to lOOX. - 33 - limitation on the number of activities in which earnings can be invested and deducted for tax purposes (and by way of successive deductions, taxable income in certain cases can be reduced to zero). If this instrument is designed to encourage the expansion of an ongoing business or the establishment of a new one, these objectives can be easily achieved by investment allowances or other tax benefits. As already discussed, tax holidays are not an efficient incentive instrument. Because the subsidy conveyed by a tax holiday is proportional to profits, this instrument is also subject to the risk of providing greater and unnecessary benefits to profitable and already established industries. 2.27. The following are some of the criteria for choosing among different instruments. First, incentives should be offered for a limited number of years. This would minimize the fiscal cost of these benefits, and it would become more difficult to create firms that could survive only because of the concessions. Second, incentives should be easy to award, administer and monitor. The elimination of discretion in the approval process and in the size of benefits would reduce the administrative costs of running the scheme, which potential investors may find more attractive. Third, only the most efficient instruments should be chosen, that is those which can affect in a direct and unique way the activities to be encouraged. For example, if the Government wishes to encourage employment, it should provide a subsidy to labor (for example, a temporary exemption on payroll taxes) rather than a tax holiday to labor-intensive industries. 2.28. Three incentive instruments most closely respond to the criteria of cost effectiveness and efficiency as outlined above: Accelerated depreciation allowances allow firms to write off the value of their capital over a shorter time than normally permitted. They reduce the cost of capital and are clearly attractive to firms that have a high capital stock and/or renew and upgrade their capital investments. They would reduce corporate taxes, however, only for companies with sufficient investment. Unlike tax holidays and reductions in corporate taxes, they would raise revenue from profitable projects.37 Rates could vary across different regions or between exporting and domestic companies. Investment allowances, like accelerated depreciation, benefit firms not in proportion to their profits but to the size of their investments. Investment allowances, usually calculated as a percentage of the value of the eligible assets, are attractive to firms that produce enough income to take advantage of the allowance. The simplicity of the investment allowance makes the calculation of the benefits to the investors and the revenue loss to the State relatively easy and reliable. Finally, direct grants represent an effective instrument to reduce the cost of capital. They can be focussed on any policy objective desired by the authorities and, if properly accounted for in the budget, they represent a transparent and easily monitored instrument of subsidization. 21/ The major disadvantage of accelerated depreciation allowances is that they encourage capital intensive investments (i.e., they distort the choice between capital and labor whenever such a choice is available). - 34 - III. FINANCIAL INCENTIVES TO INVESTMENT AND EXPORTS A. INTRODUCTION 3.01. Aside from the incentives created by trade policies and those embodied in the Investment Codes, the Tunisian authorities have traditionally used financial incentives and controls as a means to influence the allocation of capital among economic sectors. Until recently, the Tunisian financial system was characterized by heavy regulation in the deposit and credit markets and an inadequate development of the bond and equity market. The reforms of 1987-88 have introduced many changes. Some interest rates have been liberalized; the responsibility for credit allocation has been returned, at least partly, to the financial institutions; and a money market and new financial instruments (commercial paper, certificates of deposit, etc.) have been created, which have increased competition and flexibility in the financial system as a whole. Moreover, the tax reform of 1989-90 has eliminated the most flagrant disincentives provided by the previous tax system in holding securities. 3.02. Yet, a multitude of past controls and practices still contributes to the perpetuation of what has been defined as the "overdraft economy."8 This is an economy where investment is financed mainly with debt provided by financial institutions, as opposed to internal funds and shares. Banks are virtually the sole link between lenders and borrowers; external funds are made available at a cost that does not reflect the foreign exchange risk; and selective credit policies are eztensively used to channel credit to priority sectors of the economy, or regions, at subsidized rates of interest. 3.03. The analysis of the characteristics of this "overdraft economy" is shown in this chapter. It starts with an analysis of the sources and uses of funds (Section B) in the economy and for the corporate sector in particular. It then discusses the reasons for the choice of debt as the main source of investment financing: the segmentation of credit markets (Section C), the availability of external funds (Section D), the existence of selective credit policies (Section E), and the inadequacy of the securities markets in channelling funds directly from the savers to the enterprises (Section F). Finally, Section G gives recommendations on how to continue the transition towards an open and market-determined financial system. IV See "Epargne et D6veloppement dans les pays du Maghreb", Institut de Financement du D6veloppement du Maghreb Arabe, 1990; in particular the article by Messrs. Ben Othman and R. Bouvariz "La lib6ralisation financiAre: cas de la Tunisie". - 35 - B. THE FINANCING OF CORPORATE INVESTMENT 3.04. Flow of funds data on the borrowing and lending of the various zsectors of the economy in Tunisia and in some developed and developing countries are reported in Annex I, Tables 4 and 5. Relative to these countries, Tunisia has shown a peculiar pattern in the sectoral distribution of surpluses. First, it was one of the few countries where the Government registered a net surplus during the early eighties. Second, the business sector's deficit, averaging 13X of GNP, was one of the largest among the countries of the sample; moreover, the household sector recorded a small surplus of only 2.5X of GNP (as compared for example with 14X in Portugal, 7.7X in Turkey, and llX in Italy). Thus, the burden of closing the gap between savings and investments of the business sector was borne by foreign financing which, on average, represented 9X of GNP (as compared to 7X in Portugal. 3X in Turkey, and 2X in Italy). 3.05. A breakdown of investment financing sources for the business sector (private as well as public enterprises) is reported in Table 3.1. A striking feature of corporate finance in Tunisia is the low degree of self financing.9 Between 1979 and 1987, internal sources of funds (retained earnings and depreciation allowances) accounted for only 20X of the total sources of finance; new share issues have amounted to 19X of companies' needs for funds. Thus, a great part of financing came from government subsidies (13X), and debt in the form of short-term loans (22X), and medium- and long-term loans (26X).4 ~2/ This is very much in contrast with financing patterns of both developed and developing countries where gross internal funds represent more than 50X of corporate financing. gQ/ Both domestic and foreign loans are included. - 36 - Table 3.x: P1URIC AND LnVATE ENTERPRISES - F1ArZ13 SOURCES (in percent of total funds) 1979 1980 1981 1982 1983 1984 1985 1986 1987 1979-87 Gross internal funds 21.5 9.2 15.7 1.5 16.2 30.9 18.4 32.3 27.7 20.0 Medium-Long term loans 41.2 24.9 27.0 32.5 33.5 24.5 25.2 16.0 14.0 26.0 Short-term loans 14.0 33.3 19.6 27.6 10.3 18.6 29.1 23.2 22.4 22.0 Subsidies 14.7 16.1 17.2 14.3 9.9 8.4 14.4 11.7 15.2 13.0 Equity 8.7 16.6 20.6 24.1 30.2 17.7 12.8 16.8 20.8 19.0 Source: IEQ (1990) - La r6trospective de 1'6conomie tunisienne, 1960-1985. 3.06. What has caused private and public companies to rely so much on debt to finance their investments and so little on internally generated funds? Unfortunately, statistics on the operating profits of the corporate sector, are not available. However, indications are that even when profits were high, Tunisian entrepreneurs preferred purchasing real assets (e.g., houses, land) or even investing in short-term financial assets rather than financing their own investments. This behavior appears to be in contrast with the well-established view41 that companies prefer internal sources of funds even when faced with the economic advantages of debt. 3.07. The preference of debt over all other forms of finance has depended on the easy availability of credit from the banking system and on its low real cost. In turn, these characteristics have been the result, to a great extent, of heavy Government intervention in credit markets. Such an intervention has taken the form of: AV/ See for example S. Myers and Mayluf (1984) and G. Donaldson (1961). Donaldson, G. (1961) "Corporate Debt Capacity", Graduate School of Business Administration, Harvard University Press. Myers, S.C. and Mayluf (1984), "Corporate Financing and Investment Decisions when Firms have Information that Investors do not have", Journal of Financial Economics, 13, pp. 187-221. - 37 - (a) Regulatory controls over the banking system. For example, commercial and development bank operations have been strictly directed towards short-, medium- and long-term maturities respectively, which has resulted in a high segmentation of the credit market.42 Such segmentation has been enhanced by administrative controls on both deposit and lending rates.43 (b) The complete coverage, by the State, of the foreign exchange risk on all external loans raised by financial institutions. This has resulted in easily available external funds at a cost sometimes lower than that for domestic resources. Thus, during the eighties, financial intermediaries were encouraged not only to borrow abroad but also to borrow in currencies carrying a low interest rate (usually associated with expectations of depreciation). (c) The extensive use of selective credit policies to direct banks' resources to priority investment projects that the authorities believed would not be otherwise undertaken. Priority investments were considered those in the agricultural, export, and tourism sectors, in underdeveloped areas of the country and those made by small and medium enterprises. Selective credit instruments have included differential rediscount rates, subsidized loan rates and direct grants from budget resources. 3.08. Finally, the predominance of the monetary sector in financing investments has contributed to the underdevelopment of the securities market, together with other factors. For example in Tunisia, the private corporate sector is dominated by family-owned companies. These companies, as in most developing countries, are often reluctant to broaden their shareholdings to prevent loss of control and to avoid disclosure of financial information, possibly for tax evasion purposes. This situation combined with the tax advantages associated with debt instruments, as compared to equity financing and the lack of investor protection regulation, has inhibited the development of a strong securities market in support of corporate investments. C. REGULATORY CONTROLS ON THE BANKING SYSTEM 3.09. For selective credit policies to work, and financial institutions to be unable to circumvent them, credit markets need to be kept segmented and controls over the banking system kept tight. Segmentation and fragmentation are in fact still the principal characteristics of the financial system in Tunisia, W/ Deposit banks lend up to seven years. A/ Administrative controls, coupled with high inflation rates, have resulted in negative real interest rates during the late seventies and early eighties (around 7X-8% for demand deposits and OX-lX for lending rates). - 38 - although less now than in the past. The financial system comprises a Central Bank (BCT), twelve commercial banks, eight development banks, one specialized savings institution, eight offshore banks, several insurance -nmpanies, and a postal checking system." Half of the commercial banks are owned by the Government which also has a 50% stake in most development banks. Legally, development banks have not been allowed to take deposits or to benefit from Central Bank rediscounting. They have funded their operations through their equity base, domestic bond issues and special government resources including the proceeds of loans raised by the Government in international markets and from official agencies. Development banks have represented, and largely still represent, the main channel through which medium- and long-term credit has been allocated to the economy, according to Government priorities. Starting in 1988, they were allowed to take deposits over one year maturity from clients and to provide short-term lending. Long-term lending, however, still accounts for most of their portfolio (about 85%)45 and represents a significant portion of total credit to the economy (15% at the end of 1989). Development banks can freely set their lending rates, although strong moral suasion is exercised by the Authorities in order to keep the rates of both commercial and development banks in line. 3.10. Commercial banks, on the other hand, have been traditionally confined to their deposit taking and short-term lending function, although they have enjoyed Central Bank refinanc4,g facilities and have had access, like the development banks, to special governme.lt and foreign resources. Until the reforms of 1987-88, they were required to place 20% of their deposits into Treasury bonds (Bons d'Equipement), 5% into bonds issued by housing savings institutions, and 18% into medium- and long-term loans. In 19886, the 18% requirement was replaced by the RAP (Ratio des activites prioritaires): since then, 10% of their sight and time deposits have to be held in medium- and long-term loans to the priority sectors. Except for the fulfillment of this A±/ The most important commercial banks are: the Banque Nationale de l'Agriculture (BNA), the Arab Tunisian Bank, the Banque Franco-Tunisienne, the Banque Internationale Arabe de Tunisie, the Banque de Tunisie, the Cr6dit Foncier et Commercial de Tunisie, the SociAt6 Tunisienne de Banque, the Union Internationale des Banques, and the Banque de l'Habitat. The development banks are: the Banque de Cooperation du Maghreb Arabe (BCMA), the Banque de D6veloppement Economique de Tunisie (BDET), the Banque Nationale de D6veloppement Touristique (BNDT), the Banque de Tunisie et des Emirats d'Investissement (BTEI), the Banque Tuniso-Koweitienne de Developpement (BTKD) and the Socit6 Tuniso-Saoudienne d'Investissement et de D6veloppement (STUSID). The specialized savings institution is the Caisse d'Epargne Nationale Tunisienne (CENT). A2J The remaining 15% is represented by equity participation. ij/ Circulaire no. 89-16, May 17, 1988. - 39 - requirement, commercial banks are not allowed, yet, to offer long-term credit. Thus, short-term credit represents about 65% of their asset portfolio (and 57% of total credit to the economy). Moreover, unlike development banks, commercial banks' lending rates are subject to a ceiling of 300 basis points over the money market rate (TMM). The original aim of the ceiling, established in 1988, was to prevent a general rise in interest rates that could have been engineered by the banks' inexperience with competitive loan pricing. The ceiling was therefore meant to allow banks a learning period for developing price competition. Price competition has, indeed, developed. In fact, it is estimated that 20% of loans are currently granted at the TMM plus 1.5%, another 50% at 2-2.5 percentage points above TMM, and only 30% of total loans are priced at TMM plus 3%. Thus, it appears that the ceiling is less justified; on the contrary, it may hamper the banks' ability to further differentiate their lending rates according to the quality and credit worthiness of borrowers. 3.11. Should commercial banks be allowed to grant long-term loans, they would probably be precluded from entering this market segment by their weak capital structure. Aggregate information47 on the balance sheet structure of commercial and development banks indicates that in 1989 the ratio of assets to equity was 11.4 for commercial banks and 1.8 for development banks. Such discrepancy is likely to perpetuate the segmentation of the credit market, to inhibit competition and to distort the emergence of market determined lending rates. Contrary to commercial banks, whose supply and cost of funds is market determined, development banks rely almost completely on their strong equity position to develop long-term loans.48 Because their cost of funds largely depends on the return on equity expected or agreed upon by their shareholders, their pricing policies may not completely reflect market forces and may therefore result in unfair competition with the commercial banks.49 A/ Data on the capital structure of financial institutions should be treated with caution. The lack of uniformity in audit standards is a clear impediment to the transparency of the financial statements of banks and development banks. it./ At the end of 1989, long-term loans (from own resources) and equity participation of development banks were 100% funded out of their share capital. £@i/ For example, it is striking to observe that, following the recent increase in the money market rates (to 11.8% in the second half of 1990 from 10.3% p.a. at the end of 1989 and 8.6% p.a. at the end of 1988) commercial banks increased their lending rates to a much larger extent than development banks. As a result, maximum commercial banks' short- and medium-term rates now stand at about 15%, while development banks lend long term at lower rates. - 40 - Table 3.2: SI4PLIPIED BALACE SHEETS OF WX4MCIAL & DEVELOET BA9 Decmber 1989 Commercial Banks Development Banks (TD million) EQUITY 529 686 LIABILITIES 5,489 530 Deposits 4,194 39 Bonds 0 82 Special Resources 570 386 - govern. funds 369 12 - external funds 201 374 Rediscount 725 23 ASSETS 6,018 1,216 Loans 5,382 847 Securities 141 173 Current Assets (net) 495 196 ASSETS/EQUITY 11.4 1.8 Source: Central Bank, Statistiques Financi6res, June 1990. 3.12. Fostering competition in the banking system and promoting market pricing policies crucially depends on the capital structure issue being addressed. Such issue is currently being discussed by the Central Bank in the context of a major reform aimed at strengthening prudential regulations, accounting practices and the supervisory framework. 0 A radical measure would be to merge selected commercial and development banks, along the same lines as that for the Banque Nationale Agricole (BNA).61 However, this is obviously a complex matter, both from a political and a technical standpoint, which should be carefully investigated. In the short run, the authorities should ensure that the credit and pricing policies of development banks are consistent with those of other lending institutions and of the market as a whole through their representatives on the banks' boards of directors. 3.13. Credit guidelines. A further source of segmentation in the lending market derives from the current system of credit guidelines applied to commercial banks. This system, however, represents a marked progress over the one in IQ/ The design of the reform is still at a preliminary stage. IJ/ In October 1989 the Banque Nationale de DUveloppement Agricole (BNDA) was merged with the Banque Nationale de la Tunisie (BNT) to create the Banque Nationale Agricole (BNA). - 41 - existence until the end of 1986 whereby all banks needed prior approval from the Central Bank for all their lending operations. The prior approval system was replaced in 1987 by a comprehensive set of guidelines which defined the purpose, the maturity and the amount of the loans banks could grant.52 Compliance with these guidelines is checked by the Central Bank on the basis of detailed loan reporting from commercial banks. 3.14. The actual effect of the guidelines on credit allocation is difficult to assess. The Central Bank has indicated that they were mainly motivated by prudential considerations at the time that the prior approval system was suppressed and that they basically spelt out existing banking practices. Potential distortions, however, may obviously take place if the type of loans defined administratively do not correspond to the borrowers' needs. The guideline system may induce banks to provide, and borrowers to request, loans because they are available in a certain form, irrespective of real needs. Also, lenders may be induced not to perform a genuine risk appraisal as long as they feel that the loan features comply with the Central Bank guidelines. In that sense, the current system may perpetuate, though to a lesser degree, the main shortcoming of the previous one, that is to dilute the banks' responsibility in deciding, screening and monitoring loan projects. Moreover, it appears to be ill-adapted to its stated objective of setting prudential regulations. For example, it does not precltide a bank portfolio from being excessively exposed to certain economic sectors, or poorly matched to resources in terms of maturities. The revision of the credit guidelines system is expected to be a major component of the proposed reform of the prudential and supervisory framework. D. EXTERNAL FINANCIN: 3.15. As shown in Table 3.2, a importanit proportion of the total liabilities of the banking system is represented by special resources which include budget resources (from the Government funds) and the counterpart of long-term loans raised abroad. At end-1989, the sum of the funds originated from foreign credit lines by deposit and developmei.t banks amounted to TD 575 million (about 8% of total liabilities and 30% of development banks' liabilities) and included funds from multilateral creditors 50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Source: IEQ Tariff Files. ANNEX I Page 6 of 21 - 69 - Table 8: THE OPEUNESS OF TUNISIA'S ECONOW 1970 1975 1980 1985 1988 1989 Export/GDP 22.0 31.0 40.2 32.6 42.2 44.6 Import/GDP 26.5 35.8 45.6 38.7 41.8 48.7 Openness ind. 48.5 66.8 85.8 71.3 84.0 93.3 Source: Comtrade Data Base. Table 9: THE STRUCTURE OF EXP!S (in percent) 1970 1973 1976 1979 1982 1985 1988 1989 Fuel 26.4 30.0 42.3 48.6 46.0 41.8 16.1 20.0 Agriculture 14.0 17.1 7.7 7.1 5.9 9.6 9.0 8.1 Manufact. 70.0 73,4 78.9 88,0 90.3 87.2 89.7 90.8 Note: Agricultural and manufacturing exports in percent of non-oil exports. Source: Comtrade Data Base. Table 10: THE STRUCTtUE OF DIHPRTS (in percent) 1970 1973 1976 1979 1982 1985 1988 1989 Food 20.7 16.8 12.3 13.0 10.5 12.1 15.0 13.2 Cons. 16.9 19.6 15.8 17.8 16.9 19.8 25.3 25.7 Intermed. 35.4 29.3 28.8 27.8 31.0 32.5 36.4 33.3 Capital 23.2 28.0 31.9 24.1 30.5 22.0 16.7 19.1 Source: Comtrade Data Base. ANNEX I Page 7 of 21 - 70 - Table 11: DISTRIBUTION OF EMPORTS BY WORT REGIME, 1982-1989 (Values in millions of dinara) 1982 1983 1984 1985 1986 1987 1988 1989 RESTRICTED 1571 1561 1641 1342 1465 1254 1316 1255 imp. lic. 991 1068 1070 837 955 789 989 1052 annu. imp. auth. 580 493 571 505 510 465 327 203 UNRESTRICTED 430 409 454 542 452 730 1331 1805 import cortif. 430 283 331 237 199 508 994 1407 equipment goods --- 126 123 151 31 21 77 100 temp. admission --- --- --- 154 222 201 260 298 TOTAL 2001 1970 2095 1884 1917 1984 2647 3060 Source: Central Bank of Tunisia. Table 12: THE REAL EXCHANGE RATE (1980 = 100) Year Based on rela- Based on rela- tive WPI tive GDP deflators 1980 100.0 100.0 1981 100.0 106.7 1982 108.3 109.7 1983 107.3 106.4 1984 104.3 107.5 1985 103.6 104.8 1986 93.6 81.3 1987 80.4 69.7 1988 78.8 69.7 1989 N.A 71.0 Source: IMF Financial Statistics and mission calculations. ANNEX I Page 8 of 21 - 71 - Table 13: THE CURRENT ACCOUNT IMPACT OF TRADE LIBERALIZATION 1989 1990 1991 1992 A. BASE SCENARIO ER 1.000 1.000 1.000 1.000 TB -20.369 -112.899 -188.501 -238.137 MCON 1065.730 1261.848 1418.643 1553.908 MCAP 793.389 805.251 845.513 887.591 XMANUF 1838.750 1954.200 2075.655 2203.363 B. TRADE LIBERALIZATION ER 1.000 1.000 1.000 1.000 TB -20.369 -334.224 -566.508 -735.423 MCON 1065.730 1483.173 1796.650 2051.195 MCAP 793.389 805.251 845.513 887.591 XXANUV 1838.750 1954.200 2075.655 2203.363 C. TRADE LIBERALIZATION AND COMPENSATORY DEVALUATION ER 1.000 1.072 1.112 1.137 TB -20.404 -112.899 -188.501 -238.137 MCON 1065.742 1392.033 1632.141 1828.205 MCAP 793.395 773.959 795.564 824.657 XMANUF 1838.709 2200.044 2490.111 2745.210 D. TRADE LIBERALIZATION AND TARIFF INCREASE ER 1.000 1.032 1.070 1.093 TB -20.404 -112.899 -188.501 -238.137 MCON 1065.742 1320.686 1552.918 1741.758 MCAP 793.395 790.847 813.492 843.635 XMANUF 1838.709 2062.556 2329.618 2564.685 Legenda: ER: real exchange rate TB: (partial) trade balance MCON: imports of consumption goods MCAP: imports of capital goods XMANUF: exports of manufacturing goods Source: Mission calculations ANNEX I Page 9 of 21 - 72 - Table 14: PRE-TAX RATES OF RETURN WITH INFIATION AND Z0R PERSOMaL TAX Buildings Machinery Stocks Average* Retained earnings/ new equity 6.71 6.39 7.69 6.81 Debt 4.23 4.09 5.00 4.34 Average 5.07 4.87 5.91 5.18 (1,28) * Overall unweighted standard deviation of the cost of capital in parenthesis. Table 15: PRE-TAX RATES OF RETIIRN AND TAX WEDS WITH TOP PERSONAL TAX, Z7Rn AND 7.42 INLFATION Buildings Machinery Stocks Retain. Earn./ Debt Average* New equity Zero inflation/35S Personal Tax 4.23 4.09 5.00 4.34 4.34 4.34 (0.40) Tax wedge 0.98 0.84 1.75 1.09 1.09 1.09 7.4S inflation 352 pers. tax rate 1.82 2.23 5.00 2.57 2.57 2.57 (1.41) Tax wedge 1.16 1.57 4.34 1.91 1.91 1.91 * Overall unweighted standard deviation of the cost of capital in parenthesis. ANNEX I - 73 - Page 10 of 21 Tabl 16: PRE-TAX RDTA O RETlUM WITH TOP PESBXItL TAX AND 7.4Z INlLATIOX Average over assets and finance types Buildings Machinery Stocks Retain. Earn. Debt Average* /New equity 1. INDUSTRIAL INVESTMENT CODE l. Totally export. 0,66 0.66 0.66 0.66 0.66 0.66 (0) lb. Partly export. 2.63 3.76 4.22 1.81 4.01 3.26 (I '.3) lc. Decentr. zones 1.96 5.15 4.83 3.02 3.64 3.44 (1.43) 2. AGRICULTURE CODE 2a. Basic -0.01 0.27 0.62 0.96 -0.20 0.19 (0.61) 2b. Priority areas -0.55 -2.15 0.47 -0.37 -1.01 -0.80 (1. 14) 3. TRADING COMPANIES INVESTMENT CODE 3a. Non resident trading companies 2.77 2.89 4.11 1.41 3.92 3.07 (1.32) 4. TOURISM INVESTMENT CODE 4a. Hotel Constr. 0.26 3.18 3.23 3.03 3.03 1.67 (1.66) 4bc. Invest. in Sahara -0.737 1.93 2.75 2.73 -0.36 0.69 (1.95) 5. SERvICES rNvESTMENT CODE 5a. Exporters 0.66 0.66 0.66 0.66 0.66 0.66 (0) 5b. Partly export. 2.63 3.76 4.22 1.81 4.01 3.26 (1.23) ^ Overall unweighted standard deviation of cost of capital in parenthesis. Note: Incentives considered in the calculation of the required pre-corporate tax rates of return (inflation - 0.074 in all cases). ls. Exemption from corporate tax, no import duties, no taxes on interest income. lb. Corporate tax reduced on the basis of 35X of the firm's output being exported; tax exemption on profits from domestic sales up to 20X of profits on exports; exemption from 35X of import duties. lc. Tax holiday at 10X of the corporate tax rete for 7 yeers. 2as Corporate tax at 10S, reduced to 02 for 10 years and then 5X for five additional years; exemption from import duties; a 52 soft loan rebate; a grant equal to 20X of the investment cost. 2b. On the assumption that 502 of turnover in exported, the corporate tax rate is 0.05; tax holiday at 0X for 10 years; exemption from import duties; a 5S soft loan rebate; and a 20S grant. 3s. On the assumption that 501 of output is exported, the tax rate is 17.52; exemption from import duties. 4e. Tax holiday at 102 for five years, and a 3X soft loan rebate. 4b. Tax holiday at 02 for 10 years plus a 5X soft loan rebate. e. Exemption from corporate tax, no import duties. 5b. Corporate tax reduced on the basis of 352 of the firm's output being exported; tax exemption on prof:.ts from domestic sales up to 201 of profits on exports; exemption from 352 of import duties. ANNEX I 74 Page 11 of 21 Table 17: TAX EIDGES WITH TOP PERS(OL TAX AND 7.41 INFLATION Average over assets and finance types Buildings Machinery Stocks Retain. Earn. Debt Average /New equity 1. INDUSTRIAL INVESTMENT CODE la. Totally export. 0 0 0 0 0 0 lb. Partly export. 1.97 3.10 3 56 1.15 3.35 2.60 Ic. Decentr. zones 1.30 4.49 4.17 2.36 2.99 2.76 2. AGRICULTURE CODE 2a. Basic -1.35 -1.84 -0.05 0.63 -1.50 -1.21 2b. Priority areas -1.21 -2.81 -0.19 1.03 -1.67 -1.46 3. TRADING COMPANIES INVESTMENT CODE 3a. Non resident Trading companies 2.11 2.23 3.45 0.75 3.26 2.41 4. TlorsM INVESTMENT CODE 4a. Hotel Constr. -0.4 2.52 2.57 2.37 0.32 1.01 4b. Invest. in Sahara -1.43 1.27 2.09 2.07 -1.02 -0.03 S. SERVICES INVESTMENT CODE 5a. Exporters 0 0 0 0 0 0 5b. Partly export. 1.97 3.10 3.56 1.15 3.35 2.60 Note: Incentives considered in the calculation of the required pre-corporate tax rates of return (inflation - 0.074 in all cases). la. Exemption from corporate tax, no import duties, no taxes on interest income. lb. Corporate tax reduced on the basis of 35S of the firm's output being exported; tax exemption on profits from domestic sales up to 202 of profits on exports; exemption from 351 of import duties. lc. Tax holiday at 10X of the corporate tax rate for 7 years. 2a. Corporate tax at 101. reduced to 0X for 10 years and then 5S for five additional years: exemption from import duties; a 52 soft loan rebate; a grant equal to 205 of the investment cost. 2b. On the assumption that 502 of turnover is exported, the corporate tax rate is 0.05; tax holiday at 0X for 10 years; exemption from import duties; a 52 soft loan rebate; and a 202 grant. 3a. On the assumption that 502 of output is exported, the tax rate is 17.5X; exemption from import duties. 4a. Tax holiday at 102 for five years, and a 3X soft loan rebate. 4b. Tax holiday at 0S for 10 years plus a 5 soft loan rebate. 5a. Exemption from corporate tax, no import duties. 5b. Corporate tax reduced on the basis of 351 of the firm's output being exported; tax exemption on profits from domestic sales up to 202 of profits on exports; exemption from 352 of import duties. Table 18&A: CCEL.M&TED BALMCK SHUT OF CENTRAL PAW AND DEPOIT RABR - LIABILITIES (End of period - Milliona of TD) 1975 1980 1983 1984 1985 1986 1987 1988 1989 1990 Money + quasi money 663 1441 2429 2712 3087 3262 3711 4416 4909 Money 475 950 1653 1765 1987 2057 2025 2456 2526 2649 - Currency outside banks 162 299 533 573 632 651 704 800 874 1005 - Demand deposits 313 651 1120 1191 1355 1406 1321 1565 1652 1644 of which: households 160 322 367 407 425 438 559 496 enterprises 406 635 664 773 801 785 982 1008 Quasi money 188 491 776 947 1100 1205 1686 1960 2383 2571 of which : Comptes Sp6ciaux d'Epargne 49 178 407 487 576 688 863 1079 1300 1511 M3 (1) 667 1445 2431 2715 3090 3265 3713 4420 5163 5480 kU Special resources (2) 108 186 292 351 405 460 501 542 664 743 l of which: Long-term Foreign Borrowing 49 94 130 163 179 189 191 187 201 Capital 83 159 324 413 487 570 612 448 574 640 Total resources of Central Bank & Development banks 892 1781 3301 3647 4136 4316 4777 5217 6230 6840 (1) M3 - (Money + quasi money) - Housing savings. (2) Government funds make up the difference between Special resources and Long-term foreign borrowing. Source: Central Bank, Statistiques Financieres. O o: Table 18.B: CDSLrA MLhM SMEET OF CEURh PAhR AND DI1MIT B*KS - hSSETS (REd of period - MiMllons of TD) 1975 1980 1983 1984 1985 1986 1987 1988 1989 1990 Net foreign assets 148 191 389 289 201 53 176 532 668 596 Credit to the Government 69 218 354 451 554 628 721 677 722 807 Credit to the economy 674 1372 2558 2907 3381 3635 3878 4008 4840 5335 - Short term 473 953 1708 1940 2305 2492 2574 2796 3201 3365 of which: Rediscount 74 117 255 338 433 453 386 381 401 146 Special resources 8 28 57 72 93 106 102 112 112 - Medium term 110 236 512 562 609 631 694 725 901 881 of which: Rediscount 11 21 188 204 209 221 220 212 337 332 I Special resources 36 39 58 70 80 94 117 103 148 - Long term 58 124 204 224 257 257 298 333 597 329 of which: Rediscount 0 0 21 21 13 10 3 0 2 9 Special resources 50 108 157 177 210 211 217 256 286 - Bonds 33 59 134 182 210 255 312 154 141 154 Total assets 891 1781 3301 3647 4136 4136 4775 5217 6230 6840 Source: Central Bank, Statistiques Financi6res. W D >< '0 39 _- ANNEX 1 - 77 - Page 14 of 21 Table 19: S-4ARCET RESOUbM (F DEPOSIT BAhKS (End of Period; iillcma of TD) 1983 1984 198S 1986 1987 1988 1989 1990 1) Total 834 1000 1137 1256 1273 1334 1608 1440 2) Central Bank Rediscount 542 649 732 796 772 792 944 772 3) Special Resources 292 351 405 460 501 542 664 668 - Government Funds * 136 159 193 214 224 259 366 - Foreign Resources 130 164 179 190 191 187 201 - Counterpart Funds 26 28 33 56 86 96 97 Memorandum 4) Credit to the Economy 2148 2396 2752 2953 3217 3527 4668 5194 5) - 1/4 36.8 41.7 41.3 42.5 39.5 37.8 34.4 27.7 * FOSDA, FOSEP. FOPRODI, FONAPRA, FNAB Source: Central Bank, Statistiques Financi8res. Table 20: BALANCE SEEET OF DEVELOPHENS BARXS (End of period. KiLlions of TD) 1986 1987 1988 1989 1990 - Foreign assets 89 45 45 59 70 - Claims on Government -- 0.6 0.7 9 -- - Claims on Economy 1026 1184 1292 847 983 Credit from ordinary resources 655 743 738 510 597 Cradit from special resourcea 371 441 554 337 387 - Securities portfolio 168 159 182 173 208 - Others 190 390.4 464.3 342 382 Total assets - Total liabilities 1473 1779 1984 1430 1643 - Foreign liabilities 83 73 28 51 57 - Monetary deposits 11 17 13 6 8 - Time and saving deposits 247 256 329 115 186 -Special resources 457 566 699 386 417 - Capital accounts 461 662 710 686 739 - Others 214 205 205 184 236 Source: Central Bank, Statistiques Financieres. - 78 - ANNEX I Page 15 of 21 Table 21: (Zn)IT RED)ED BY THE CENTRALE DES RISQuES BY BRANCH OF EMCIC CTIviTY (End of Period; Killioms of TD) 1980 1983 1984 1985 1986 1987 1988 1989 Shgrt-Term Credit (1) 901 1568 1918 2174 2458 2666 2936 3411 - Agriculture and Fish 44 80 94 113 122 135 154 182 - Industry 455 867 1040 1230 1447 1577 1757 2043 of which manufact. 287 670 824 972 1147 1299 1463 1725 - Services 602 621 784 831 889 954 1025 1186 of which Tourism 51 56 79 85 102 104 104 121 Medium- A Long-Tor Credit (2) 568 1150 1373 1585 1718 1882 2024 2139 - Agriculture and Fish 84 149 161 186 208 232 317 320 - Irdustry 267 549 672 758 787 812 816 817 of which Manufact. 214 447 561 633 661 634 716 745 - Services 217 452 540 641 723 838 891 1002 of which Tourism 56 122 167 210 242 257 275 318 Total Credit (3-1+2) 1469 2718 3291 3759 4176 4548 4960 5550 - Agriculture and Fish 128 229 255 299 330 367 471 502 - Industries 723 1418 1712 1988 2234 2389 2573 2860 of which Manufact. 572 1117 1385 1605 1808 1993 2179 2470 - Services 619 1073 1324 1472 1612 1792 1916 2188 of which Tourism 107 178 246 295 344 361 379 439 Percontage of Short-Term Credit 61.3 57.7 58.2 57.8 58.8 58.6 59.2 61.4 Percentage of Medium- & Long-Term Credit 38.7 42.3 41.8 42.2 41.2 41.4 40.8 38.6 Sourco: Central Bank, Statistiques Financibres. Table 22: FCDS DE PER$CAUIoR DES CEUAUKS (FPC) (millions of TD) FOREIGN EXCHANGE PAYMENTS LIABILITIES LIABILITIES: STOCKS LOSSES MADE BY TEE FPC FLOWS AT END-YEAR 1 2 3- 1-2 1983 13.59 1984 13.72 13.72 27.31 1985 16.07 16.07 43.38 1986 14.36 19.50 -5.14 38.24 1987 21.49 17.50 3.99 42.23 1988 32.78 30.41 2.37 44.60 1989 27.27 23.90 3.37 47.97 1990 30.45 22.81 7.64 55.61 156.14 114.12 42.02 source: MLniLtry of Finance and Mission estimates. ANNEX I Page 16 of 21 - 79 - Table 23: LENDIG CONDITIONS OF GOVERM4ENT FUNDS REPAYABLE ADVANCES LOANS FROM BUDGET INTEREST RATE SUBSIDIES Maturity Grace Interest Maturity Grace Interest Period rate Period Rate (years) (years) (in X) (years) (years) (in S) FONAPRA < TD 10,000 11 1/ 0 10,000 TO 25.000 11 1/ 0 FOPRODI < TD 45,000 12 5 3 7 0 4-6.25 2/ (oxtensions) • TD 150,000 12 5 3 10 3 4-6.25 2/ (new projects) 1O to 500,000 12 5 3 FOSDA 3/ "A" projects up to 25 8-8.5 "B" projects General end-user pays Young/Technic. 12 5 4 the equivalent Technicians in 12 5 4 of 8X-8.5S priority areas "C" projects General 10 3 6 up to 3 percen- Young farmers 12 5 4 tage points Technicians 12 5 4 TOURISM PROJECTS 3X to 4X end-user pays at least money market rate 1/ Grace period of up to the maturity of loans grant6d by banks for the project. 2/ Depending on the geographical location of the project. 3/ "A" projects - isolated investments; "B" projects = integrated investments below TD 120,000; "C" projects - integrated investments below TD 150,000. Source: Information provided by the relevant agencies. ANNEX I page 17 of 21 - 80 - ._j a: AfCC8S CONDITIOA TO _VM TUNDS Project Others Repayble Capital Total Debt Broad Promoter Advances Subsidies Equity Debt 2/ FONAPRA < TD 10,000 4% 36% N.A. 40% 60% 96% 10,000 to 25,000 8% 32% N.A. 40% 60% 92% FOPRODI < TD 150,000 3% 6X 21% N.A. 30% 70% 3/ 91% 150 to 500,000 3% 6% 21% N.A. 30% 70% 91% 500,000-1 Million 6X 10.5% 13.5% N.A. 30% 70% 83.5% FOSDA 1/ "A" projects 10% N.A. 10% 20% 80% 3/ "B" projects General 10% N.A. 16% 26% 74% 4/ Young/Technic. 2% 8% 23% 33% 67% 4/ 75% Technicians in priority areas 1X 4% 28% 33% 67% 4/ 71% "C" projects General 15X 15% 1% 31% 69% 4/ 84% Young farmers 6% 24% 8% 38% 62% 4/ 86% Technicians 6% 9S 23% 38% 62% 4/ 71% TOURISM PROJECTS 4/ 1/ "A" proJects - isolated investments; "B" projects - integrated investments below TD 120,000; "C" projects - integrated investments below TD 500,000. 2/ Including repayable advances in debt. 3/ Preferential loans from budget are available. 4/ Bank loans may benefit interest rate subsidies. Source: Information provided by the relevant agencies. . ANNEX I - 81 - Page 18 of 21 Table 25: FINACIAL MABFZS - SELECTED TALES (in Nillioem of TD) 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Corporate Invest, 641 895 1148 1135 1248 1120 981 913 962 1176 1543 STOCK MARKET Share issues 53 314 156 259 300 137 200 121 134 195 New Companies 31 259 118 146 200 42 15 31 42 123 o/w cash 30 259 116 144 193 41 14 28 39 107 Cap. increases 23 55 38 113 99 96 185 91 92 72 o/w cash 16 35 34 108 65 81 169 84 78 59 Total cash 46 294 150 252 258 122 183 112 117 166 X of Investment 7.2% 32.8% 13.1% 22.2% 20.7% 10.9% 18.7% 12.3% 12.2% 14.11 Market Capitalization 366 450 455 546 577 (main market) (278) (445)* Volume Traded 20.4 18.8 23.9 23.7 33.9 24.3 18.2 29.2 S4.8 98.2 Main market 2.9 5.6 6.2 6.2 20.6 8.1 5 4.9 11.1 29.6 Occasional market17.5 13.2 17.7 17.5 13.3 16.2 13.2 24.3 43.7 68.6 Turnover (main market) 2.2% 1.1% 1.1% 2.0% 5.1% BOND MARKET Gross Bond Issues 113 154 153 180 217 258 286 246 324 308 332 Equipment Bonds 111 154 151 178 207 253 246 237 207 252 260 Other Govt Iss. 0 0 0 0 0 0 32 0 80 0 0 Other Issuers 2 0 2 2 10 5 8 9 37 56 72 Govt in % 98.2% 100.0% 98.7% 98.9% 95.4% 98.1% 97.2% 96.3% 88.6% 81.8% 78.3% Net Bond Issues 72.0 101.8 86.8 98.6 120.3 145.7 147.3 80.6 141.8 96.3 101.3 Equipment Bonds 70 102 85 97 111 142 109 80 34 69 60 Other Govt Iss. 0 0 0 0 0 0 32 -6 74 -22 -22 Other Issuers 2.0 -0.2 1.8 1.6 9.3 3.7 6.3 6.6 33.8 49.3 63.3 Govt in S 97.2% 100.2% 98.0% 98.4% 92.2% 97.4% 95.7% 91.8% 76.2% 48.8% 37.5% Market Capitalizat. 389 491 578 676 796 942 1090 1170 1312 1410 1514 Equipment Bonds 387 489 574 671 782 924 1033 1113 1147 1216 1276 Other Govt las. 0 0 0 0 0 0 32 26 100 80 60 Other Issuers 2 2 4 5 14 18 25 31 65 114 178 Marketable Bds 2 2 4 5 14 18 57 57 165 194 238 in X of capitaliz 0.5% 0.4% 0.6% 0.8% 1.8% 1.9% 5.2% 4.9% 12.6% 13.8% 15.7% * Revised data based on new admission criteria for the permanent market. Source: Stock Exchange and Central Bank. ANNEX I - 82 - Page 19 of 21 Table 26.A: KUIN FEATURES OF THE FIADS ISSUED BY DEVELOEtI4E BAMMS 1955- 1988 YEAR *ISSUER * * VOLUME *INTEREST *TAX *W,TURITY *REPAYMENT *INTEREST * 1/ * * ISSUED * RATE *BENEF r* * *PAYMENT * * *(TDmillion)* * 2/ * * * 1985 *STUSID * A * 0.62 * 7.40X* YES * 5 YEARS *BULLET *CAPITALIZED * * B * 0.51 * 9.60X* NO *10 YEARS *BULLET *CAPITALIZED * *TOT* 1.13* * * * * * * * * * * * * 1985 *BDET * A * 2.72 * 7.75%* YES *10 YEARS *EQUA1 INST. *ANNUAL * * B * 1.44 * 10.25X* NO *10 YEARS *EQUAL INST. *ANNUAL * *TOT* 4.16* * * * * * * * * * * * * 1985 *TOTAL * * 5.29* * * * * 1986 *BDET * A * 5.18 * 8.00X* YES *10 YEARS *EQUAL INST. *ANNUAL * * B * 2.44 * 11.00X* NO *10 YEARS *EQUAL INST. *ANNUAL * *TOT* 7.62 * * * * * 1987 *BDET * A * 5.83 * 8.00X* YES *10 YEARS *EQUAL INST. *ANNUAL * * B * 3.03 * 11.00X* NO *10 YEARS *EQUAL INST. *ANNUAL * *TOT* 8.86* * * * * 1988 *BNDT * A * 0.38 * 8.002* YES * 5 YEARS *BULLET *CAPITALIZED * * A * 6.66 * 8.00S* YES *10 YEARS *EQUAL INST. *ANNUAL * * B * 0.89 * 11.00X* NO * 5 YEARS *BULLET *CAPITALIZED * * B * 6.48 * 11.00S* NO *10 YEARS *EQUAL INST. *ANNUAL * *TOT* 14.41* * * * * * * * ' * * * * * 1988 *BDET * A * 1.04 * 8.00X* YES * 5 YEARS *BULLET *CAPITALIZED * * A * 4.67 * 8.00X* YES *10 YEARS *EQUAL INST. *ANNUAL l * * B * 5.59 * 11.00X* NO * 5 YEARS *BULLET *CAPITALIZED ; * * B * 1.08 * 11.00X* NO *10 YEARS *EQUAL INST. *ANNUAL . * *TOT* 12.38* * * * * i ~ ~* * * ' * * * * * 1988 *STB * A * 3.50 * 10.00X* YES *10 YEARS *BBULLET *ANNUAL * * B * 6.50 * 10.001* NO *10 YEARS *BULLET *ANNUAL * *TOT* 10.00* * * * * * * * * ot * * * 1988 *TOTAL * 36.79* * * ' * 1/ STUSID, BDET, BNDT and BTKD are development banks; BNA and BS are commercial banks; TL is a leasing company. 2/ For "A" bonds, the investment amount is deductible from taxable income (up to 35X of taxable income). 3/ Bondholders can choose repayment in equal installments. ANNEX I - 83 - Page 20 of 21 Table 26.B HA MIN FEATURES OF THE BONDS ISSUED BY DEVELOPMENT BARKS 1989-1990 YEAR *ISSUER * * VOLUME *INTEREST *TAX *MATURITY *REP4YMENT *INTEREST * 1/ * * ISSUED * RATE *BENEFIT* * *PAYMENT * * *(TDmillion)* * 2/ * * * 1989 *BDET * A * N.A. * 8.00o* YES * 5 YEARS *BULLET *CAPITALIZED * * A * N.A. * 8.002* YES *10 YEARS *EQUAL INST. *ANNUAL * * B * N.A. * 11.002* NO * 5 YEARS *BULLET *CAPITALIZED * * B * N.A. * 11.002* NO *10 YEARS *EQUAL INST. *ANNUAL * *TOT* 12.37* * * * * * * * * * * * ft 1989 *BNDT * A * 1.42 * 8.002* YES * 5 YEARS *BULLET *CAPITALIZED * * A * 1.81 * 8.002* YES *10 YEARS *EQUAL INST. *ANNUAL * * B * 4.35 * 11.002* NO * 5 YEARS *BULLET *CAPITALIZED * * B * 4.91 * 11.002* NO *10 YEARS *EQUAL INST. *ANNUAL * t*TOT* 12.49* * * * * * * * * * * * * 1989 *TL * A * 1.09 * 8.00X* YES * 5 YEARS *BULLET *CAPITALIZED * * A * 0.35 * 8.00%* YES *10 YEARS *EQUAL INST. *ANNUAL * '* B * 1.19 * 11.002* NO * 5 YEARS *BUILET *CAPITALIZED * *t B t* 0.36 * 11.00X* NO *10 YEARS *EQUAL INST. -ANN..L * *7(fTOT* 2.99 *f* * * * * ft * *t ft * * * 1989 *BS * A * 2.00 * 8.25X* YES *10 YEARS *EQUAL INST. *ANNUAL *t *t B t* 1.00 * 11.002k NO *10 YEARS *EQUAL INST. *ANNUAL * *TOT* 3.00 * * * ft 1989 *BNA * A * 17.50 * 8.00%* YES *10 YEARS *EQUAL INST. *ANNUAL * * 8* 7.50 * 11.00I* NO *10 YEARS *EQUAL INST. *ANNUAL *TOT* 25.00 * * * * 1989 *TOTAL * * 55.85 * * ** 1990 *BNA * A * 5.00 * 8.00X* YES *10 YEARS *EQUAL INST. *ANNUAL *t *t B t* 20.00 * 11.002* NO *10 YEARS *EQUAL INST. *ANNUAL *TOT* 25.00 ** 1990 *BNDT * A * 3.74 * 8.00X* YES * 5 YEARS *BULLET *CAPITALIZED t * A * ~ *A 8.002* YES *10 YEARS *EQUAL INST. *ANNUAL -t * B t* 9.77 * 11.00X* NO * 5 YEARS *BULLET *CAPITALIZED ; * * B *' * 11.002* NO *10 YEARS *EQUAL INST. *ANNUAL * TOT* 13.51 * * * *t 1990 *BDET * A * 8.95 * 8.00X* YES * 5 YEARS *BULLET *CAPITALIZED * A * * 8.00X* YES *10 YEARS *EQUAL INST. *ANNUAL * B * 4.93 * 11.002* NO * 5 YEARS *BULLET *CAPITALIZED * * B * * 11.00X* NO *10 YEARS *EQUAL INST. *ANNUAL *TOT* 13.88* * ** 1990 *BTKD * A * 5.04 * 8,00X* YES * 5 YEARS *BULLET *CAPITALIZED *t *t A * *A 8.002* YES *10 YEARS *EQUAL INST. *ANNUAL * * B * 2.79 * 11.00X* NO * 5 YEARS *BULLET *CAPITALIZED *t *t B *t * 11.00X* NO *10 YFt'RS *EQUAL INST. *ANNUAL *TOT* 7.83 * * * *t 1990 *BS * A * 2.00 *8.5-10X ** YES - 8 YEARS *EQUAL INST. *ANNUAL *t *t B * 6.00 * 11.00X* NO *10 YEARS *EQUAL INST. *ANNUAL *TOT* 8.00 * * * * 1990 *TL * A * 1.50 * 8.25X* YES * 5 YEARS *BULLET *CAPITALIZED * *A * * 8.25X* YES *10 YEARS *EQUAL INST. *ANNUAL *B * 1.50 * 11.00S* NO * 5 YEARS *BULLET 3/ *CAPITALIZED * *B * * 11.00X* NO *10 YEARS *EQUAL INST. *ANNUAL *TOT* 3.00 * * * ft 1990 *TOTAL * * 71.22 * * ft 1/ STUSID, BDET, BNDT and BTKD are development banks; BNA and BS are commercial banks; TL is a leasing company. 2/ For "A" bonds, the investment amount is deductible from taxable income (up to 352 of taxable income). 3/ Bondholders can choose repayment in equal installments. ANNEX I Page 21 of 21 - 84 - Table 27: PLACEMENT OF SELECTED BOND ISSUES WITH INDIVIDUALS AND CORPORATIONS ISSUE INDIVIDUALS CORPORATIONS VOL-.i (I,. percen:) (millions of TD) DEVELOPMENT BANKS BDET 1983 2.1 33.7 66.3 BDET 1984 3.9 5.1 94.9 BDET 1985 4.1 19.8 80.2 BDET 1986 7.6 8.4 91.6 BDET 1987 8.8 13.0 87.0 BDET 1988 12.5 73.9 26.1 BNDT 1988 14.4 3.0 97.0 BNA 1989 25.0 26.0 74.0 BNDT 1989 12.5 3.0 97.0 BNA 1990 25.0 4.0 96.0 BNDT 1990 13.5 5.0 95.0 BTKD 1990 7.8 14.0 86.0 TOTAL 137.2 16.6 83.4 TREASURY BONDS 1986 32.0 57.0 43.0 1988 80.5 60.0 40.0 Source: Stock Exchange. - 85 - ANNEX II THE MODEL OF THE TRADE SECTOR Econometric analysis has focussed on imports of consumer and investment goods and on exports. The empirical results indicate that the evolution of tariff and non tariff barriers to trade plays a crucial role in affecting import behavior, particularly of consumer goods. Remarkably, however, also imports of capital goods are found to be significantly affected by movements in relative prices. Finally, the competitiveness of exports is found to be a major contributing factor to the performance of manufacturing exports. 1. Imports of Consumer Goods The demand for imports of consumer goods modelled in this section draws on recent applications of rationing theory to import behavior.' It is assumed that consumer preferences for domestic and foreign goods can be described by a linear demand system. In the specification of the model a crucial distinction is introduced between restricted and free imports. The model permits to recover structural parameters of consumers behavior, allowing therefore to isolate the impact of import liberalization. The representative household, after having solved the first stage of its optimization problem, allocates total consumption among domestic and foreign goods. Two import cataegories are distinguished, depending on whether they are subject to quantitative restrictions. Relative prices between restricted and unrestricted imports are assumed to be constant. Consumers' preferences are described by a Linear Expenditure System. As shown by Neary and Roberts (1980),2 the demand function for an unconstrained good in the presence of binding constraints on other goods is: m,p1 - rUP+ E E - [ p-iPyf- PcmO] where T's are the subsistence levels of the different commodities; A's are marginal propensities to spend, m's are levels of consumption. Indices f (and j) refer to unconstrained goods, index c to constrained goods. Aggregating over imported goods, we find that total real imports M can be written as: v/ See Bertola G. and Faini R., "Import demand and non tariff barriers: the impact of trade liberalization. An application to Morocco", Journal of Development Economics 34, 1990. See Peary, J.P. and Roberts, K.W.S., "The Theory of Household Behaviour under Rationing", European Economic Review 13, 1980, pp. 25-42. - 86 - Annex II Page 2 of 5 PM N PCE YfmC) M.)+ [E-PD EYD-PJ(EYf + mE)] where Pm and PD denote the price of imports and domestic goods respectively. The above equation is not suitable for estimation to the extent that data on quota levels for individual goods are unavailable. For the purpose of estimation, it is assumed that: a) E O/O - E Pc/(E Of + E ft) can be approximated by the share of tariff lines subject to QR's (henceforth, q); b) that a = (yf + E m,) is approximately constant. With these assumptions, the estimating equation becomes: pNM = pE-a +p(0 [y-PDY-p1aJ Results of the estimation by non-linear least squares, after allowing for partial adjustment are presented in Table Al. Table Al: IMPORTS OF CONSUMPTION GOODS Coefficient Estimate Standard Error a 211.98 29.5 Y 1471.6 240.1 P .50 .098 d .59 .154 DV: 2.31 R2 - .99 IM - 1.45 The (constrained) average propensity to import in 1989 was equal to .173. The estimate of the (unconstrained) marginal propensity to import is significantly higher, .50. Estimated super-numerary income is positive over the - 87 - Annex II Page 3 of 5 whole sample. The Lagrange multiplier (LM) test is used as a diagnostic tool. It does not indicate any clear sign of misspecification. 2. Imports of Investment Goods Investment goods imports are modelled very simply. Investment decisions are assumed to be separable, i.e., investors first decide the aggregate amount of investment and then determine its allocation across foreign and domestic sources of supply. Only the second-stage decision is analyzed here. Following Nabli (1980)3, it was assumed that quantitative import restrictions did not have a substantial effect on capital goods imports. Preliminary investigations confirmed this supposition. Imports of investment goods (MI) are assumed to depend on aggregate domestic investment (I) and the relative price of foreign capital goods (RP, defined as the ratio of imported to domestic capital goods prices). Starting from a general dynamic specification, with one lag for all explanatory variables and no zero price homogeneity condition imposed, the following parsimonious representation of the data generating process was retained (Table A2). Table A2: IMPORTS OF INVESTMENT GOODS ExDlanatory Variable Coefficient Estimate Standard Error Ln I 2.07 * Ln I. l -1.07 .295 Ln RP -.57 .23 DV: 2.14 R2 - .74 LM - .39 *: Constrained coefficient The hypothesis of a unitary iong-run elasticity of MI with respect to I was tested and imposed on the data. It was not possible to reject the hypothesis of both short-run and long-run zero-price homogeneity. v/ See M. Nabli, La protection effective en 1977, IEQ. - 88 - Page 4 of 5 3. Exports of Manufacturing Goods The demand for manufacturing exports is modelled in a fairly standard fashion. It is assumed to depend on foreign income level and relative prices. Following Faini et al (1991),4 the standard model is however amended to allow for a %,ore general pattern of competition where a distinction is made between developing and developed countries' competitors. As a result, international demand for Tunisia's exports of manufacturing goods is assumed to be a function of the level of world income (YW, proxied by a weighted share of the GDP of Tunisia's main trade partners), the relative price of Tunisia's exports with respect to industrial countries (PN, proxied by a weighted average of the GDP deflators of Tunisia's main trade partners) and the relative price of Tunisia's exports with respect to other developing countries (PS, proxied by a weighted average of the wholesale price index of Tunisia's main developing countries competitors, i.e. Turkey, Morocco and Greece). The equation was estimated by a two-stage least-squares procedure, starting again from a general specification with one lag for all explanatory variables and no zero price homogeneity condition imposed. The instruments included all the relevant appropriately lagged explanatory variables, real value added in domestic manufacturing and the GDP deflator. The last two variables would presumably belong to the exogenous determinants of export supply. A Sargan test was used to determine the adequacy of the set of instruments. The results are presented in Table A3. Table A3: EXPORTS OF MANUFACTURING GOODS Explanatory Variable Coefficient Estimate Standard Error Ln (YW) 6.12 .28 Ln (PS) -1.71 .43 DV: 2.44 R2 . .99 Sargan - 4.77 Note: The Sargan test is distributed as a X2 with 7 degrees of freedom. Faini, R. et al., "The Fallacy of Composition Argument", European Economic Review, forthcoming. - 89 - Annex II Page 5 of 5 4. Simulation of the Trade Sector Model The model comprises the three equations described so far. It also includes a (partial) trade balance equation. The model was simulated starting in 1989. All variables are defined at 1989 prices. The activity variables in the import equations, i.e. investment and private consumption, are assumed to grow at a constant 5% rate each year. World demand is assumed to grow at approximately one percent each year. In the first simulation, quantitative restrictions on imports of consumption goods are assumed to affect only 5X of import commodities. The exchange rate is kept at its base level and the trade balance is allowed to deteriorate. In the second simulation, the trade balance is fixed at its base case level and the real exchange rate is the adjusting variable. In the last simulation, a 10% across-the-board import tariff is introduced. The real exchange rate is still determined endogenously. The GANS program was used for all the simulations. 90 - ANNEX III THE TUNISIAN TAX SYSTEM 1. In recent years the Tunisian tax system has been largely reformed. The process started with the reform of external trade taxes in 1986, followed in 1988 by the replacement of the existing system of indirect taxation with a Value Added Tax (VAT). A new personal income tax was introduced in 1990, and the corporate income tax is expected to be effective in 1991. Only the Investment Codes, designed to encourage export volumes and investments by residents as well as foreigners in particular sectors and areas of the country remain largely unreformed. The aim of this Annex is to describe the most important characteristics of the tax system in Tunisia. The features which are not relevant to the empirical analysis developed in Chapter II have not been included. A. PERSONAL INCOME TAXES Personal Income Tax 2. In December 1989 a unified income tax was introduced in Tunisia (Law number 89-114). It replaced the old system based on a schedular system , with different tax rates according to the source of income. Based on a broader definition of income, since most previously existing indemnities and exemptions were eliminated, the new unified tax is levied on the incomes realized on an individual basis or in the context of a partnership by all Tunisians, residents or not. Eleven different tax rates are applied up to a maximum rate of 35%. Income (Tunisian Dinars) Rate (X) 0- 1500 0 1500- 2000 18 2000- 2600 20 2600- 3200 22 3200- 4000 24 4000- 6000 26 6000-10000 28 10000-20000 30 20000-40000 32 40000-80000 34 more than 80000 35 Dividends distributed are exempt from personal tax. - 91 - AnnexllII Page 2 of 7 Taxation of Labor 3. Payroll taxes. Two taxes are levied on the payrolls of companies: (a) The Business Training Tax is levied on the wage bills of all companies (whether subject to the personal or corporate income tax) at a rate of 1% for industrial companies, 2% for ether companies, but 0% for agricultural activity; and (b) The Contribution to Funds for Workers Lodgings is set at 1% on the wages bill. 4. Social security contributions. As in most countries,the social security system in Tunisia is funded by compulsory levies on both employers and employees. In the non-agricultural sector, employers pay 17.5% of total salaries to the national social security fund, while employees pay 6.25%. In the agricultural sector, employers pay 4.4% and employees pay 2.05%. B. TAXATION OF GOODS AND SERVICES Value Added Tax 5. The value added tax, introduced by Law 88-61 of June 2, 1988, has replaced the existing production tax, the consumption tax and the tax on services. The VAT applies to most economic transactions which are carried out in Tunisia. The standard rate is 17%, with a reduced rate of 6% applying to professional services and to the import, production and sale of fertilizers, electricity, gas, television and other minor products. An increased rate of 29% is levied on luxury goods. Custom Duties 6. (i) Custom duties on imports; a three-column tariff system is used - minimum, preferential, and general rates of duty are levied according to the origin of imports. Preferential treatment is given to 15 North African and Middle Eastern countries. General treatment is &Iven to most other countries. Some exemptions and deductions apply to certain categories of importers. The range of tariff rates is 15%-43%X The minimum tariff rate on equipment goods is 10%. All duties are assessed ad valorem at c.i.f. value. 7. (ii) Custom duties on exports; most exported goods are exempted from custom duties. They apply only to a very small number of goods (fish waste, animal products unfit for human consumption, waste and scrap metal of iron and steel) with a rate of 25% of the export value. - 92 - Annex III Page 3 of 7 C. TAXATION OF CAPITAL Corporation Tax 8. The comprehensive reform of direct taxation enacted by Law 89-114 of December 1990 included a new corporate tax system, which is expected to become effective in 1991. Instead of six rates and a top rate of 44% as in the old system, the new corporation tax has a general tax rate of 35%. This is reduced to 10% for agricultural, fishing and handicraft activities. The tax base excludes interest payments. Assets are depreciated for tax purposes using the straight line method. Commercial buildings can be depreciated at 2-5%, industrial buildings at 5%, and equipment at 15%. Inventories are valued on the first-in-first-out (FIFO) basis. A system of accelerated depreciation is allowed for a limited number of equipment goods. 9. The Law no. 89-114 of December 1990 also attempted to reduce some of the fiscal advantages included in the Codes. In particular, Article no. 12 states that, regardless of the provisions of the Codes, all companies will be subject to a minimum corporate tax at a rate of 10% with the exception of wholly exporting enterprises, of financial intermediaries and of firms under the Agriculture and Fishery Code. Articles 9 and 10 concern the incentives accorded by Law no. 62-75 (and increased in the various Codes) in the form of tax exemption on the part of income or profits reinvested in shares and bonds. Previously, the exemptions varied from 35% to 100% according to the sectors of the enterprises issuing these assets. With the new reform, the percentage of reinvested income that can be deducted from the tax base has been reduced, by a minimum of 30%, except in the case of reinvestments in firms in decentralized areas, in agriculture and in particular deposit accounts (the "comptes d'epargne projets"). Taxation on Gains from Immovable ProRertv 10. This tax is levied on profits from sales or gifts of buildings and land. If the asset is held for one year or less the tax rate is 30% on the gain, 20% if it is sold after being held for 2-3 years, and 10% if it is held for between 4 and 10 years. Tax on Immovable ProRertv 11. The taxation of immovable property differs from the regime of immovable property gains tax. There is a municipal tax on buildings, whose base is the gross rental value and with rates varying from 10% to 15%, depending on the rent control legislation. A supplementary tax applies to buildings subject to the rent control legislation and is payable by any occupant of a building. There is also a land tax on underdeveloped land, at a rate up to 5% of the rental value. - 93 - Page 4 of 7 D. OTHER TAXES 12. There are a range of other less important taxes. A wide range of registration and stamp duties are levied on financial transactions. There is a tax on currency transactions - 0.5% of all sales of foreign currency is taken in tax. A hotel tax (1% of turnover) is levied on tourist establishments, and there is a similar tax on non-tourist activities, at a rate of 0.2%. E. THE INVESTMENT CODES 13. Most economic activities in Tunisia are covered by separate investment incentive schemes, managed by the sector Ministries and administered by the Promotion Agencies: the Industrial Investment Code, the Agricultural and Fishing Investment Code, the Incentives for Resident and Non-Resident International Trading Companies, the Tnvestment Code for Tourism, and the Services Investment node. In addition, there are other promotional 1*s8 (not analyzed in this Annex), providing fiscal incentives to banking and financial institutions, encouraging investment in Government securities, and establishing provisions for the production of Hydrocarbons. The Industrial Investment Code (Law no. 87-51. August 1987T 14. The Industrial Investment Code was completely reformed in 1987. It replaced the provisions in Law no. 81-56 (establishing incentives for manufacturing industries and industrial decentralization) and in Decree-Law no. 85-14 of October 1985 (establishing incentives for investments in exporting industries). In the new Code, the procedures for granting incentives have been highly simplified. Incentives to wholly exporting firms are granted only upon presentation of a statement of investment intention. For all other firms, a request is needed, followed by a decision by the Ministry as recommended by the Industrial Promotion Agency. 15. (i) Manufacturing industries producing fully for export. On deposit of a corporate plan outlining an investment project which is intended to produce mainly (i.e., more than 80%) for export, the firm is entitled to be exempt from corporate taxes, registration duties, customs duties and VAT. Nor is the firm liable to the payroll taxes - the business training contribution, and workers housing contribution. The only compulsory levy is the employers social security contributions. However, if the firm operates in a decentralized area, a 5 year tax holiday is granted. If the project is financed by debt, then the interest received by debt hoLders is exempt from personal tax. Profits or income reinvested in the equity of a wholly exporting firm are 100% tax deductible (if the firm operates in decentralized ateas) or 70% (in non decentralized areas). Wholly exporting firms are also entitled to a refund of customs duties and VAT on equipment goods, spare parts and semi-finished goods imported and/or acquired in the local market. There are also a range of other administrative benefits, such as unconstrained repatriation of profits, no control over recruitment policy, etc. If in addition the company is in an underdeveloped region of the - 94 - Annex III Page 5 of 7 country, further advantages result, including State payment for any infrastructure costs, and the payment of the employers contribution to the social security fund for five years, extensible to ten. 16. (ii) Manufacturing partially for exports. A company exporting only part of its production is entitled to fixed registration fees for 10 years and it i. exempt from corporate tax in proportion to the ratio of exports to total sales. In addition, the company is granted a tax exemption on profits from domestic sales up to a ceiling of 20% of profits on exports. The company also enjoys: a) a 40% tax exemption on annual taxable income or profits reinvested in the company, if it operates in non decentralized areas (70% in non decentralized areas); b) suspension of VAT on capital goods imported or purchased locally and needed for the firm productions activities; c) suspension of turnover taxes on local purchases needed for the exports; refund of custom duties and charges on import of inputs needed for the exports; d) refund of custom duties and charges on import of equipment goods not manufactured locally, in proportion to exports; e) exemption of 40% of the income received from export activities from individual income tax. It also benefits from the temporary admission procedures for imported goods and products intended to be processed for re- exportation. 17. (iii) Manufacturing in decentralized areas. Manufacturing firms in decentralized areas are entitled to a reduced corporate tax rate of 10% for seven years. In addition, if companies can show that they export 20% or more of their annual turnover, this incentive can be extended for another 3 years. The state pays for any infrastructure costs, and the employers social security contributions for five years. The employer is exempt from contributing to the employers housing fund. The Agriculture and Fishing Investment Code (Law ;o. 88-18. Anril 1988) 18. All businesses in this sector, regardless of whether they qualify for the investment code, are entitled to the reduced corporate tax rate of 10%. Investments qualifying for this Code (those intended to increase, develop and modernize agriculture and fishing production, and to promote the interests of farmers and fishermen) must be approved by the Minister of Agriculture on the advice of the Agency for the Promotion of Agricultural Investments. Investments in the agriculture and fishing sectors are classified as : "A", investments made by small and medium-sized farms and fishing concerns in the form of "specific actions:; "B", investments undertaken by small and medium-sized farms or fishing concerns in the form of small and medium-sizid integrated projects; "C", investments undertaken by large companies as specific actions or integrated projects or within the framework of the realization of highly productive projects. General advantages include: (a) the exemption from corporate tax during the first 10 years followed by a reduced rate of 5% for a further five years; (b) reduction of custom duties to the legal minimum leviable on imported capital goods and suspension of VAT payable on the acquisition of capital goods locally manufactured; (c) tax exemption, up to 70% of taxable income of earnings reinvested in the subscription or increase of the company's capital; (d) fixed - 95 - Annex III Page 6 of 7 registration duties on acts setting up a corporation and on those relating to an increase in capital during 10 years from the date of incorporation of the company; (e) financial advanteges. Investments in category "Al' can benefit from tho financial advantages provided in Law no. 63-17 and Law 69-11 in the form of direct subsidies (up to 15% of the projects), loans (up to 75% of the projects) and interest rate subsidies (preferential rediicounting. see Chapt. III). For priority projects, these rates are higher. The minimum self-financing ratio required is 10%, while an additional 10% is provided by the State as a capital subsidy. Investments in category "B" benefit from a subsidy (up to 15% of the investment). The minimum required self-financing ratio is 10% and up to 25% is usually provided by the State as repayable advances. Some particular investments in "B" enjoy the same benefits as those in "A". Category "C" investments are entitled to preferential rediscounting, interest rate subsidies and repayable advances. Investments in priority activities (that is in the fields of food self-sufficiency and exporting activities) benefit from exemption from corporate tax within the timit of the ratio of priority activity turnover to total turnover. Similarly, firms that export agricultural or fishing products are exempt from corporate tax within the limit of the ratio of export turnover to total turnover. Investments made in priority activities benefit, beside the general financial advantages, from an added rebate of the interest rate as investment credit. The Trading Companies Investment Code (Law no. 88-110. August 1988) 19. International trade companies are companies whose main activity is the import and export of goods as well as any other international trade or brokerage operation. 20. (i) Non-resident international trade companies. To be eligible for this code, at least 66% of the capital of the company must have been provided in foreign currency. If this condition is satisfied, the company becomes entitled to the following: exemption from corporation tax; exemption from paying VAT and from import duties on capi.tal goods and inputs; 100% reimbursement of import duties on goods purchased locally; fixed registration duties; reduction in the income tax on the salaries of foreign members of staff. The company is also allowed to the free transfer of profits to non residents abroad and to keep the net proceeds from export in foreign currency. 21. (ii) Resident international trade companies. These companies are entitled to: (a) exemption from corporation tax in proportion to the ratio of exports to overall sales; (b) exemption from VAT on inputs purchased locally for export; and (c) refund of customs duties, and various other administrative benefits. Reinvested earnings are deductible up to 70% of taxable income (irn the case of the initial capital) and 35% (for increases in the capital). The Tourism Investment Code (Law no. 90-21. March 1990) 22. Tourism investments are those in the "tourist zones" in the folloving activities: (a) providing accommodation; (b) recreation; (c) tourist transport. - 96 - Annes III Page 7 of 7 To be entitled to the advantages of this Code, tourism investments must be approved by the Minister in charge of the tourist sector on the advice of the Office National du Tourisme Tunisien. The basic advantages include: (a) fixed rate of registration duty for acts relating to the setting up of an enterprise and to increase in the capital; (b) exemption of earnings reinvested in authorized capital of an enterprise up to 35% of taxable income; (c) suspension of payment of VAT and other taxes and payment of the minimum tariff rate on the imported new capital goods; (d) suspension of payment of VAT for new capital goods produced in Tunisia. 23. (i) Hotel construction for tourist accommodation and recreation. Companies involved in hotel construction get the same benefits as those investing under the basic code. In addition, they pay corporation tax at a rate of 10% for five years (in the case of tourist accommodation) or three years (in the case of recreation). The state also takes charge of some of the necessary infrastructure costs. In addition, up to 60% of the investment costs may be funded by a soft loan. The maximum interest rebate is 3% (4% if the company is in a decentralized area). 24. (ii) Investment in the Sahara. Companies which invest in the Sahara get the same benefits as those investing under the provisions of hotel construction. They are also exempted from payment of corporation tax for 10 years. In addition, up to 70% of the investment cost of the project can be financed by a loan with a 5% interest rebate. The Services Investment Code (Law no. 89-100. November 1989) 25. Service investments are defined as invastments made in computer services, engineering, building and public works, research, counseling and assistance. As in the Industrial Code, wholly export-oriented firms are exempt from corporate taxes, from payment of customs duties and of VAT. Interest payments received from companies are exempt from personal tax. 26. Partially exportilna firms are entitled to paying fixed registration duties. Corporate taxes are reduced in line with the proportion of turnover which is exported. In addition, domestically generated profits of up to 20% of the profits made by exporting services are exempted from tax. Individuals contributing to nev equity of a service industry may reclaim up to 70% of the personal tax paid on the contribution. The company is exempt from customs duties and VAT on capital goods imported or purchased locally. - 9'7 - ANNE.' IV METHODOLOGY OF CONSTRUCTING TAX WEDGES AND EFFECTIVE TAX RATES A. THE KING-FULLERTON METHODOLOGY 1. The marginal effective tax rate is a measure of the tax due on an additional hypothetical investment. The most popular method of measuring the marginal tax rate is that derived by King and Fullerton (1984).' They consider a small increase in the level of real investment by the domestic non-financial sector financed by an increase in the savings of domestic households. The technique consists in calculating the minimum rate of return (which we call p) that an investment, net of depreciation, must yield before taxes in order to provide savers with the same gross tax return (which we call s) they would receive from lending at the market interest rate. In the absence of taxes, these rates will be equal. With distortionary taxes, however, they can differ, giving a tax wedge of p-s. The size of the wedge depends on the way projects are financed (e.g., in the form of shares, retained earnings, bonds and bank loans), on the types of assets in which companies invest (e.g., buildings, stocks, equipment, etc..) and on all the parameters of the tax system. Thus, at the margin, a company would undertake a project if the rate of return, after payment of all taxes, is sufficient to persuade potential savers to finance that project. The tax wedge can be used to define an "effective" tax-inclusive mar-tnal tax rate, h-(p-s)/p. The link between the saver and the company is the rate of return the company pays on the savers financial claims (denoted by r). In the case of debt finance, for example, r is simply the real rate of interest paid by the company to the lender. 2. The investment project considered is a single asset such as a machine. It is expected to generate a flow of revenues which declines exponentially over time in real terms, at rate 6 as the asset depreciates. These revenues include the financial rate of return, p, plus the cost of depreciation, 6. At the same time, revenue increases in nominal terms at the general rate of inflation, i, which is expected to remain constant. The flow of net revenues is taxed at rate t. These flows are discounted back to the current period, period 0, at the company's nominal discount rate, p. The present value, V, of this stream of returns is therefore: 1/ King, M.A. and Fullerton, D. (eds) (1984), "The taxation of income from capital: a comparative study of the United States, the United Kingdom, Sweden and West Germany". University of Chicago Press. - 98 - Annex IV Page 2 of 8 - (p+S - r)u v- f (1- t) (p +6)e du 0 (1 - t) (p + 6) p + 6 - i 3. The cost of the project is unity, the initial payment for the asset, less the present discounted value of any grants or tax allowances given for the asset, A. Hence the total cost, C, is C- 1 - A 4. For a marginal investment, V-C; the project is just worth undertaking. This allows us to solve for the value of the pre-tax rate of return, p given the tax parameters and the company's discount rate: (1 - A) 1- t) (p+6 The next step is to relate the company's discount rate, p, to the interest rate, r. In the absence of taxation, p-r+r-i, i.e., the discount rate equals the nominal interest rate. In the presence of taxation, it depends on the source of finance used for the investment. For debt finance, since nominal interest is tax deductible, the rate at which companies will discount post-tax cash flows is the net of tax interest rate. Hence p - i (1 - t) However, if the tax treatment of interest payments from companies is different from the tax treatment of interest payments from other sources, then the discount rate will be that rate which equalizes the after tax return on the two forms of interest income. If ml is the personal tax rate on normal interest income, and mic is the personal tax rate on interest income from companies, then we can write (1 - mIC) (1 - ml)i - p - 99 - Annex IV Page 3 of 8 This leads to a discount rate of [ I For retained earnings and new equity finance, the discount rate also depends on the interaction of the corporate and personal tax systems. If retained earnings are used, then the value of shares held in the firm will increase, so the return to the financier is in the form of a capital gain. The investor will require an after capital gains tax return sufficient to match the return on an alternative use of the funds. Thus, if the yield of a project is p, the investor would require a yield such that (1 - mf)i - p(z). The discount rate for the retained earnings is therefore: (1 - MI) p -i i i(1 - z) In Tunisia, however, there is no tax on capital gains, so z-0 and p - (1 - mI)i For investments financed through new share issues, the required rate of return needed by financiers is affected by the tax treatment of distributed profits. Again the return to shareholders after tax must be sufficiently large to match the return they could get from another use of their funds, which is (1 - md)i. The personal tax rate on dividends is denoted by md. However, in many tax systems, an imputation tax credit is given, so the return to shareholders needs to be grossed up by the imputation rate, c. Hence overall, (1 - md) (1 - m)i - p In Tunisia, neither is an imputation tax credit nor personal tax on dividends, so the discount rate is p - (1 - ml)i This is exactly the same as the one for retained earnings. Hence the basic Tunisian tax system is neutral between retained earnings and new equity. The discount rates on projects financed by the three different types of finance for zero rate taxpayers and for higher rate tax payers are summarized in the table belov. - 100 - Annex IV Page 4 of 8 DISCOUNT RATES ON PROJECTS WITH DIFFERENT TYPES OF FINANCE Type of Discount rate for Discount rate for finance zero rate taxpayers higher rate taxpayers Debt (1 - t)i (1 - t)(l - mI)i (normally) Debt (no tax on (1 - t)i (1 - t)i interest income from companies) Retained earnings i (1 - m')i New equity i (1 - m')i The basic case is completed by introducing the personal taxation of the investor. The real post-tax return earned by the saver, s, is equal to r less personal taxation. If income tax is charged on the nominal return, s (1 - m) (r + ff) - X 5. Finally, for the case of investment in inventories (stocks), there may be an additional tax charge if for tax purposes inventories are valued by the FIFO (first in, first out) method. Essentially, the additional tax is equal to tw in each period. This leads to a modification of the definition of p to (1 - A) P ~ (1-t) (pi-6 -ir +t) -6 6. This completes the description of the basic methodology. The approach used in this paper is to choose a value of r - 5X is chosen - which applies to all investments. Given r, it is possible to calculate p and s, and hence the tax wedge, p-s, and the effective marginal tax rate, (p-s)/p. - 101 - A I Page 5 of 8 B. TUNISIAN INVESTMENT INCENTIVES 7. There are five main relevant forms of investment incentive in Tunisia: exemption from import duties, tax holidays, soft loans, the permanent reduction in corporate tax rates and the exemption of interest payments from companies from personal tax. The circumstances in which these can be claimed vary, as described in Annex III. The intention here is simply to show how each has been modelled within the context of the King-Fullerton approach. Each of the five is taken in turn. ExemDption from Import Duties 8. It is not so much the exemption from import duties as the import duties themselves which must be added to the King Fullerton framework. They are easily included. The basic principle is that if the asset purchased as part of the investment is imported, its effective price rises from unity to 1+X, where X is the rate of import duty. If the whole of the effective price can be depreciated for tax purposes, then the overall effective price of the asset is (l+X)(l-A), where A is, again, the present value of tax allowances and grants. This yields an expression for p as (1 - A) p - + X) ,, (p+6 -XT) -6 (1 - t) If the asset is exempt from import duties, then X is set to zero. Tax Holidays 9. Tax holidays are more complex to model. This is because the King- Fullerton framework was set up only to consider a tax system which was not expected to change at any time in the future. Under the Tunisian system, however, companies may face a different statutory tax rate for five, seven or ten years, to that faced thereafter. To consider a general case, suppose that the proportion of the statutory tax rate faced in the first period of n years was a; thus a-0 and n-5 shows that the company faces a zero tax rate for five years. In addition, the proportion of the statutory tax rate faced in the second period of m years (immediately following the first period) is b; thus, for example, if b-l the tax holiday is not continued after n years. The simplest way to capture this in the King-Fullerton framework is to construct two new variables t and A'. These directly replace t and A in the earlier definition of p, so that, in the basic case (1 - A-) p_ (p + 6-_)6 (1 - t') - 102 - Annex IV Page 6 of 8 To construct the variables t and A it is necessary to return to the definitions of V and C, the marginal benefit and marginal cost of the project. Consider first the definition of V. Allowing for tax holidays as described above, this can be written: num - (p+6 -wr)u V f (p + 6) (1 - at) e du 0 n+m -P6f) + f' (p+ 6) (1 -bt) e dui n + I (p + 6) (1 - t) e du n+m Using this definition, V can be written as (1 - t*) (p + 6) (p + 6 - X) where -(p+6-ir)n -(p+6-fr) (n+m) t- t {a+(b-a)e + (l-b)e } Clearly, if a-b-1, t*-t. Turning to the definition of C, the present value of allowances, A, depends on the tax depreciation rate and the length of the tax holiday. Since tax depreciation is on the straight line method, the number of years for which an allowance can be claimed is T-l/d, where d is the depreciation rate. Taking first the case in which T>n+m, the present value of allowances can be constructed from A - I dtae du n+ - pr) + J dtbe du n + dte (I du n+m - 103 - Annex IV Page 7 of 8 which simplifies to dt -(p-r)n -(p -r) (n+m) -(p-N)T A -_ - {a+(b-a)e + (1-b)e +e } p - E The approach is similar if T