Selected Issues

This Selected Issues paper serves as background to the 1997 Article IV Consultation with Tunisia and sheds light on economic developments in Tunisia during 1990–96. The paper aims to identify more closely obstacles to higher growth and to draw lessons from past experience to help formulate the economic strategy for the coming years, taking into account the prospective evolution of the exogenous environment. The paper also reviews issues related to the further integration of Tunisia into world markets.


This Selected Issues paper serves as background to the 1997 Article IV Consultation with Tunisia and sheds light on economic developments in Tunisia during 1990–96. The paper aims to identify more closely obstacles to higher growth and to draw lessons from past experience to help formulate the economic strategy for the coming years, taking into account the prospective evolution of the exogenous environment. The paper also reviews issues related to the further integration of Tunisia into world markets.

Tunisia: Basic Data

article image
article image

Includes all economic agents except the Central Government

Includes the social security system.

Consolidated operations; payment-order basis.

Tunisia: Basic Social and Demographic Indicators, 1980-96

article image
Sources: IBRD, Social Indicators of Development, 1996, Republio of Tunisia-Poverty Alleviation; 1995; Ministry of Plan and Regional Development, INS, Enquête nationale population emploi 1989, 1991; VIIIème Plan de développement 1992-1996,1992, Volume I; Budget économique, 1994,1995; Recensement général de la population et de I’habitat 1994, principales caractéristiques démographiques de la polupation 1995; and IMF staff estimates.

Most recent estimates, 1989-94 unless otherwise indicated.

Data for 1981.

Data for 1996.

Data for December 1996.

Data for 1984.

Data for 1989.

I. Introduction

1. In recent years, Tunisia managed to sustain a solid growth rate, averaging 4.5 percent during 1992-96, while further consolidating macroeconomic stability. Public expenditure emphasized support for basic social services, and social indicators are among the best in the region. At the same time, economic growth fell short of the ambitious targets set under the recent 5-year development plans (6 percent annual average for 1992-96) and unemployment continues to hover around 15 percent. The reasons for these shortcomings are complex, but likely include the relatively gradual pace of structural reforms, and shortfalls in public sector saving from targeted levels in virtually every year.

2. The following studies which serve as background to the 1997 Article IV consultation, attempt to shed more light on recent economic developments in Tunisia. They aim to identify more closely obstacles to higher growth and to draw lessons from past experience to help formulate the economic strategy for the coming years, taking into account the prospective evolution of the exogenous environment.

3. Chapter II reviews issues related to the further integration of Tunisia into world markets. Against the background of the relatively modest trade liberalization in recent years, which leaves Tunisia still a quite highly protected economy, it identifies the challenges arising from the comprehensive trade liberalization being pursued under the Association Agreement with the European Union (AAEU) as well as global liberalization under the WTO.

4. Chapter III reviews recent trends in investment, highlighting the divergent evolution of investment in different sectors as well as shifts between public and private investment. Investment regulations may explain these trends in part, as sector-specific incentives were in recent years replaced by benefits applying uniformly across sectors. An econometric exercise attempts to identify the main policy variables that seem to be correlated with investment levels.

5. Chapter IV looks at the agenda for future regulatory reform. The focus is on obstacles to the functioning of goods markets (price and sometimes also marketing controls) and factor markets (labor regulations, land issues). However, the weaknesses regarding data on employment and unemployment as well as on wages do not permit to establish a strong link between labor market regulations and the evolution of unemployment.

6. Chapter V explores the consolidation of public finances, the core area of future macroeconomic adjustment. It depicts the recent evolution of public finances, going to the extent possible beyond the central government budget, and identifies the key areas for fiscal reform. On the revenue side, the challenge confronting the Tunisian authorities will be to stabilize revenues in the face of steadily declining import tax receipts, given the dismantling of tariffs under the AAEU, as well as declining revenue from the hydrocarbon sector as oil and gas fields are gradually depleted. On the expenditure side, the challenge will be to reform the civil service in line with a changing role of the state in the economy, and to overhaul an expensive food subsidy system that partly benefits middle- and upper- income groups.

7. Chapter VI reviews Tunisia’s record of financial sector reform. The challenge for the coming years is to narrow the remaining gap vis-à-vis the top performers among developing countries including through the strengthening of the financial soundness of the banking system, the establishment of functioning clearing and settlement mechanisms to give depth to secondary markets in government paper, and laying the ground for the full liberalization of all capital account transactions.

8. In a related context, Chapter VII discusses issues of the exchange system and exchange rate policy. It reviews the operation of the interbank foreign exchange market and the conduct of exchange rate policy which for a number of years has aimed successfully at keeping the real effective exchange rate of the Tunisian dinar broadly unchanged.

9. The final Chapter VIII returns to the central theme of economic growth. After reviewing the recent growth experience, including in an international context, a growth accounting exercise tries to identify the different sources of growth and examines the evidence of linkages between economic policies and capital and productivity growth.

10. Annexes provide basic economic and financial statistics and an updated description of the tax system.

II. Association Agreement with the European Union


11. Recent studies have highlighted the empirical association between increased openness of an economy and higher growth performance (Sachs and Warner (1995)). While causality does not go exclusively in one direction, evidence indicates that policies aimed at greater integration in the world economy raise productivity through the acquisition of new technology and improved cost discipline and innovation arising from increased competition (Pack and Page (1994); Sarel (1996)).

12. Tunisia has been opening up its economy mainly since 1986. At that time, its development strategy shifted from import substitution and state intervention toward a more market-based and export-oriented policy. The policy shift aimed to improve the competitiveness of the economy through better resource allocation while narrowing the external imbalance. Over the past ten years border protection has been lowered significantly, as Tunisia reduced tariffs and quantitative restrictions, including within the framework of its formal accession to GATT in 1990 and the WTO in 1995.

13. In the years ahead, a main impetus for further trade liberalization will come from the Association Agreement that Tunisia concluded with the EU in June 1995 as part of the broader Euro-Med partnership initiative. The Association Agreement has become a catalyst for Tunisia’s overall economic reform strategy by committing the country to a course that can be completed successfully only through the implementation of a range of supplementary economic policy reforms.

Trade liberalization: 1986-96

14. As of 1985, the Tunisian economy was inward oriented and highly protected. High levels of oil export revenue and foreign borrowing had fueled an investment boom, mainly by public enterprises and with emphasis on creating employment (Morrison et al. (1996)). The inefficiencies that resulted implied high production costs that necessitated extensive import protection, and were partly offset for consumers through subsidies and price controls and for exporters through extensive tax concessions and other benefits. Only 18 percent of imports requiring payment in foreign exchange were exempt from prior authorization requirements and less than 10 percent of local production was without protection in the form of quantitative import restrictions. Custom duties ranged from 5 percent to 236 percent, with the average rate exceeding 40 percent. The rate of effective protection in 1986 was 70 percent on average and 124 percent for manufacturing (World Bank (1995b)). Control over import flows was further enhanced as a result of government monopolies for food imports, the magnitude of procurement by the state, and the weight of public enterprises in the economy.

15. By end-1996, significant trade liberalization had been achieved. Quantitative import restrictions wore limited to a negative list that provided protection for an estimated 8 percent of local production. Import duty rates ranged from 0 percent to 53 percent with a weighted average of 32 percent; however, receipts from import duties during 1996 amounted to only 10.3 percent of total imports, reflecting the importance of the off-shore sector, which took in 32 percent of total imports, and exemptions extended to imports entering the domestic economy. The average rate of effective protection in 1995 was estimated at 56 percent.

16. Building the necessary consensus at each phase of the liberalization process took longer than envisaged. The original goal under the 1986 structural adjustment program called for the removal of all quantitative restrictions and the establishment of a uniform rate of effective protection of about 25 percent by 1991. An initial phase (1986-88) saw significant accomplishments including: (i) the merging of tariff schedules; (ii) a narrowing of tariffs to a range of 10-43 percent; and (iii) a reduced prevalence of quantitative restrictions, in particular on equipment goods and inputs in order to streamline protection and remove the anti-export bias. Notwithstanding the lowering of tariffs, protection for locally produced goods was largely maintained. By 1990, still only 26 percent of domestic production was without protection in the form of quantitative import restrictions. The subsequent reduction of such restrictions in several steps, beginning in 1991, was each time accompanied by the introduction of transitional compensating duties (droits compensatoires provisoires or DCP) on finished and semi-finished goods, ranging from 10 percent to 30 percent and to be phased out over three years, thus temporarily reversing the decline in the top import duty rate that had taken place in the late 1980s (see also GATT (1994)). The DCP were largely eliminated by January 1997.

17. As of 1996, the distinguishing characteristics of import protection might be summarized as follows:

  • The average duty rate on textiles, apparel, and leather was 50 percent and the effective rate of protection of domestic production of wearing apparel was estimated at over 100 percent. High tariffs prevailed on many agricultural products, reflecting the government’s objectives of food self-sufficiency and income support for rural areas.

  • An active and generous export promotion policy has been in place since 1972. This policy, in conjunction with the availability of low cost labor, at least initially, and proximity and preferential access to the European market allowed the rapid development of “offshore” processing facilities. Exports and imports of goods and nonfactor services in relation to GDP rose from 61 percent in 1986 to 85 percent in 1996, and the share of manufactured products in total exports rose substantially. However, offshore enterprises produce limited value added, have developed few linkages with the domestic economy, and have provided minimal competition on the local market which has remained sheltered (World Bank (1994)).

  • Prior import authorization requirements1 remain on: (i) passenger cars, owing to implications for the mechanical industry sector through partnership arrangements; and (ii) certain other goods including boats, marble, glassware, lamps, jewelry, refined petroleum, footwear, and fertilizer.

  • Some imports are protected through multiple mechanisms. Imports of grain, edible oil, and citrus (which are export products) are protected through tariffs and are further controlled as a result of an effective import monopoly by the public sector.

18. How does trade liberalization in Tunisia compare with developments elsewhere? With the exception of South Asian countries, Tunisia appears to have moved more slowly and to have kept higher average tariff protection than several other comparable countries and regions. East Asian and Latin American countries generally moved faster during the 1980s and Central European countries during the 1990s in bringing down tariff protection (Hoekman (1995)). By 1996, Tunisia’s weighted average tariff level (32 percent) exceeded that of other MENA countries (e.g., 20 percent in Morocco) and was higher than the average for developing countries (21.4 percent). The average for East Asia was 21 percent, Latin America and subsaharan Africa about 15 percent, and for the world economy 8.2 percent (Havrylyshyn (1996)).

19. Multilateral agreements have played an important role in trade liberalization in Tunisia. Consistent with its new outward looking policy, Tunisia formally acceded to the GATT in 1990 and made a further round of commitments in 1993 under the Uruguay Round. As a result, total bindings represent about 60 percent of all tariff items (100 percent in agriculture), which is comparable to what developing countries have offered. The tariff levels bound are quite high and well above actually applied tariffs. On textiles, a uniform rate of 90 percent will be bound starting in 1996 and will decline gradually to 60 percent in 2005. In connection with the entry into force of the WTO Tunisia replaced its QRs on agricultural products by tariffs in 1996: all agricultural items were bound at tariff rates varying between 25 percent and 200 percent, with some limited tariff reductions on these products to be phased in through the year 2004. The consultation process with the GATT/WTO Committee on Balance of Payments Restrictions has been instrumental in the reduction of quantitative import restrictions in the 1990s.

20. Regional integration has encountered substantial obstacles. With the aim of strengthening regional ties, Tunisia has since 1960 established links to several Arab and sub-Saharan African countries through bilateral trade agreements that specified either preferential or most-favored-nation treatment. A more extensive multilateral arrangement with Algeria, Libya, Mauritania, and Morocco was formed as the Arab Maghreb Union (AMU) in 1989. The ambitious goals under the AMU included the establishment of a free-trade area by end-1992 and a customs union by 1995. However, trade relations continue to be governed by bilateral agreements as reaching consensus at the AMU level on a list of commodities that will circulate freely and on harmonization of tariffs with nonmembers has not yet bean accomplished. Factors responsible for the slow progress include: political differences, unequal progress in different member countries in pursuing economic adjustment policies and in removing distortions from state intervention and price controls, and concern about the cost of trade diversion. Tunisia’s trade flows with other AMU countries represent about 7 percent on the export side, mainly semi-processed goods, and about 4 percent of its imports, mainly energy products from Algeria.

Association Agreement with the EU

21. Underlying the signing of the Association Agreement by Tunisia and the EU is the shared view that the dismantling of tariff protection in Tunisia over a 12 year period as required for the establishment of a free trade area in nonagricultural products, combined with a harmonization of trade-related regulations and stepped-up financial and technical assistance from the EU, will enhance prosperity and employment in Tunisia and thereby help ensure political stability and security on both sides of the Mediterranean.2

22. One limitation of the benefits of the agreement that has been emphasized is the absence of a significant increase in access for Tunisian products to the European market. This largely reflects the fact that Tunisia already had privileged access to the European market under the 1976 trade and cooperation agreement. In particular, the main agricultural exports, namely olive oil, citrus fruit, and wine benefited from export quotas at the Union’s support prices (i.e., were fully or partially exempt from tariffs and variable levies). However, these quotas were granted only for a limited period of time and were renegotiated from time to time in the context of new multi-year protocols. Exports of manufactured goods have been admitted free of duty either for unlimited amounts or, as in the case of textiles, for specified quotas under a bilateral “voluntary” agreement. Actual exports usually exceeded the quotas. Thus, a main concern on the Tunisian side during the negotiation of the agreement from 1992 onward was the need to firmly lock in guaranteed access for agricultural exports, in particular in light of the implications of the expansion of the EU to the North and eventually the East, and of the Uruguay Round.

Fiscal implications

23. Estimates indicate a gradually rising revenue loss as a result of the agreement from 0.2 percent of GDP in 1996 to 0.5 percent in 1997 and further to 2.4 percent of GDP by 2008, when the agreement will be fully implemented. These estimates are based on the simplifying assumption of no change in the geographical origin and level of imports and of GDP from their 1995 magnitudes.3 Revenue from custom duty in 1995 was 4.3 percent of GDP. Hence, nearly half of custom duty receipts would still remain owing to the exclusion from the agreement of some non-exempt imports (such as industrial commodities with agricultural content) and the fact that an estimated 31 percent of imports subject to taxation originates from outside the EU. In comparison, collection of value added tax (VAT) amounted to 5.2 percent of GDP and excise taxes to 3.5 percent of GDP in 1995, while tax revenue (excluding social security contributions) was 20 percent of GDP and total central government revenue (including social security contribution) 30 percent of GDP. Thus, the estimated revenue loss, once the agreement is fully in place, corresponds to slightly more than one fourth of the receipts of VAT and excise taxes combined. It should be noted, however, that the revenue loss is likely to be higher if account is taken of the impact of trade diversion and a possible transitional setback in industrial production.

24. Absent compensating fiscal measures, the tariff reduction is expected to aggravate external and domestic imbalances. Fiscal neutrality, however, could be preserved by raising the rates or broadening the base through removal of exemptions of domestic indirect and direct taxes. At first sight, the revenue loss of the government is the gain of the nongovernment sector which benefits through lower prices for imported goods. In practice, however, instituting compensating revenue measures may be complicated if the incidence of the tax break and the tax hike falls on different groups of consumers and producers. Furthermore, regional (more than multilateral) liberalization carries the risk that a portion of the foregone fiscal revenue could be captured by EU exporters to the extent that they have the market power to raise their export prices (see Panagarya (1995)), thereby causing a deterioration of Tunisia’s terms of trade. Such considerations have been advanced to help underpin the case for stepped-up financial assistance from the EU. Alternatively, fiscal neutrality could be preserved through expenditure reduction, which would also be consistent with reducing the size of the government sector in the economy.

25. The early experience with compensating fiscal measures in 1996 and 1997 seems to indicate that the authorities find it difficult to compensate fully for the fiscal losses incurred. The estimated revenue loss from trade liberalization (including the reduction of DCP under the WTO and the lowering of duties on a multilateral basis) in 1996 amounted to TD 55 million. Revenue measures taken in 1996, inter alia with a view to offsetting this loss, amounted to about TD 80 million, equivalent to 145 percent of the loss, and included the removal of VAT exemption on certain imported capital goods at the beginning of the year and an increase in the VAT rate on tourism from 6 percent to 10 percent from September 1996 onward. Although the extent of compensation appears high, tax revenue fell to 19.9 percent of GDP from 20.5 percent in 1995, suggesting that the measures taken would have been needed in any event to sustain fiscal revenue, even in the absence of the AAEU. For 1997, the estimated loss (compared with the 1995 base) is TD 136 million, including the impact of the elimination of DCPs, equivalent to 0.7 percent of GDP. By comparison, revenue from compensating measures taken in 1996 and from those included in the 1997 budget are estimated at TD 125 million.

Impact on economic growth and employment

26. Estimates in the literature of the overall welfare gains for the Tunisian economy as a result of the full implementation of the AAEU range from a small loss to a net increase in overall welfare of up to 4.6 percent of annual GDP once the agreement is fully implemented. The high welfare gains estimated in some studies (in particular, Rutherford et al. (1995)) derive from the inclusion of various dynamic gains. Traditional static analyses of the benefits of trade creation net of trade diversion as a result of the reallocation of factors of production to areas of comparative advantage find magnitudes that are positive and range up to 1.6 percent of GDP.

27. Estimated benefits rise by about 3 percentage points of GDP annually as a result of “deeper integration” elements in the Association Agreement that go beyond trade liberalization. The harmonization of Tunisian product standards, customs arrangements and competition rules with those of the EU and the liberalization and upgrading of financial and other services with EU assistance are expected to enhance productivity and lower costs. Contractual assurances of access for Tunisian exports to European markets and the enhanced credibility of the Tunisian authorities’ economic reform commitment are expected to stimulate domestic and foreign investment and accelerate the transfer of technology with additional gains estimated in a range of 2 percent to 3 percent of GDP.

28. The “static gains” increase in welfare results from a process involving job destruction and creation as the economy reallocates labor and capital to sectors of comparative advantage. Different methods have been used in the literature to capture the social cost of the transition. Computational general equilibrium models that assume full employment seek to measure the magnitude of shifts in the sectoral composition of production and the accompanying movement of labor. One model (Rutherford et al. (1995)) equates the cost of shifting and retraining labor to one year of salary per worker who moves to a different industry. Total adjustment costs have thus been calculated as the equivalent of 4 percent of GDP. This is a one-time cost spread over the full 12-year period and beyond. Other models examine the employment that can be expected to be created as a result of the agreement. In one such study (Mahjoub (1995)) the overall impact is a modest 1 percent increase in employment, which however rises substantially once it is assumed that the AAEU will spur large foreign investment inflows and economic growth.

29. The process of upgrading Tunisian industrial enterprises and the environment in which they operate in order to enable them to meet increased foreign competition in their home market and abroad (mise à niveau) has been estimated to cost up to TD 2.5 billion for 1996-2000 (equivalent to 2 percent of GDP annually). Of this total, TD 1.5 billion would need to be spent on strengthening individual enterprises through modernization of equipment, recapitalization, and debt restructuring; the balance on infrastructure and support activities. Most of these outlays will occur as part of enterprises’ investment programs and the public investment program under the IXth Plan. Specific public programs aimed at facilitating technology transfers, strengthening vocational training, and retraining programs, and diffusing knowledge about technologies and export possibilities via sectoral “technical centers” are expected to be financed through a combination of own resources of the enterprises, foreign grants, and budgetary expenditure. A program granting subsidies of 10 percent to 20 percent for investment projects aimed at technological upgrading, amounting to an estimated TD 180 million, is being implemented. The EU is likely to contribute to the financing of such programs; other EU-supported programs aim at reducing the social cost of restructuring.

30. Apart from the specifically designed mise à niveau program, there is widespread consensus that the net benefits of the agreement can be maximized through a coherent and credible reform package that is well understood by the markets. Flexible labor market arrangements, combined with well designed social safety net provisions for labor redundancy, and reforms conducive to the provision of modern financial services at internationally competitive costs would allow economic agents to react swiftly and freely to the new incentive structure for the reallocation of resources. In addition, regulatory reform that confers an increased role to market forces and reduced government intervention in the allocation of resources, including through stepped up privatization and increased competition in the provision of basic services (transportation and communication), should allow improvements in productivity and spur domestic and foreign investment, with its attendant benefits for technological upgrading.

31. What is the likely impact of the EU agreement on the balance of payments?4 The reduction in the average effective tariff level from 10.5 percent to an estimated 6.0 percent (or even lower if trade diversion is significant) implies a reduction in import prices by 4 percent or more. This can be expected to raise demand for imports. Depending on the elasticity of export supply, a real exchange rate depreciation of up to 4 percent might thus be required to restore the current account to its initial level (in foreign currency terms); the required depreciation would be less the higher the elasticity of export supply.

32. The above estimate, however, represents an upper bound, and currency depredation may, in fact, not be required. First, the implementation of fiscal policy measures with a view to neutralizing the adverse budgetary impact will dampen import demand. Second, the agreement can be expected to also exert positive shocks on the balance of payments. These include additional inward portfolio investment and foreign direct investment and additional aid flows, to the extent they do not imply corresponding levels of additional imports. To give some perspective to the amounts involved, the EU assistance Tunisia expects to receive during 1997 in the context of the EU-Mediterranean partnership, inter alia to help meet the costs of the trade liberalization, overall amounts to roughly SDR 100 million, which corresponds to about 2 percent of imports. Third, in the longer term, the adverse balance of payments impact would be reduced further as a result of improved export competitiveness stemming from reduced input prices and the indirect benefits of harmonization of standards and lower cost services. Taking all this into account, market pressure toward a currency appreciation rather than depreciation may also be a possible outcome.

Remaining issues

33. Tunisia has implemented significant, yet limited trade liberalization over the past ten years. Skillful export promotion has been important in overcoming the drawbacks of a still fairly protective environment. The AAEU has engaged Tunisia much further in the direction of trade liberalization than had been achieved through multilateral channels (WTO), and its full and timely implementation is likely to bring considerable benefits. The closer anchoring to the EU through contractual assurance of market access and stepped-up cooperation with the EU including in customs, competition policy, and harmonization of standards, as well as the greater credibility of policy reform commitments and the reduced market protection can be expected to raise economic growth and investment. Nevertheless, the question arises whether further complementary policy steps by the authorities could address some of the limitations of the agreement and help attain its full benefits.

34. First, additional trade liberalization beyond the framework of the AAEU will be important. The relatively long transitional period under the AAEU, while allowing to spread out the adjustment cost, also delays the benefits and could be counterproductive. Furthermore, the sequencing of the tariff reductions, with reductions of tariffs on capital goods and inputs phased in faster than tariffs on locally produced consumer goods, will enhance the rate of effective protection during the early years. The backloading of the more difficult adjustment, in the wake of a slower-than-originally-expected pace of trade liberalization during 1986-96, and combined with the possibility of a slowing of the implementation of the agreement under the various safeguard provisions could leave some uncertainty about the pace of implementation of the agreement, particularly in later years. Accordingly, within the agreed 12-year time frame, the authorities should consider deciding unilaterally to move at a faster pace, to reduce high effective rates of protection.

35. Second, the AAEU is a second-best solution and should serve as a stepping stone towards more universal liberalization. East Asian countries have been successful without entering into preferential trade agreements. The AAEU could divert imports away from more efficient non-EU sources, thereby reducing the benefits of trade liberalization. By contrast, the unilateral reduction of tariffs by Tunisia from all geographical sources on a most favored nation basis would allow the full benefits of the competitive impact. One study (Rutherford et al. (1995)) calculated that this measure would raise welfare by the equivalent of a further 0.7 percent of GDP annually. The net additional cost in terms of labor dislocation would be relatively small because the adjustment costs deriving from trade diversion are avoided under this scenario. Generalizing the agreement to non-EU sources would require compensating fiscal measures to be 22 percent higher than under the EU agreement.5 If full-fledged extension of the tariff reduction is difficult for political reasons, one way for Tunisia to reduce the scope for trade diversion would be to unilaterally align its tariffs on non-EU imports to the external tariff schedule of the EU, pari passu with the implementation of the free trade agreement over the 12 year transition period of the Association Agreement.

36. Third, the deferment of negotiations of liberalization of trade in agricultural products until the year 2000 is largely explained by the political difficulties of reforming the EU’s common agricultural policy and resistance on the part of European countries who compete with Tunisia’s export products. Tunisian agricultural production is also heavily protected through tariff and non-tariff barriers, and liberalization would require major adjustment, for example in the cereal sector. Reciprocal liberalization of agriculture would provide Tunisia further benefits, possibly including increased access for olive oil at EU support prices, although this advantage might be offset as a result of closer correspondence of the latter with world market prices. At the same time, faster liberalization of agriculture even unilaterally (only limited steps are programmed under WTO rules) would help avoid new distortions that may result if resources flow from the relatively unprotected industrial sector into agriculture.

37. Fourth, Tunisia would derive additional benefits from increased market access for its industrial exports to other AMU and EU association countries. The greater market would allow economies of scale and would enhance the attractiveness of Tunisia as a central location for foreign investors, and limit adverse “hub-and-spoke” effects on foreign investment. Thus, concluding free trade agreements with other countries outside the European Union would be mutually beneficial. A free trade agreement with Morocco is already under discussion and may help to advance the transformation of AMU into a well-functioning free trade area.

38. Fifth, beyond trade liberalization, complementary policy reforms could help ensure the attainment of the important dynamic gains of the agreement, help minimize the transitional costs, and bolster the credibility of the authorities’ strategy. Specific measures could include: (i) further deregulation by easing restrictions on foreign investment especially in the service sector; (ii) increased flexibility in the labor market combined with stepped-up social safety net provisions; (iii) a greater role for the private sector and stronger competition in the financial system; and (iv) a bold privatization program and increased competition in key service sectors such as transportation and telecommunications.

39. Finally, Tunisia needs to ensure that trade liberalization does not come at the expense of macroeconomic stability. The adverse budgetary impact should be fully offset through high quality measures of revenue enhancement and expenditure cuts and external stability should be preserved consistent with the authorities’ intention to lower the external debt/GDP ratio. Enforcement of hard budget constraints on state enterprises and civil service reform, consistent with the altered role of the government, as well as further reform of the food subsidy system, should be the focus on the expenditure side. On the revenue side, efforts should concentrate on a streamlining of the VAT system and the curtailing of income tax exemptions as discussed more fully in Chapter V below.

III. Investment—Recent Experience And Prospects

40. During 1960-96, capital accumulation is estimated to have contributed about half of Tunisia’s GDP growth (Chapter VIII), with the capital stock growing on average by about 6 percent a year. Increased levels of investment continue to hold the key to higher growth in Tunisia, and raising both private and public investment levels is a central tenet of the IXth Economic Development Plan. This chapter reviews the past behavior of investment with a view to identifying the economic policies required to stimulate investment, in particular in view of the relatively weak levels registered in the private manufacturing sector in the most recent years.

Recent trends in investment patterns

Overall investment

41. Investment levels have varied considerably over time. Following independence (1956) investment was initially low, but the ambitious, public sector-led development strategy pursued led to more than a doubling of the average investment ratio to 22 percent of GDP during the 1960s (Table 1 and Chart 1). The public sector on average accounted for about two-thirds of total investment during this period. As Tunisia was a net energy exporter (until the mid-1990s), the two oil shocks during the 1970s provided the state with additional resources and the investment ratio rose further during the 1970s and early 1980s, peaking at 34 percent of GDP in 1982; investment by public enterprises reached 14 percent of GDP in that year.

Table 1.

Investment Indicators

(In percent of GDP; unless otherwise stated)

article image
Chart 1
Chart 1

Tunisia: Gross Fixed Capital Formation by Type of Investor, 1961-95

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Sources: Tunisian authorities; and IMF, International Financial Statistics.

42. Investment levels of this magnitude became unsustainable in subsequent years as the economic environment turned less favorable with the fall in oil prices, more frequent and severe droughts especially after 1988, and a decline in workers’ remittances because of dismissal of expatriate workers from a neighboring country. While investment by households (mainly for housing) and the central government rose further during 1982-85, enterprise investment fell from 25 percent of GDP in 1982 to 17 percent in 1985, with the bulk of the decline taking place in public enterprises. During the period of Fund-supported adjustment programs (1986-92), household and central government investment remained broadly stable, while enterprise investment trended upwards.

43. During 1993-95, overall investment levels rose further, mainly reflecting the two large foreign financed projects (Gazoduc and Miskar) in the energy sector which in 1993 alone accounted for about 3 percent of GDP. Private sector investment (including Miskar) trended upwards and government investment also rose. Overall investment in 1996 is estimated at about 24 percent of GDP, unchanged from 1995.

Sectoral composition of investment

44. The sectoral allocation of investment has remained relatively stable in recent decades. Investment in services, including transportation and telecommunications, has dominated total investment (Charts 2 and 3) throughout, and in fact increased its share at the expense of investment in agriculture and in the hydrocarbon sector. The recent surge in 1991-93 reflects inter alia the construction of a second transmediterranean pipeline (Gazoduc) and a high rate of investment in telecommunications. An upward trend in manufacturing investment through 1985 has given way to a broadly constant level at around 4 percent of GDP since.

Chart 2
Chart 2

Tunisia: Gross Fixed Capital Formation by Economic Sector, 1961-95

(In percent of total)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source: Tunisian authorities.
Chart 3
Chart 3

Tunisia: Gross Fixed Capital Formation by Selected Economic Sectors, 1961-95

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Sources: Tunisian authorities; and IMF, International Financial Statistics.

Public versus private investment

45. There has been a shift in the last two decades from public enterprise investment to private enterprise investment (Chart 4). The share of the overall public sector in total investment (including housing) declined steadily from a peak of 74 percent in 1965 to 48 percent in 1995, in the main driven by a slowdown in central government investment in the 1970s and a decline in public enterprise investment in the 1980s and 1990s. Data for the composition of sectoral investments between public and private sector are available only through 1991. On this basis, public enterprises have invested mainly in capital-intensive sectors such as transport and telecommunications, manufacturing, mining, electricity and water. Private investments have been highly concentrated in housing, the hydrocarbon sector (mainly foreign enterprises), agriculture, and tourism (Chart 5). The shift in investment from public to private enterprises would thus reflect mainly the strong expansion of certain services (such as tourism) that are dominated by private enterprise, and more recently also strong growth in private manufacturing.

Chart 4
Chart 4

Tunisia: Gross Fixed Capital Formation by Type of Investor, 1961-95

(In percent of total)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source: Tunisian authorities.
Chart 5
Chart 5

Tunisia: Gross Fixed Capital Formation by Economic Sector and Type of Investor

(In percent of total; averages 1961-91)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source: Tunisian authorities.1/ Excludes hydrocarbons.

Comparison with other countries

46. Gross fixed capital formation in Tunisia, at an average of about 27 percent of GDP during 1980-96, exceeded the average for the Middle East and North Africa (MENA) region (20 percent of GDP) and was closer to the levels prevailing in the fast growing regions of Asia (Chart 6). However, in contrast to most high-growth developing countries, Tunisia’s gross fixed capital formation has been heavily dominated by the public sector, while private sector investment at about 13 percent of GDP during the 1990s was well below the levels prevailing in Asia in recent years (21-22 percent of GDP).

Chart 6
Chart 6

MENA Region. Public versus Private Investment, 1980-95

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source: Tunisian authorities; and World Economic Outlook.

Efficiency of investment

47. The evaluation of capital productivity is complicated by data insufficiencies, including statistical breaks in the time series for GDP and investment. The overall ICOR seems to have been on a downward trend, indicating increasing capital efficiency (Chart 7). Total factor productivity has also been rising (Chapter VIII). The analysis in Chapter VIII provides furthermore some evidence that investment by the public sector was less efficient than private sector investment (see also Morrisson et al. (1996)).

Chart 7
Chart 7

Average ICOR for Selected Countries, 1980-95 1/

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source: World Economic Outlook.1/ ICOR for a given year is defined as the 3-year average ratio of gross fixed capital formation to GDP divided by the 3-year average growth rate of real GDP.

Investment under the IXth Development Plan

48. The IXth Development Plan aims to raise investment to 28 percent of GDP by 2001, with a growing share expected to take place in the private sector. The emphasis will be on productivity enhancing investment in infrastructure, transportation and services, including through build-operate-transfer contracts with private operators. Public investment programs will be specified each year as part of the budget, while structural reforms will be the main instrument to achieve the targeted levels of private investment.

Investment incentives

49. Investment was highly regulated during the 1970s and 1980s. Under the 1969 investment law, all investment required prior government approval. The law provided the same fiscal incentives for all sectors, such as temporary income tax holidays, investment tax credits, exemptions from import duties on imported capital goods, and government guarantees facilitating access to bank loans. It also allowed the government to grant additional advantages for large projects, including a monopoly position and protection from competing imports. During the 1970s, a special incentive system to promote exports was developed, and during the 1980s incentives were further differentiated between sectors and regions, and were extended to include preferential interest rates. Thus, the effective tax rate was virtually nil in agriculture and tourism, while it approached the statutory 35 percent in manufacturing destined for the local market (World Bank (1995b)). The fiscal cost of the tax exemptions under the various investment codes reached an estimated 1 percent of ODP during the 1980s; additional social costs arose because of the resulting distortions of the incentive system. The incentive structure is likely to have favored investment in relatively capital-intensive technology, biased the sectoral allocation of capital, notably toward tourism, and favored debt over equity financing.

50. A new industrial investment code in 1987 eliminated the need for prior authorization for industrial investments that did not seek fiscal benefits under the various codes. However, to date investments in certain service activities, as well as mining and energy production, remain subject to authorization by the relevant authorities. Investment subject to authorization still exceeded 40 percent of total investment in 1993 (World Bank (1995)). Foreign investment in the offshore sector is unrestricted and enjoys national treatment, except in agriculture. Foreign companies or individuals may not own agricultural land, but a system of long-term leases has been developed. In other—mostly services—sectors, foreign direct investment beyond 50 percent of capital in a company requires approval by the Investment Commission (Commission supérieure d’investissement).

51. Fiscal benefits were again harmonized under a new investment code (code unique) introduced in 1993, which replaced the various sectoral codes and most other investment incentive schemes. Most incentives do not discriminate among sectors of production, and are linked to economy-wide objectives such as protection of the environment, promotion of technology, and human capital formation. Tax benefits include accelerated depreciation, reduction of certain taxes (tariffs, VAT, and excises) on imported capital goods, and income tax exemptions on reinvested income. Various subsidy schemes are aimed at encouraging absorption of new technology: under the mise à niveau program launched in 1996, industrial enterprises may obtain subsidies of up to 20 percent of the cost of investments aimed at upgrading production technology, within the context of an approved restructuring plan underwritten by a Tunisian bank.

52. Some sectors continue to benefit also from specific incentives. Agricultural investment is eligible for certain subsidies, and selected agroprocessing activities may qualify for additional benefits. Firms that qualify for full export status (manufacturing or services producers exporting at least 80 percent of output, or agricultural producers exporting at least 70 percent of output) qualify for offshore-status Benefits include a 10 year income tax holiday followed by imposition at 17.5 percent (half the standard rate), as well as full exemption from all taxes on imported inputs or capital goods. Such enterprises may sell up to 20 percent (30 percent in the case of agricultural producers) in the local market, unless their product competes with that of domestic producers. In the latter case, such firms may sell in the domestic market only up to the equivalent of their purchases of local inputs. Partially exporting firms receive similar tax incentives in proportions related to their exports, except that incentives for reinvesting profits are limited to the general tax benefits available for all firms, and refunds of import duties are granted only if there is no locally produced similar equipment available.

53. While the incentive system has contributed to the strong growth of Tunisia’s offshore sector in recent years, it has also introduced some distortions between totally and partially exporting firms. The authorities are currently preparing the operational modalities for implementation of a recent decision to liberalize some of the restrictions on sales in the domestic market by offshore firms. More importantly, to the extent that special export incentives were designed to offset distortions arising from import protection, trade liberalization under the AAEU will reduce the need for these incentives schemes.

Quantitative analysis of private investment behavior

54. One of the central elements of Tunisia’s development strategy is a greater role of the private sector in economic activity. Accordingly, under the IXth Plan, the share of the private sector in overall investment is targeted to increase. In order to identify and quantify the main determinants of investment by households and private firms in Tunisia, a simple regression analysis was carried out. The initial specification of the function for private investment as a share of GDP draws on the recent literature on determinants of investment in developing countries,6 and includes the following explanatory variables: (1) the growth rate of the flow of bank credit to the nongovernment sector, to capture the influence of domestic borrowing constraints on investment; (2) the real effective exchange rate (REER)—the expected sign of which is ambiguous as a depreciation encourages investment in the traded goods sector, but also results in a higher relative price for imported capital goods; (3) the one-period lagged private investment ratio, to capture adjustment lags and other inertial factors; (4) the real growth rate of GDP, lagged one period, to capture demand effects;7 (5) the real interest rate, lagged one period, measured by the nominal rediscount rate of the central bank adjusted by the GDP deflator, as a proxy for the real cost of capital; (6) the terms of trade, measured by the income effect of changes in export and import prices,8 (7) the ratio of public investment to GDP (defined as government plus public enterprise investment) to examine possible complementarity or substitutability between private and public investment; and finally (8) the foreign exchange premium (defined as the ratio of the parallel market rate to the official rate), as a measure of macroeconomic uncertainty.9

Empirical results

55. The regression results are reported in Table 2, underlying data in Table 4. Estimates of the private investment function including all potential explanatory variables yielded implausible (wrong-signed) and insignificant parameter estimates for several variables (foreign exchange premium, import, and export price shocks). An initial specification of the investment function, excluding these variables was then estimated (equation 1); subsequently, statistically insignificant explanatory variables—public investment, and terms of trade shocks—were successively removed, to arrive at the final specification of the investment function (equation 3). The lack of significance of public investment, which holds both for central government and public enterprise investment separately, as well as the sum of the two—is somewhat surprising, given that some degree of correlation between government and nongovernment investment would appear likely.10 The overall results of equation 3 nevertheless appear robust, with over 70 percent of private investment explained by its determinants, and the significance of the estimated coefficients quite high11

Table 2.

Investment Function - Specification Search 1/2/3/

(Dependent variable: gross fixed private investment/GDP, in logs)

(OLS estimates)

article image

Values in parentheses under the coefficients represent t-statisttcs; the stars indicate significance with *=10%, and **=5%.

Tests for serial correlation, heteroscedasticity, and normality were conducted and produced results implying that no correction to the least squares specification is needed, except in EQ3 estimated for the period 1971-85, where standard errors and t-statistics were computed using White’s heteroscedasticity-consistent-variance-covariance estimator.

An ADF test for unit roots on the residuals of EQ3 indicated stationarity of the residuals at the 1 percent level.

Table 4.

Data Underlying Regressions

article image
Sources: Data provided by the Tunisian authorities; INS; IFS; and Fund staff estimates.

56. The regression results point to the fact that the following domestic policy variables were the main determinants of private investment in Tunisia:

  • The lagged real interest rate has a negative effect on private investment. An increase in the real interest rate of 1 percentage point, from recently observed levels of about 4 percent, would reduce investment in the order of 0.3 percentage point of GDP.

  • The rate of growth of domestic bank credit to the nongovernment sector has a positive but very small effect, which may reflect biases deriving from the inclusion of public enterprise credit in the explanatory variable.

  • The real effective exchange rate is found to have a small negative effect. An exchange rate depreciation of 10 percent in real effective terms would increase the private investment ratio in the order of 0.3 percentage points of GDP. This would suggest that any negative “cost” effect of a real depreciation by raising the real price of imported investment goods would be more than offset by the positive “competitiveness” effects for the economy as a whole, notably by raising the profitability of the tradable goods sector.

  • The lagged real growth rate of GDP exhibits sizable positive effects on the private investment ratio. A 1 percent increase in the rate of growth in economic activity in the preceding year would raise the private investment ratio in the order of 0.5 percentage point of GDP. This effect of lagged growth may not only operate through increased economic activity, but may also play an important role on private sector expectations of short-term growth and the investment climate.

57. While the preferred equation tracks the observed path of investment reasonably well, the model misses some turning points, for example in the period 1993-95, and overpredicts investment during the late 1980s (Chart 8). This may reflect the rote of uncertainty—not captured in equation 3—on investment. The period 1986-95 was marked by comprehensive structural reform, likely associated with considerable uncertainty. As a result, the incentive for investors to wait before embarking on irreversible investments may have been strong.

Chart 8
Chart 8

Tunisia: Actual and Estimated Private Investment, 1971-95

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Stability of the estimated coefficients

58. The implementation of structural reforms in Tunisia since 1986, including liberalization of most domestic prices, financial sector reforms, an outward-oriented trade strategy, as well as weather-related shocks may have altered the way in which investment responded to policy changes. Accordingly, in order to investigate how sensitive the estimated coefficients are to changes in the sample period, the stability of the estimated investment function over time has been examined via forward recursions, based on an initial sample of 1971 to 1980, to which annual observations were added until reaching the full sample (Chart 9). This exercise points to considerable stability in coefficient estimates for all the explanatory variables except for credit to the nongovernment sector.12 Furthermore, the dear reduction in the standard errors as observations for the late-1980s and early 1990s are included further substantiates the robustness of the estimated coefficients.

Chart 9
Chart 9

Tunisia: Recursive Least Square Estimates of Private Investment Equation

(Slope Parameters in Forward Recursions) 1/

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

1/ Recursive least squares estimates were based on samples starting in 1971 and ending in the year displayed on the time axis.Forward recursions added annual observations one-by-one to an Initial sample covering 1971-80.

59. The analysis also points to some limited structural change in the coefficients. The impact of the real effective exchange rate on investment declines somewhat when observations after 1986 are included while the significance of the coefficient increases; tins may indicate that with greater openness, and a growing share of manufacturing in GDP, the (negative) effect of a real appreciation via its impact on the relative price of imported capital goods became stronger. By contrast, the negative effect of the real interest rate on private investment increases as observations are added after 1986, consistent with a more market-based allocation of credit through interest rates rather than quantity controls.

60. However, there is only limited evidence of a structural change in the determinants of private investment during the “pre-adjustment period” (1971-85) and the period starting with the implementation of Fund-supported programs (1986-95). Estimating the preferred equation for these periods separately (Table 2) preserves the signs of all the coefficients, and implies an even stronger role of the real exchange rate in determining investment. However, the limited number of observations makes clear-cut inferences difficult.

Explaining the performance of private investment

61. To assess how policy and exogenous factors influenced private investment, the preferred investment equation has been used to decompose the estimated change in private investment into the contribution of each explanatory variable. Table 3 presents this decomposition for four subsamples based on different phases of the private investment cycle: first growth stage (1971-75), second growth stage (1977-81), strong decline (1981-87), and recovery (1987-94).

Table 3.

Tunisia: Contribution of Explanatory Variables to the Changes in the Private Investment

(Based on equation 3, Table 2, 1971 to 1995)

article image

In percent of total estimated change.

62. A common feature of all subperiods is that the growth of credit to the nongovernment sector contributed little, whereas the real effective exchange rate contributed significantly to changes in private investment. The early 1970s were marked by significant fluctuations in the explanatory variables. Declining real interest rates, accompanied by a real effective depreciation of the Tunisian dinar, and high growth rates fueled the rise in private investment from 1971 to 1975. Similar trends during the period 1977-81 with, in addition, increased growth of credit flows to the nongovernment sector caused private investment to surge by almost 4.5 percent of GDP. However, only 50 percent of the change in investment between 1977 and 1981 is explained by the equation. Moreover, the equation underpredicts the path of actual investment for the period 1978-81.

63. The sharp decline in the private investment-to-GDP ratio from its peak of 11 percent in 1981 to 5.7 percent in 1987 is attributable to a large extent to the decline in GDP growth over the period, coupled with an increase in the real interest rate, which was only partially offset by the continued depreciation of the Tunisian dinar in real terms. The economic reforms beginning in 1986 coincided with a recovery in private investment. Both domestic price liberalization, which helped stabilize real interest rates, and other financial and external sector reforms may have contributed to the increased economic growth, which in turn contributed to the recovery in private investment and to the investment climate in general.


64. Notwithstanding data insufficiencies that limit the preceding quantitative analysis, it appears that real interest rates and access to credit matter for private investment, pointing to the importance of prudent fiscal policies to leave sufficient resources for private investment. The importance of the real effective exchange rate indicates that profitability in the tradables sector, where most of private investment is likely to occur, also semis to be crucial. These findings thus support the macroeconomic strategy adopted by the authorities under the IXth Plan. The impact of structural parameters, including the investment incentive system, could not be quantified. However, the increase in the investment ratio during 1987-94, may well be related to growth-enhancing structural reforms during this period, including the overhaul of investment incentives. Some of the impact of time reforms is picked up through the growth-variable, but its positive impact may have been much larger, given that investment rose much more than predicted by the estimated equation. Again, this underscores the importance of bold structural reforms to achieve the targets of the IXth Plan.

IV. Regulatory Framework Issues


65. Since abandoning a state-centered development strategy in the late 1960s, Tunisia has implemented substantial reforms to establish an institutional and regulatory framework conducive to the functioning of goods and factor markets, integrating the economy with world markets, and allowing the private sector to act as an engine of growth. To a large extent, price and marketing controls have been removed, except in agriculture, where they remain linked to socio-political concerns and the subsidization of basic food staples. The gradual liberalization of the trade and exchange system (Chapter II) has established a closer linkage of the structure of domestic relative prices to international prices. Financial sector reform has improved the allocation of capital, although important reforms have yet to be taken (Chapter VI) and foreign investment remains restricted in a number of areas. The functioning of land markets, in particular for agricultural land, is improving as weaknesses in the titling system are being addressed. Important reforms such as the modernization of the commercial code, a strengthening of creditors’ access to collateral, and the clarification of legal issues falling under several national jurisdictions, are under way. However, only recently has the divestiture of the large public enterprise sector—a legacy from the state-centered development strategy of the past—gained momentum, its pace determined by the socio-political constraints the government feces. Furthermore, restrictions on worker lay-offs and binding minimum wages, motivated by the strong concern of the government to preserve social peace and enlist support of Tunisia’s labor union for economic reform, may partly explain relatively high unemployment.

66. Tunisia’s reforms, while gradual, never suffered serious reversals, and have contributed, together with generally prudent and responsible macroeconomic policies, to a growth performance among the best in the region, though it fell well short of East Asia’s most successful economies (Chart 10). The shortfall is essentially explained on account of lower levels of investment and lesser gains in total factor productivity (Chapter VIII), reflected also in relatively low levels of foreign direct investment outside the energy sector.

Chart 10
Chart 10

Per Capita GDP, 1970-95

(In current U.S. dollars)

Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source; IMF, International Financial Statistics; WEO for Tunisia.

67. This points to an important unfinished agenda for future reform. Most importantly, there is a need to further reduce the role of the public sector in production and distribution, eliminate remaining price and marketing controls and rigidities in factor markets, and relax entry restrictions notably for foreign investors in sectors that will be crucial for future export growth. Without such reforms, adjustment to increased import competition from the implementation of the AAEU, and to increased competition on export markets because of globalization and the reduced importance of preferential access to the EU, may be costly in terms of output and employment.

Current labor market regulations

68. Current labor market regulations are the result of a long-standing social dialogue (tripartite) between the government, the labor union (Union générale des travaitteurs tunisiens-UGTT) and the employers association (Union tunisienne de I Industrie du commerce et de I ‘artisanat-UTICA). This dialogue became institutionalized in 1989, and succeeded in putting an aid to the often turbulent relations of preceding periods, thus ensuring social peace and broad-based support for economic reform. Wage and workplace-related issues have been dealt with by the tripartite in the context of 3 year sectoral agreements within the framework of an overall national agreement; the third such agreement, covering 1996-98, was concluded in mid-1996 and covers 48 sectors and an estimated 80 percent of the workforce. As during the negotiations of previous agreements, the government played a role in arbitrating between the positions of employers and workers, notably to contribute to an outcome which was consistent with its broader macroeconomic objectives. Changes in the social security system and labor legislation usually complemented the sectoral agreements.

69. The current labor code features a number of provisions that substantially affect the functioning of the labor market, although greater flexibility was introduced in 1996. Strict firing regulations stipulate that any dismissal for economic reasons is subject to prior notification and consultation with the Inspection du travail; its recommendations can be appealed to the Commission de contrôle du licenciement (CCL), a tripartite body.13 Both institutions typically propose alternative solutions often developed in lengthy negotiation involving the CCL, the employer, and the employees concerned, such as early retirement for older workers (with the cost borne by the social security fund), or higher severance pay than the minimum levels stipulated in the labor code. Frequently, these negotiations involve substantial delays.14 Revisions to the labor code in July 1996 clarified the role of both institutions, and fixed at 15 days the maximum statutory delay for the Inspection du travail and the CCL to issue their recommendations, respectively. The revisions also eased some of the other remaining labor market constraints, by eliminating the requirement to obtain approval for new hiring from the official employment agency.15 Under the revised code, fixed-term employment contracts may be freely concluded, but may not exceed a total of four years (renewals included).16 The revisions to the code also increased the employers’ flexibility regarding holidays, vacation, and length of the work week.

70. Another important feature of Tunisia’s labor market is the strict enforcement of minimum wage laws. The government sets minimum hourly wages for the non-agricultural sector (SMIG)17 and minimum daily wages for workers in agriculture (SMAG).18 The lower wage for agriculture refects the difference in cost of living between rural and urban areas but is also consistent with an estimated 20 percent difference in labor productivity (Hakim (1995)). In international comparisons, minimum wages in Tunisia tend to be rather high as a share of per capita income. Minimum wages in real terms have trended downward during the 1980s, but have been broadly constant during the 1990s (Chart 11). Although the salary grid established in the triennial wage rounds is based on the minimum wages, time is little evidence of minimum wages influencing the average wage over the longer term; wages appear to be broadly in line with productivity developments (Azam (1995)). Differentials between administratively set wages (SMIG, SMAG, and public sector wages) have been broadly constant in the last decade. The 1996 revisions also provided somewhat greater scope for firms to link wages at the firm level to individual productivity, as long as the resulting salaries are not lower than those established in the framework agreement.

Chart 11
Chart 11

Wages Real Wage Indicators, 1981-96 1/


Citation: IMF Staff Country Reports 1997, 057; 10.5089/9781451837742.002.A001

Source: Tunisian authorities; and staff estimates.1/ Nominal wages deflated by CPI.2/ Minimum wage in the nonagricuttural sector.3/ Minimum wage in the agricultural sector.

71. Finally, nonwage costs are important. Until recently, employers were charged a 17.5 percent payroll tax to the social security fund (Came nationale de sécurité sociale (CNSS))19 while the levy on employees was 7.75 percent. The employers’ contribution was lowered effective October 1996 to 15.5 percent, and a planned increase in the contribution of workers was canceled. While these contributions are generally lower than in Europe, the demographic dynamics are likely to enable the CNSS in coming years to record a modest surplus, while maintaining benefits, which currently include health care,20 pensions, family allowances and some direct transfers to the poorest, but no unemployment benefits. A new law was adopted in November 1996 which, inter alia, extended social security coverage for up to one year to unemployed members dismissed for economic reasons; the law also established that the CNSS will ensure payment of severance benefits in the case of insolvent enterprises. At the same time, the legal framework is being developed to allow employers to reduce their contributions provided they contract group health insurance for their employees with private insurers. Less favorable demographics for the civil service will, however, likely require adjustments in the operation of the CNRPS over the medium term.

72. An assessment of the employment impact of labor regulations in Tunisia is hampered by insufficient data. Labor market surveys are conducted only every few years in the context of a census or a comprehensive living standards survey (most recently in 1989 and 1994); unemployment data from the 1994 survey have not yet been made public.21 While the methodology and classifications underlying labor statistics corresponded to prevailing international norms in the 1960s, the definition of the labor force has since been modified to include actifs marginaux (unemployed workers not actively seeking a job but who were partly or fully employed in the three months preceding the poll) and actifs partiels (household workers who would accept some additional salaried employment, but are not actively seeking it).22 At the same time, Tunisian labor statistics differ also from international norms in not including among the unemployed job seekers of less than 15 or more than 60 years of age. Thus, while official statistics point to fairiy stable unemployment rates of around 15-16 percent in recent years (Table 5), applying international norms to the same data would indicate a lower rate of unemployment at around 11 percent as of 1989 (Rama (1995)). For less-qualified workers and/or first-time job seekers, available data point to particularly inefficient labor intermediation, as both vacancies and unemployment rates are high. Also, labor market regulations are likely to have contributed to the emergence of an informal sector, which according to the 1989 survey employs nearly 25 percent of the work force.

Table 5.

Tunisia: Wage and Employment Indicators, 1981-96

article image
Sources: Ministry of Economic Development; World Bank; and Fund staff estimates.

Labor force as percent of population; World Bank data (SID).

Labor surveys and census; 1981 column corresponds to 1980 survey; 1994 data are preliminary estimates based on 1994 census.

Deflated by the average annual increase in the consumer price index.

Agenda for later market reform

73. The analysis points to a number of ways to improve the functioning of the labor market and thereby raise the potential for economic growth. More flexible labor market policies will be essential to help the Tunisian industry adjust to the removal of trade protection under the AAEU by facilitating the reallocation of labor toward sectors of comparative advantage. In this context, a better monitoring of developments in the labor market is required to provide policymakers with the statistics necessary to formulate well-informed policy responses. High unemployment among the young/nonqualified points to the need to exercise caution in setting minimum wages that may represent insurmountable hurdles for low-productivity job seekers. Building on Tunisia’s good record in providing education, training could be even more attuned toward requirements of the economy; in particular vocational training, traditionally geared toward absenting dropouts from the school system, should address better the needs of enterprises. Easier lay-off procedures and more touted severance benefits would make hiring less irreversible and mistakes in hiring less costly, and thus reduce existing disincentives for offering employment. The collective bargaining system could leave more room for decentralized negotiations at the subsector/firm level; in particular, to broaden the scope for trade-offs between various benefits, firms could be allowed to negotiate wage agreements outside the conventions collectives, without being required to offer better conditions in all areas. Increased scope for fixed-term contracts and for temporary reassignment of workers could further enhance the functioning of the market.

Price controls

74. As part of its overall strategy to allow market forces a greater role in the allocation of resources, Tunisia in the mid-1980s started a process of gradually liberalizing the pervasive system of price control. Adoption of a new price and competition law in 1991 accelerated the process. The law establishes that prices are free unless explicitly regulated, and stipulates rules to prevent collusive behavior and combat restrictive practices, while promoting full disclosure of conditions governing commercial transactions. A consultative body (Conseil de la concurrence) was set up; it may initiate investigation of anti-competitive behavior, can be approached by firms or individuals who consider to have been affected by anti-competitive practices, and may be consulted by the Ministry of Commerce when deciding on the approval of mergers or acquisitions that could result in excessive market concentration. The law also specifies that the government may reinstate price controls for a period of up to six months, if necessary to protect low-income groups or in cases of insufficient competition. The number of remaining price control systems was reduced to two: (1) under the cost-plus system, which is applicable either to the production and distribution stages or to the production stage only, the government establishes a price for a certain product that applies to all producers and distributors in the sector or only to a given enterprise; and (2) under the automatic cost-plus system, which is applicable only to the distribution sector, enterprises establish their prices by applying a maximum margin set by the government. A law on domestic trade introduced in mid-1991 abolished the requirement that traders obtain authorization before engaging in retail and wholesale trading activities, thereby easing entry into the sector. New rules for the marketing of agricultural and fishing products, and establishment of a commerce register, were promulgated in 1996; and a supervisory agency (Observatoire national de l’approvisioimement et des prix) was set up to monitor the evolution of supplies and prices in Tunisian markets.

75. The list of items subject to price controls was reduced in annual tranches during 1991-95, and by end-1995 about 13 percent of domestically produced goods (measured by value added) remained controlled at the production stage; price controls at the retail stage were reduced to 19 percent of total sales. Remaining controls at the production stage mainly affect certain basic commodities, notably subsidized consumer staples including bread, flour, couscous, edible oils, sugar, powder milk—accounting for about 40 percent of goods under price control; certain goods produced or distributed by state monopolies, such as tea, coffee, tobacco, alcoholic beverages, and cement, altogether accounting for another 40 percent of controlled products; medicine, subsidized paper and books, gasoline, and motor vehicles; and finally services such as medical fees, passenger transportation, water and electricity, port handling and telecommunication services. At the retail level, controls apply to the above-listed products, as well as fruits and vegetables—despite high competition among small traders and no controls at the production stage—and cars, to limit the scope for oligopolistic profits under the protection of existing import controls.23


76. Tunisia has phased out consumer and input subsidies on a wide range of items, most recently on fertilizers and animal feed; consumer subsidies remain only on certain basic food staples (cereals, edible oils, sugar, milk) and on school books for social reasons. Efforts have been made to improve the targeting of subsidies by limiting them to variants of the product that are predominantly consumed by the poor. As a result, the amount of subsidies remained constant at about 2.2 percent of GDP during 1994-96. As nominal budgetary outlays for subsidies were tightly limited under the VIIIth Plan, the government accumulated arrears toward the public enterprises (mainly Office des céréales (OC), and Office national de l’huile (ONH)) who administer the subsidies; this quasi-fiscal part was financed by banks, mainly the Banque nationale agricole (BNA). Since 1996 most of the operations are handled directly by the Ministry of Commerce while before they were administered through the Caisse générale de Compensation (CGC). However, the subsidy system remains closely interrelated with a web of marketing and entry controls in the agricultural sector.24

77. The Agency for Cereal Marketing (OC) collects the local harvest through two cooperatives at administratively set prices (which have exceeded world market prices since the mid-1980s). Its quasi-monopoly for cereal imports was abolished in 1996 with the elimination of quantitative restrictions on cereal imports, but given the low fixed prices for subsidized cereals products, private imports remain unprofitable. The OC sells the cereals purchased or imported to private millers, below cost, with the difference covered by a subsidy from the CGC. Millers must sell the flour eligible for subsidy to bakeries and other food processors at a fixed price, which is below their cost (including a fixed profit margin), with the difference covered by another subsidy from the CGC, administered through the OC. Producers must sell subsidized bread, couscous, and pasta at the official price, which is set to cover cost and a fixed margin, and they do not receive subsidies. Extensive controls are aimed at preventing leakages of the subsidy, in particular diversion of subsidized flour for use in bread or pastry that is not subject to price controls.

78. Similarly, the ONH maintained until mid-1996 the monopoly on grain-based oil imports, as well as a dominant role in olive oil exports, as it is charged with administering the export quota granted by the EU. For basic varieties of edible oil covered by the subsidy scheme, the ONH allocates imported raw oil for processing to private refineries; this allocation takes place on the basis of traditional quotas reflecting capacity, rather than competitive bids, and thus does not seek to minimize costs. A practice to subsidize transportation to refineries located far away from ports (péréquation) was abolished in 1996. The ONH sells the refined grain oil to retailers at a fixed price below cost; the difference is covered by a subsidy from the CGC.

79. Three parastatals dominate the sugar sector. The Office du commerce de la Tunisie is the sole importer of refined white and unrefined brown sugar; the Société tunisienne du sucre refines brown sugar and local sugar beet and the Complex sucrierde Tunisie produces white sugar for industrial use. The latter two enterprises have to sell their output at regulated pices, and as the CGC pays the difference between total revenues and total documented costs, they have no financial incentives to be cost efficient.

80. Two parastatals (Société des industries laitiéres and Tunisie-Lait) dominate imports, processing, and distribution of milk reconstituted from milk powder although their formal import monopoly was abolished in 1990. Both receive a subsidy to cover the difference between the controlled sales prices and their cost; private distributors/importers are not eligible for the subsidy. By contrast, the government instituted in 1980 a fixed unitary subsidy for fresh local milk paid to producers, for which private factories are also eligible.

Entry restrictions

81. The current subsidy system has sustained the existence of public marketing controls or monopolies in the agricultural sector, either de jure or de facto. Cereal marketing remains largely restricted to the OC and a few cooperatives; the management of the EU export quota by the ONH has strengthened its de facto monopoly on olive oil exports; and state trading monopolies remain for tobacco and alcohol imports. Public monopolies or restrictions on private sector access are also maintained in the utilities sector such as air transport on certain lines (charter flights are liberalized), certain shipping lines, and port handling.25 It is estimated26 that resulting inefficiencies in the transport/port sectors have added about 17 percent to the transport cost of Tunisian exports, or 1.2 percent of average export value, implying a serious impediment for Tunisia’s export competitiveness.

82. Outside agriculture and public utilities, restrictions on investment have been substantially liberalized over the past ten years, although controls are maintained in a number of areas, (see Chapter III for the evolution of investment regulations). Tax incentives for investment since 1993 are granted on a horizontal basis, although certain sector-specific incentives persist. A number of additional restrictions apply to foreign investment in certain sectors.

Land markets

83. The land tenure system is an important constraint on the further development especially of agriculture (World Bank (1995)). Many existing titles are outdated as recent changes in ownership have not been recorded in titles, although regional commissions were set up in 1992 to resolve tilts issue, in the context of new legislation aimed at speeding up the process of changing title. The law specifies that from 1998, only changes in ownership recorded in title will be legally recognized. Large areas of government land are awaiting title registration.27 Farmers in any area may register their land and obtain title by paying a small fee. In addition, in rural areas, selected lands—particularly in irrigated areas—are being registered in newly established cadastres free of charge under specific programs spelled out in the five-year plans. A program initiated in 1991 that leases prime state land formerly managed by cooperatives to private farmers or cooperatives has been highly successful in boosting productivity.

Reform agenda

84. Ensuring high growth in the coming years, despite the drying up of traditional sources of growth (hydrocarbon discoveries, preferential access to European markets), and despite likely transitional strains in the industrial sector as tariff protection is dismantled under the AAEU, will require an acceleration of structural reforms to achieve the private investment levels and gains in factor productivity called for to achieve the objectives of the IXth Development Plan (1997-2001).

85. A number of reforms are already under way in the context of the mise à niveau program supported by the EU, as well as reforms supported by World Bank lending. For example, in the context of a World Bank-supported program, Tunisia committed to specific targets for 1997 to accelerate privatization, lift restrictions on private sector access to shipping and port handling; revise the corporate law notably to modernize merger and takeover rules, and clarify the legal basis for holding or single-shareholder companies; adopt a new Private International Law to clarify legal rules for cases that fall under several national jurisdictions; and prepare by mid-1997 a medium-term strategy for the telecom sector that would include incorporation of Tunisia Telecom (currently a government body) as an autonomous public enterprise, while opening certain services using the public network to the private sector. Preparatory work is also under way to modernize the commercial law (especially as regards payment rules, and regulations for new type of commercial contracts such as factoring and franchising), and strengthen real and private property regulations to allow better enforcement of creditor rights. Efforts to facilitate the functioning of land markets are underway.

86. Other reforms would be needed to complete the transition to a market-based, private sector-driven economy. These include notably abolition of remaining price controls. Abuses could be dealt with through the anti-monopoly commission, while accelerated import liberalization can also help to reduce the scope for domestic market power, which seems anyway rather limited for most of the food products still under price control. Remaining controls in marketing will need to be removed concurrent with subsidy reform. If food subsidies were to be maintained, the authorities could explore ways to move to a system of fixed unitary subsidies for the varieties consumed by the poor, and payment of subsidies could furthermore be made dependent on world market prices exceeding certain thresholds. This would establish transparent and binding budgetary envelopes for subsidies outlays, and would allow to: (i) eliminate controls on the collection of cereals from farmers; (ii) fully liberalize the importing and refining of edible oils; and (iii) enhance competition in the sugar industry.

87. Remaining entry controls in the transportation sector constitute serious handicaps for the overall economy. For public utilities more generally, a bolder pace in involving the private sector in the provision of such services (including through build-operate-transfer agreements) would improve infrastructure services. Finally, investment controls in particular sectors, including on foreign investment, run counter to Tunisia’s aim to boost investment levels and accelerate technology transfer, and likely constrain Tunisia’s export opportunities. While it may be politically too difficult to lift controls on foreign land ownership soon, other controls could be eliminated swiftly. Environmental, health, or moral concerns motivating such restrictions can be addressed more efficiently through economy-wide regulations that need to be strictly enforced.

V. Fiscal Policy Over the Medium Term

88. Fiscal policy remains at the center of Tunisia’s efforts to consolidate macroeconomic stability and achieve higher economic growth. Generally prudent budgetary policies in recent years have contributed to sustainable macroeconomic balances and resulted in a declining trend of public debt in relation to GDP. At the same time, during the period of the VIIIth plan (1992-96), the deficit exceeded the budget target each year, which may have contributed to the shortfall in economic growth performance compared with plan objectives. This chapter analyzes Tunisia’s fiscal profile and compares it with that of high growth countries in other regions. It also describes the authorities’ fiscal objectives for 1997-2001 and discusses the challenges for achieving them.

Current fiscal profile

89. The deficit of the consolidated central government operations amounted to 4.7 percent of GDP in 1996 (Table 6). The concept used includes the social security system, takes into account the quasi-fiscal outlays of the public food agencies, but excludes grants.

Table 6.

Tunisia: Consolidated Financial Operations of the Central Government, 1991-96 1/

article image
article image
Source: Data provided by the Tunisian authorities.

Includes special funds, fonds de concours, operations financed abroad, net treasury operations and the social security funds.

Includes interest on arrears related to operations of the consumer subsidy fund (CGC).

Includes arrears of the consumer subsidy fund.

Excludes social security system.

Includes government debt instruments held by the social security system; excludes debt of the social security system.