Tax reform in Malawi

Policy, Research, and External Affairs < e 6) ' '/ j L WORKING PAPERS Public Economics Country Economics Department The World Bank August 1990 WPS 493 Tax Reform in Malawi Zmarak Shalizi and Wayne Thirsk Malawi's comprehensive reform of its tax system in the 1980s illustrates many of the issues that developing countries must address when altering the way they levy taxes. ThePolicy, Rescarch. and Extcrnal Affairs Complex distributesPREWorking Papers todissefninatethefindings of woik in progress and to encounge the exchange of ideas among Bank staff and all others interested in development issues. These papers cary the names of the authors, rflect only their views, and should be used and cited accordingly. The findings, interpretations, and conclusions are the authors' own. They should not be attributed to the World Bank, its Board of Directors, its management, or any of its mcmber countries.  Policy, Research, and External Affairs Public Economics WPS 493 This paper- a product of the Public Economics Division, Country Economics Department - is part of a larger effort in PRE to evaluate tax reform efforts in developing countries (RPO 674-52). Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Ann Bhalla, room NIO-059, extension 37699 (55 pages, including tables). Malawi embarked on a comprehensive tax ing surtax. The change was intended to reduce reform in the mid te late 1980s with World Bank production distortions arising from the taxation assistance. This paper looks at the problems in of inputs. The tax was also applied to large Malawi's revenue system that prompted the agricultural and trading establishments. The reform, and examines the nature of the solutions import surtax was modified to be consistent with offered and their rationale. the domestic surtax and the non-protective elements of import tariffs were merged with the In the early 1980s, the flow of extemal funds reformed surtax. These changes helped shift the dropped precipitously. This decline coincided base of taxation from production and trade to with the loss of Malawi's primary foreign trade consumption. Equity features were introduced artery (80-90 percent of exports and imports) due through a redesign of excise taxes and their to the closure of rail lines in neighboring merger with the surtax. For companies, existing Mozambique. These shocks resulted in a sharp investment credits and allowances were amal- increase in the servicing of Malawi's external gamated into a single allowance at a higher rate debt and defense spending, thereby creating a to benefit new investment, keeping the average pressing need for more revenue. At first the effective tax rate high for revenue reasons. Government raised the rates on those tax bases Small changes were made to simplify personal that were administratively the easiest to tax, such income taxes. The ground was laid for improv- as trade. By 1985, the tax to GDP ratio had ing procedures and computerizing tax adminis- increased by almost 50 percent. However, it was tration and creating a tax analysis unit. increasingly apparent that the ad hoc, temporary measures were inconsistent with the creation of a Most of the recommendations have been more liberal economic environment in the long successfully implemented over a three year run. This led to a reexamination of the tax period despite difficult economic circumstances system as a whole. in the country. The new system has even raised more revenue than the old, though that was not a It was decided that the reformed tax system specific goal of the reform. However, since the should build as much as possible on existing budget deficit remains high and inflation contin- instruments, generate at least as much revenue ues to be a problem, further base expansion will and be at least as equitable. With these con- be necessary before the currently high statutory straints in mind, a de facto VAT was created company tax rates and basic sunax can be through the manufacturing/import stage by reduced. introducing a crediting mechanism in the exist- The PRE Working Paper Series disseminates the findings of work under way in the Bank's Policy, Research, and External AffairsComplex. Anobjective of the series is to get these fundings out quickly, even if presentations are less than tully polished. The findings. interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by PRE Editorial Services  TAX REFORM IN HALAWI Executive Summary .................................................. i-vi I. INTRODUCTION . .......................................... 1 II. AN OVERVIEW OF MALAWI'S ECONOMY AND TAX SYSTEM PRIOR TO 1985 ............*....................*..**. . 1 1. Salient Features of Malawi's Economy ............. . 1 2. Tax Structure up to 1978 ................... ... 2 3. Revenue Crisis in the Period 1978-84 ......... 3 4. Ad Hoc Revenue Measures in response to the revenue crisis ............................................. 5 III. 1985 TAX STUDY AND TAX REFORM PROPOSALS .... ........... 10 1. Potential Problems Arising from the Ad Hoc Revenue Measures ........ .................................. 11 2. Framework of the Tax Reform . . 14 3. Taxes on Goods and Services .. 18 4. Company Income Taxes .............................. 26 5. Personal Income Taxes ............... .. ............ 30 6. Tax Administration .... .................. 33 7. Revenue Impact ..................... ............... 34 IV. IMPLEMENTATION OF THE TAX REFORM PROPOSALS ............ 35 1. Handover from Study Team to Technical Assistance Team .......... .................................... 35 2. Reform of Taxes on Goods and Services ............. 37 3. Reform of Taxes on Income ......................... 38 4. Reform of Tax Administration ....................... 39 5. New Issues ............................ 40 V. SOME LESSONS FROM TAX REFORM IN MALAWI ................. 41 Bibliography . .... .. .................. . 45  TAX REFORM IN MALAWI Zmarak Shalizi and Wayne Thirsk Executive Summary Malawi embarked on a comprehensive tax reform with World Bank assistance in the latter half of the 1980s. In January 1985, the government asked the World Bank to locate and analyze the weaknesses in its tax system and to make recommendations for change. Subsequently, it asked for help to finance the technical assistance required to implement those of the reform proposals that were approved by Government. This paper examines that effort. The paper discusses the revenue problems up to the mid-1980s that created the need for reform measures and the nature and rationale of the solutions offered, concluding with a brief review of what has and has not been implemented to date. Since many reforms are still in the process of being introduced, their ultimate success camnot yet be assessed. When Malawi gained independence in the mid-1960s, it inherited a tax system based on that of the U.K. As much as 50 percent of total tax revenue was collected through personal and company taxes -- a pattern more reminiscent of a developed economy than that of a developing Sub-Saharan country, particularly one with little or no mineral wealth. Rates of investment and economic growth in the 1970s were high by historical standards, fueled in part by easy access to external capital markets. The tax structure remained reasonably steady in this period, with personal and company taxes accounting for as much as 47 percent of tax revenue as late as 1978. The 1980s, however, ushered in a serious reversal of economic fortunes. Not only was there a global recession and a sharp decline in  - ii - foreign private lending to developing countries, but also Malawi, a landlocked country, lost its principal international trade rout( -- the rail link through Mozambique -- which had carried up to 90 perzent of its external trade. This series of shocks highlighted a number of weaknesses in Malawi's fiscal system. In the short four year period from 1978 to 1982, servicing of external debt doubled to 28 percent of current expenditures. Defense spending alco grew as a result of the unrest in Mozambique. Even after cutbacks in other expenditures, there was a pressing need for more revenue. This pressure was relieved by increasing the rates on those bases which were administratively the easiest to tax, such as trade. First, the rates were increased on consumer goods. Then, when foreign exchange rationing was introduced with a low priority for luxury and non-essential consumer imports, it became necessary to tax imports of intermediate and capital goods which had hitherto not been taxed or only at very low rates. By 1982, the yield from taxes on imports (including the import surtax with its discriminatory features) had surpassed the combined yield from personal and company taxes for the first time in Malawi's post-independence history. Finally, as revenue pressures continued, taxes on income were increased. In 1984 many of the allowances and deductions in the personal income tax code were repealed and the company tax rate was increased by five percentage points, from 45 to 50 percent. This high rate further weakened the cash position of many companies, particularly those already suffering from declining demand and, with the cutback in investments, from a loss of the offsets from investment related tax incentives and allowances. All of these piecemeal and ad hoc changes made during the six years from 1978 to 1984 were principally motivated by growing revenue pressures.  - iii - In 1984 and 1985 both the Fund and the Bank were recommending that the Malawian government create a more liberal economic environment by dismantling the foreign exchange allocation system and the prevailing system of administered prices. However, in the course of the dialogue it became clear that, once the obstacles to a freely functioning price system had been removed, the currently non-binding distortions of the new ad hoc tax system would start to operate. So it became necessary to reexamine the incentives provided by the system as a whole. It was recognized that additional revenua raising could not be a prime objective. From 1978 to 1984 Malawi had dramatically increased its tax to GDP ratio from 12 to 17 percent -- an unprecedented rise of five percentage points or almost 50 percent of tax revenue in 1978. As rises in tax to GDP ratio in excess of three percentage points in a couple of years are normally difficult to sustain, it seemed more important to consolidate and rationalize this already impressive increase. The priority objective, therefore, was to reduce tax distortions in production, trade and investment. Nevertheless, in the light of the country's persisting revenue requirements, it was decided early on in the analysis that the reformed structure should generate at least as much revenue as the current system. It was also specified that the reformed structure should be at least as equitable in practice as the existing structure, even though equity considerations were not the driving force in redesigning the tax rates and bases. Finally, given the co.ntry's limited administrative capacity, it was decided that the reformed structure should build rs much as possible on existing instruments. Malawi already had a tax called the "surtax" with a functioning "ring" system for removing .ax from transactions among registered domestic  - iv - produceri It was recommended that the "ring" be replaced by a system of "crediting". This would de facto create a VAT through the manufacturing stage and ensure that production distortions due to the Axation of inputs would be lowered. the next step was to map as much as possible of the revenue function of the new trade taxes into domestic taxes. This was to be accomplished by realigning the coverage and rate structure of the import surtax to be consistenc with zhe domestic surtax so that the former could also be incorporated within the crediting network. By this device, a proto-consumption tax would be created, at least through the manufacturing and import stage. The restructured surtax would also apply to large firms in agriculture and in wholesale/retail distribution. Exports were to be zero rated anc refunds made on the presentation of documentar, evidence arising from the crediting process. This would ensure that the tax structure did not interfere with export competitiveness as inputs used in exports would not be taxed. Finally, high tariff rates on luxury imports were to be merged into a second (luxury) rate for the surtax, thereby taxing the consumption of luxuries without providing protection for their production. This overall approach contained two important components: merging the revenue function of the newly instituted taxes on trade with the domestic production tax, and realigning the non-protective components of existing trade and production taxes to function as a rudimentary consumption tax. In the case of the company tax, the recommendation was to lower the level and dispersion of marginal effective tax rates as these can adversely affect investment decisions. It was initially not recommended that Malawi follow the trend in some developed economies towards reducing statutory tax rates as it was felt that this would create a revenue loss  -v (and a windrall to existing capital) even if no new investment materialized. Instead, as the intention of tax relief was to trade off revenue for new investment, it was recommended that existing investment credits and allowances be amalgamated into a single initial allowance at a higher rate to benefit new investments and that the average effective tax rate remain high for revenue reasons. Minor changes to pereonal income taxes were also recommended to simplify the system and make it easier to administer. Changes were proposed in tax administration to enable the mechanizatiox. and eventually the computerization of the system, including the redesign of some processes and forms. A system was also introduced to identify each taxpayer uniquely (whether an individual, a partnership, a company or a corporation) to facilitate auditing and cross-checking. Most of the recommendations have been successfully implemented over a three year period despite difficult economic circumstances in the country. One unusual feature has been the location of VAT administration in the Department of Customs an6 -'cises (which administered the old domestic and import surtaxes) rather than in the Income Tax Department as is customary elsewhere. This was done to ensure continuity and to minimize disruption in the administration of the surtax. The adoption of the taxpayer ID system, the introduction of the crediting system (including training of staff), the development of simple computer procedures for timely receipt estimation and the flagging of problems in tax accounts have all proceeded smoothly. The new system is now yielding mo-e revenue than the old system, even though that was not explicitly the goal of the reform program. However, since the budget deficit remains high and inflation continues to be a problem, further base exparnsion is necessary before the currently high  - vi - basic surtax rate and statutory company tax rate can be reduced. The two areas in which progress has been slow are: (i) the restructuring of the protective elements of the tax system, and (ii) the staffing of a smali new tax policy analysis unit in the Ministry of Finance. Establishing this unit was intended to create the ability within the country to identify problems with the tax systei as conditions change and to analyze the interaction of tax changes with changes ill other economic variables.  TAX REFORM IN HALAWI I. INTRODUCTION 1. This paper examines Malawi's recent efforts to reform 4ts tax system. Malawi embarked on a comprehensive tax reform with World Bank assistance in the latter half of the 1980s. At the government's request, the Bank played a prominent role in analyzing the weaknesses in Malawi's tax system in 1985, in making recommendations for change and in financing the implementation of many of the tax reform proposals accepted by the government in the period 1987 to 1990. This paper concentrates on the problems in the revenue system that provided the impetus to reform, the nature of the solutions that were offered, and the issues raised irh implementing the reforms. Since many of these reforms are still being introduced into Malawi's economy, their ultimate success will ha.e to await future analysis. The paper begins with an overview of Malawi's economy and tax system and discusses its evolution up to the time of the Bank's involvement in the mid-1980s. The reform agenda is outlined next and then in the third section the implementation of that agenda to date is discussed. The pap_- concludes with a brief set of lessons that can be learned from the Malawi experience. II. AN OVERVIEW OF MALAWI'S ECONOMY AND TAX SYSTEM PRIOR TO 1985 1. Salient Features of Malawi's Economv 2. Malawi achieved its independence from Great Britain in 1964. The country is small and landlocked, with a poor but relatively open economy. In the 1970s Malawi had one of the highest economic growth rates in developing countries. In the early 1980s per capita Income was estimated to be about US$180. In 1984 about 80 percent of Malawi's labor force was  -2- employed in the agricultural sector which accounted f r approximately 50 percent of GDP. Agriculture had a dualistic structure in that numerous smallholdings coexisted with a few large estates and commercial farms specializing in growing tea, tobacco and sugar for export. The smallholder sector grew maize, the major e.bsistence crop, and some crops for export such as tobacco, cotton, and groundnuts. In .984 exports comprised almost one-third of GDP while agricultural exports accounted for nearly half of the total exporL.4. During the 1970s Malawi also exported a significant fraction of its labor force to neighboring countries and workers' remittances contributed significantly to Malawi's foreign exchange earnings. By the mid-1980s, however, this flow had declined substantially. 2. Tax Structure up to 1978 3. After independence, Malawi inherited a tax system in which personal and company income taxes provided as much as 50 percent of total tax revenue. By any standard this represented a very high reliance on direct taxes in a non-mineral producing developing country in Sub-Saharan Africa. However, the tax base was very narrow. Most personal income tax revenue came from workers in the public sector and in large companies, while company income tax revenue was obtained mainly from a few large private firms active in primary processing and distribution. 4. The primary source of indirect tax revenues in the 19609 was custom duties. They c0iktributed around 40 percent of total tax revenue. The rate structure of trade tariffs was designed to promote import substitution of consumer goods and to discourage luxury consumption. Accordingly, the highest rates of duty were imposed on durable and luxury consumer goods while the tariff rates on capital goods and intermediate goods were low or nil.  -3- 5. To broaden its indirect tax base, Malawi introduced a sales tax, known a^ the surtax, in 1970. Initially this was applied at a five percent rate which was assessed on the sales price of domestic manufacturers and the duty paid value of imports. The surtax rate was increased to 10 percent in 1971 and to 15 percent in 1977. In the case of imports, an additional uplift factor of 1.2 was imposed on the base rate to place imports on a more even and competitive footing with domestic manufacturers (ostensibly to offset part of the currency overvaluation). Under the new surtax, capital goods were exempted and intermediate inputs were eligible for rebates. To avoid taxing transactions between producers, a "ring" system was introduced. Through this mechanism, those manufacturers who registered with the controller of customs and excise could qualify for a rebate of surtaxes paid on the purchase of intermediate inputs. The amount of the rebate varied according to the nature of the final commodity. If the intermediate product could be obtained from within Malawi, only a partial rebate of duty was available. 3. Revenue Crisis in the Period 1978-84 6. As shown in Table 1, the middle to late 1970s constituted the halcyon days in Malewi's economic development. Rates of economic growth and investment were high and were fueled by easy access to foreign lending. However, the early 1980s brought a serious reversal of economic fortunes. The global economic recession, the sherp decline in foreign private lending to developing countries and the abrupt termination of Malawi's rail access through Mozambique combined to produce a substantial slowdown in economic activity. The rail closure was a critical blow. As Malawi is a landlocked country, 80 to 90 percent of its exports and 60 to 70 percent of its imports ;Ad been transported by the rail line. The loss of this route  -4- Table 1 Macro-performance in Matawi. 1974-87 Const. 78 Prices Current Prices (1) ~~(2) (3) (4) (5) ReaL GDP Gross Domestic Gross Domesitc Implicit Foreign Inflation Year Growth Rate Investment/GDP Saving/GDP Savings/GDP Rate 1/ (X) (X) CX) CX) (X) (2) - (3) Pre-Study: 1973 20.4 12.4 8.0 1974 7.2 18.9 16.4 2.5 18.2 1975 6.1 24.9 17.0 7.9 8.3 1976 5.0 22.1 17.8 4.3 10.0 1977 4.9 22.2 20.1 2.1 12.5 1978 9.7 30.9 20.5 10.3 1.0 1979 4.4 27.1 12.7 14.4 2.6 1980 0.4 22.2 10.8 11.4 15.0 1981 -5.3 15.2 11.9 3.3 17.0 1982 2.5 14.6 14.9 -0.3 9.8 1983 3.7 13.7 15.2 -1.5 10.9 1984 5.4 13.0 15.4 -2.4 12.9 Post-Study: 1985 4.6 13.3 13.1 0.2 8.5 1986 -0.2 11.1 9.2 1.9 12.7 1987 1.1 12.1 13.2 -1.1 24.2 1988 4.8 13.3 8.0 5.3 24.4 1/ Annual change in impticit price deflator for GOP. Source: Volume 2, Tax Policy for Malawi, WorLd Bank, November, 1985. These data are drawn from Malawi NationaL Accounts Report, 1973-79 and Malawi, Economic Report, 1985. Post study data are drawn from Malawi Country Economic Memorandun (#8140-M%I). This source includes minor revisions to earLier data.  -5- dramatically increased the cost of producing goods in Malawi and of exporting them. 7. This series of external shocks acted to highlight a number of weaknesses in Malawi's fiscal system. As can be seen in Table 2 (col. 10), the servicing requirement on Malawi's accumulated debt put strong upward pressure on current expenditures. By 1981 debt servicing consumed about 34 percent of current expenditures (inclusive of amortization), more than double the rate of three to four years earlier (approximately 15 percent in 1977-1978). Although IMF loans provided some temporary financial relief, Malawi was soon forced into two consecutive debt reschedulings. At the same time, national defense spending grew as a result of the unrest in Mozambique. Declining overall economic activity and import capacity, therefore, coincided with rising expenditures to produce a strong demand for more government revenue. 8. As a result of this pressure on the revenue system, Malawi chose the administratively expedient option and raised tax rates on existing bases in virtually all areas of taxation but particularly in the realm of trade taxes. 4. Ad Hoc Revenue Measures in response to the revenue crisis 9. Increased taxation of domestic transactions: As early as 1978 the shift in Malawi's economic fortunes was reflected in its fiscal balances (Table 2, col. 8). The overall budget deficit increased dramatically by a third from eight to nine percent of GDP in 1976-77 to 12 to 14 percent of GDP in 1978-79. The government's fii-st response on the revenue side was to increase the domestic surtax rate by five percentage points overall in two steps: from 15 to 17 percent in 1979 and then to 20 percent in 1980. The import surtax rate moved up in tandem though still subject to the 1.2  Table 2 MaLawi: Central Goverrnent Fiscal Operations (Budgetary revenues and expenditure), 1974-88 a/ As a percent of GDP at market prices Debt Servicing (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Year Interest Non-interest Current Tax Revenue Non-Tax Current Current Overalt Interest/ Interest + Expenditure Current Expenditures Revenue Revenue Deficit Budget Current Amortization/ Expenditure 1 bI Deficit c/ Expenditure I Current a/ Expenditure II d/ (%) (X) (%) (X) (Z) t%) (X) (X) (X) (%) Pre-study: 1973 1.1 15.8 16.9 12.0 5.4 17.4 +0.5 7.9 6.4 17.0 1974 1.1 15.9 16.0 11.7 5.4 17.1 +1.1 7.8 7.1 18.8 1975 1.0 14.9 15.9 12.6 4.4 17.0 +1.0 12.2 6.4 17.2 1976 1.5 13.8 15.3 11.8 3.7 15.5 +0.2 7.7 10.0 16.8 1977 1.0 13.9 14.9 12.3 3.6 15.9 +1.1 8.5 0.7 16.0 1978 1.9 15.6 17.5 15.2 3.3 18.5 +1.0 12.4 10.6 14.4 1979 2.6 17.7 20.3 16.8 5.1 21.9 +1.6 13.9 12.9 21.1 1980 3.7 17.0 20.7 16.6 3.2 19.8 -0.8 15.8 18.0 27.4 1981 5.5 19.1 24.7 16.2 3.8 20.0 -4.7 15.5 22.4 34.1 1982 5.1 17.2 22.4 16.7 2.9 19.6 -2.7 12.5 22.9 27.9 1983 4.2 17.7 21.9 16.7 3.3 20.0 -1.9 10.1 19.3 26.0 1984 7.1 17.3 24.4 17.4 3.3 20.7 -3.7 8.8 28.9 37.8 Post-study: 1985 5.8 20.8 26.6 19.1 3.5 22.6 -4.0 8.7 21.7 35.8 1986 8.2 21.8 30.0 17.8 4.6 22.4 -7.5 13.9 27.2 37.9 1987 7.7 18.7 26.4 16.4 4.8 21.2 -5.3 9.9 29.3 40.2 1988 5.8 20.0 25.6 18.1 3.4 21.5 -4.1 7.6 22.0 39.0 a/ Budget data are in fiscal years (FY) and GDP in Calendar Years (CY) b/ Current Expenditure I Excludes Amortization. ci Budget Deficit before Grants d/ Current Expenditure II Includ.es Amortization. Source: Tax Poticy for Malawi, World Bank, November 1985. Post study data are drawn from MaLawi Country Economic Memorandum (#8140-MAI). This source includes minor revisions to eartier data  uplift factor (from 18 to 25 percent in 1979 and 30 percent in 1980). Excise rates (per unit of quantity) were also increased on domestic production. However, the domestic production base was not large enough to generate adequate additional revenue -- hence the shift in emphasis to taxes on external trade. 10. Increased taxation of all imports: Initially tariff rates on consumer imports were increased in the early 1980s. Then, to compensate partially for the overvaluation of the Kwacha and to discourage overall imports, the bovernment introduced a three percent import levy in 1981 as a "temporary" measure. This import levy was applied to the c.i.f.1 value of all merchandise imports and its rate was raised to four percent in 1982 and to five percent in 1983. As Table 3 suggests, by 1981/82 implicit average effective tax rates on private sector imports had nearly tripled while implicit average effective tax rates on domestic producers (primarily manufacturing) showed a much more modest rise. 11. Increased taxation of intermediate and capital imports: By 1983 imports were still too high for balance of payments purposes. Therefore, a foreign exchange licensing or allocation system was introduced to limit the importation of "non-essential" imports. This measure de facto favored public imports (in other words, imports by government, state owned enterprises [SOEs] and projects assisted by foreign aid) which were seen as essential for growth. As a result, the share of "essential" imports as a proportion of total imports grew from 12 percent in 1978 to 16 percent in 1981 and to approximately 20 percent in 1984. Since these imports were exempt from import duties and duties were highest on consumer goods imports I Cost, insurance and freight.  -8- Table 3 Malawi: Imnticit Indirect Tax Rates a/ (Percentage) Import Tax rates on Private Sector Domestic Tax Rates on Large Imnorts Manufacturers Commodity Class 1975-76 1981-82 Average Compound 1975-76 1981-82 Average Compound Annual Growth Rate Annual Growth Rate Consumption goods 27.5 40.6 6.6 8.2 10.6 4.32 Mixed use (Consumption or intermediate) 2.3 71.2 20.8 8.8 3.99 -12.36 Intermediate inputs 1.9 8.3 28.13 0.9 2.3 16.12 Capital goods 1.9 14.3 40.4 - - - Average 11.7 32.1 Source: Table 2.2, Tax Policy for Malawi, World Bank, November 1985. a/ Implicit tax rates are defined as the ratio of actual tax colLections from a particular base to the size of that base.  -9- (particularly luxuries), the negative revenue consequences of this policy forced the government to look for new revenue bases. This resulted in an increase in duties on the importation of intermediate and capital goods.2 12. In addition, in 1984 the base of the surtax was extended to include intermediate and capital goods that had previously been exempt. These were now taxed at a five percent rate. For those intermediate and capital 6ivds imports that had previously been subject to a lower surtax rate, the rate was increased from 20 to 25 percent. These changes accompanied a further increase in the basic surtax rate on domestic output in 1984 from 20 to 25 percent and on imports from 25 to 30 percent. 13. Explicit taxation of exports: With traditional tax bases subject to increasingly high tax rates (for example, the 30 percent basic surtax rate on imports), totally new bases had to be activated. In 1985 exports also became subject to explicit taxation. After the 10 percent devaluation of the Kwacha in 1985, the government introduced taxes on the major exports of tea and tobacco, ostensibly to absorb the windfall gains due to the devaluation and to ensure that the agricultural sector was subject to the same increasing rates of taxation as the manufacturing sector. 14. Increased taxation of income: Eventually increases in income taxes became unavoidable. In 1984 the company income tax rate was raised from 45 to 50 percent and, to broaden the base of personal income taxes, the personal exemptions, dependency allowances and a select set of other allowances were eliminated. 2 These had been introduced for the first time at very low rates in 1981. As a result effective tax rates on imported and domestically produced capital and intermediate goods rose at a much higher percentage rate than effective tax rates on consumer goods, as shown in Table 3.  - 13 - III. 1985 TAX STUDY AND TAX REFORM PROPOSALS 15. None of the tax and tariff changes in the early 1980s were a part of any long-run plan for the development of Malawi's revenue system. On the contrary, these changes evolved into a revenue system with a number of worrying features that could be ignored only as long as they were not binding. However, in 1984 and 1985 both the IMF and the World Bank recommended that the government create a more liberal economic environment by dismantling the foreign exchange allocation system and the prevailing system of administered prices, including a restructuring of SOEs. However, once the obstacles to a freely functioning price system were removed, the currently non-binding distortions of the new ad hoc revenue measures would become binding. 16. In the spring of 1984 a World Bank mission outlined some of the difficulties that the existing tax structure could pose for future growth of trade and investment. In the same year, another mission from the World Bank, this time dealing with a structural adjustment loan, expressed some concerns about the impact of existing taxes on production incentives in both agriculture and manufacturing. 17. By 1985 it had became increasingly clear to government officials -- following the Bank studies which had identified preliminary problems -- that recent revenue measures were not desirable in the long run and that reform of the overall tax system was necessary. The stage had been set for a restructuring of the Malawian tax system. 18. In January of 1985, the Government of Malawi asked the World Bank for assistance in reviewing its whole tax system (not just selected issues) and suggesting recommendations for reform. In response to this request, the Bank assembled a study team which went to Malawi to investigate its tax  - 11 - system. After an intensive period of study and extensive discussions with all of the major government agencies and the private sector, the team produced a two-volume report in November 1985.3 1. Potential Problems ArJsinR from the Ad Hoc Revenue Measures 19. The study identified a number of problems: - First, most tax instruments were serving multiple objectives. The surtax, primarily a revenue tax, had a number of built-in protective features not normally part of a sales tax, particularly a higher rate on imports (30 percent) relative to its domestic counterpart (25 percent) due to the 1.2 uplift factor (which was less justified after the Kwacha had been devalued) and the partial rebating of taxes on competitive imports of intermediates. On the other hand, import duties served as a major means of generating revenue and of influencing consumption behavior, in addition to their protective function for the prevailing industrialization strategy. - Second, the import levy, introduced as a proxy for devaluation, was being retained even after the devaluation. Even as a proxy it was not appropriate since non-merchandise imports were not subject to the import levy and exports were not subsidized by an equivalent rate. In addition, since intermediate and capital goods imports were subject to the import levy with no relief provided for exports, the import levy was increasing the cost of exports (a distortion in resource allocation which does not occur under a devaluation) at a time when both traditional and non- traditional export expansion needed to be encouraged. 3 C. Cbamley et al, "Tax Policy for Malawi", Washington, D.C., World Bank.  - 12 i Third, while the extension of import duties to purchases of capital and intermediate goods reduced effective rates of protection for many competitive imports, it also undesirably: (i) distorted incentives against exports, again at a time when there was a premium on expanding exports; (ii) created negative protection for the production of "essential" final goods whose imports were exempted from import taxes; and (iii) to the extent that the tariff measures existed for revenue raising rather than for protective reasons, transformed the indirect tax system into a set of production taxes that closely resembled a network of turnover taxes. Fourth, even though a duty drawback system existed, it was not an effective mechanism for compensating exporters. Malawi did make an effort to rebate duties paid by exports but the drawback system was narrow in scope. It provided tax relief for only about 30 exported products and then only in the case of ir.puts that were considered to be physically incorporated into the final product.4 Moreover, producers could not claim a rebate unless they exported their entire product since there was no mechanism to allow inputs to be pro-rated between domestic and export sales. Finally, if an exporter marketed his product through a distributor, he was ineligible for the drawback since he did not physically export the product and the distributor was not the entity that had paid the input taxes. Consequently, exports were being harmed by the growing trend towards the taxation of inputs used in production. Exports were also adversely affected by the expansion of the 4 To restrict abuse and unwarranted claims, for example, duties paid on an imported saw blade used to fashion furuiture for export would not be eligible for rebate as it was not incorporated in the product.  -13- surtax to intermediate and capital goods despite the ring or suspension system that was in place for manufacturers (though this mechanism was considerably more effective than the duty drawback). Fifth, explicit taxes on agricultural exports were making it more difficult for Malawi to compete in international markets. In fact, agricultural activities were caught in an increasingly tight fiscal squeeze because the new tax measures diminished the prices that producers received for their output while, at the same time, higher taxes on imported intermediates were raising the cost of their purchased inputs. As an attempt to capture the windfall arising from devaluation, the export tax was inappropriate since it drove a wedge between international and domestic prices and altered the long-term signal to producers who were basically price takers in the international market. As an attempt to proxy for income taxes on the agricultural sector, the export tax was appropriate since it overlooked the fact that smallholders were already subject to an implicit tax levied by the agricultural marketing board, ADMARC, (which did not show up in revenue accounts because it was not transferred to the Treasury) and that large agricultural estates were already subject to formal personal and company income taxes. Sixth, the ad hoc changes in trade and commodity taxes as well as in personal income taxes raised the possibility that the equity features of the tax system were being adversely affected. Seventh, the increase in taxation of capital imports and the increase in statutory company tax rates were both likely to have an adverse impact on investment.  - 14 - - Eighth, overall tax rates on those incomes and activities subject to tax were now at relatively high levels. All of the various tax instruments were applied to a relatively narrow tax base consisting of public sector employees, employees of large firms, the income of formal sector private firms, both domestic and foreign (but not SOEs), and traditional excisable products plus imports. It was unlikely that those incomes and transactions that were subject to explicit taxation in Malawi accounted for more than a third of GDP. Consequently, the tax to GDP ratio of roughly 17 percent in 1983 and 1984 translated into a tax of approximately 50 percent on the value added of the modern sectors of the economy.5 2. Framework of the Tax Reform 20. Objectives: The goals of the reform exercise were to be: (a) efficiency: to achieve a more neutral system of taxation, one that would interfere less with the efficient allocation of resources in production, trade and investment; (b) equity: to shift, if possible, a larger fraction of the overall tax burden from the poor to the rich; (c) administration: to identify instruments that would be administrable and to promote institutional changes that would improve the quality of the tax administration; and (d) revenue: to lay a stronger foundation for any future increases in total revenue, should this be necessary. 5 Where taxed sectors (such as the modern sector) interact with untaxed sectors (such as the informal or subsistence sectors) working out the ultimate incidence of taxation -- who carries the actual burden -- is problematic. However, even if through this interaction tax burdens were diffused throughout the economy, important distortions in resource allocations would remain as measured by differences in effective tax rates. Thus approximating the average tax rate on the taxed component of GDP, rather than on all of GDP, is potentially a useful measure. It suggests that if the average rates are very high there might be barriers to the further expansion of: (i) formal activities and transactions (since the latter are much more heavily taxed than the Informal sectors); and (ii) revenue (unless base expansion is feasible).  - 15 - 21. Constraints: The tax study recognized four constraints in its wide ranging packapge of recommended reforms for Malawi. First, it was recognized that revenue raising could not be a prime objective. From 1978 to 1984 Malawi had dramatically increased its tax to GDP ratio from 12 to 17 percent -- an unprecedented rise of five percentage points or almost 50 percent of tax revenue in 1978. Since rises in tax to GDP ratio in excess of three percentage points in a couple of years are normally difficult to sustain, it seemed more important to consolidate and ratiUnalize this already impt_ssive increase. On the other hand, even though lowering the tax to GDP ratio might reduce the tax induced inefficiencies in the economic allocation of resources, it .,as deemed necessary that the new recommendations as a whole should generate at least as much revenue as the existing system did. Failure in this area would probably result in the introduction of ad hoc tax measures to cope with revenue pressures, measures which might run counter to the study's major recommendations for reform. Thus, if a particular tax reform measure was expected to produce a loss in revenue, it was required that this loss be recouped in other features of the reformed system. Second, the proposals were constrained from reducing the de facto equity features of the tax structure, even though the focus of the reform would be on reducing tax induced distortions in production, trade and investment. A third constraint was that the reform proposal should not be so sophisticated as to overwhelm the capacity of the existing tax administration. The commitment to build on and, where desirable, modify existing tax instruments rather than to introduce new ones was made at an early stage in the reform process. Fourth, given the severe shortage of empirically estimated data, the reform would be based on judgements utilizing the best data available, rather than attempting to construct a quantitative model of the Malawi economy.  - 16- 22. Approach: In rractice it is the interaction of various tax instruments that affect attainment of objactives. The assignment of single instruments to each of the objectives is, in general, not possible. In the Malawi tax reform, however, an attempt was made to emphasize the use of different instruments for different objectives. This was done by clearly delineating the principal role of each instrument, though not necessarily to the exclusion of overlaps with other instruments. For example, import duties can be used to influence consumption, to provide protection and/or to generate revenue. The recommendation in the Malawi exercise, however, was that iwport duties be limited to a protective role rather than that they should serve as a major source of revenue. One suggestion was to limit the aggregate size of producer subsidies and to provide them in the form of short term duties to a selected set of goods which would change as industrial priorities changed. Such an approach would result in a very narrow tax base. Alternately, a broader import base could be utilized but with lower tariff rates. In either case, even though duties would clearly continue to generate revenue, the amount of revenue generated would be marginal relative to revenue generated from domestic taxes on the import base. By subjecting all competing as well as noncompeting imports to an "embryonic" consumption tax (see sections on surtax reform) revenue would continue to be collected at the point of import, to take advantage of the administrative convenience of this arrangement, but through domestic indirect taxes rather than import duties. This would remove much of the revenue function of import duties and shift it to an instrument which is equally convenient to collect but without introducing unnecessary distortions between imported and domestically produced goods. By also shifting luxury tariff rates to domestic indirect taxes, the import tariffs  - 17 - would no longer be used to influence the pattern of consumption. Thus, after a suitable rearrangement, the trade tariff system would be more or less focused on a single objective -- that of providing protection as deemed necessary by trade and industrial policy. 23. Similarly, the surtax could be made the primary revenue raiser within the system of indirect taxes by removing the protective features that were built into it. Thus, even though in general the interaction of instruments is crucial in evaluating the effects of the tax system on objectives, in the case of Malawi, the number of instruments in use were few and their effects were potentially separable. Hence, existing instruments could be redesigned to address specific objectives such as generating revenue, providing equity and providing protection. Whatever other effects they might have as by-products would be minor compared to their principal roles. 24. It is interesting to see how these reform principles were applied in the framing of specific proposals to improve the performance of the major tax instruments in Malawi. At no point in the exercise did the tax study team take the position that there was only one correct answer to a problem. This made it possible to explore ideas that initially were not considered feasible. It is important to emphasize that the focus was on the reform of the existing tax system rather than on designing a new one from scratch. In other words, every attempt was to be made to build on existing instruments and to take advantage of the familiarity that tax administrators already had with them. The instruments that were retained in the reformed system were those with many, but not all, of the desirable characteristics of the best tax for a given objective, provided there was evidence that each of these instruments had generated an increasing amount  - 18 - of revenue over time. The fact that increasing amounts of revenue were being collected through a sophisticated instrument with desirable characteristics was taken as prima facie evidence that it was an administrable instrument. Modifications were then introduced into these selected instruments (to expand the range of their desirable characteristics) in preference to introducing new instruments whose administrative feasibility was not known. This can be illustrated by the way in which the reform of the surtax was determined. 3. Taxes on Goods and Services 25. Shifting from a "ring" system to a credit system in the surtax: The study team initially recor-ended the introduction of a value added tax (VAT) in Malawi. There was tremendous resistance to the recommendation from both the Department of Income Taxes and the Department of Customs and Excise Duties because the VAT was thought to be too complicated to administer. Therefore, the idea was shelved at the time the team commenced field work. But when the study team did a detailed assessment of the existing "surtax" procedures, it found that the paper work for tha prevailing "ring" system was the same as would be required in the case of a VAT but in reverse. Malawi had adopted the ring (or suspension) method of commodity taxation in an effort to control the degree of tax cascading uue to the application of the surtax to purchases of business inputs. Under the ring system, the surtax was applied only to sales made to purchasers outside the ring. Transactions among registered manufacturers were exempt from application of the surtax. Thus, in order to qualify for an exemption, the seller had to request documentation from the buyer that proved that he was "registered". A reversal of procedures, with the vendor indicating to the purchaser the amount of taxes for which he was being  - 19 - assessed, would be sufficient to transform the existing surtax into a tax similar to the VAT with more or less the same amount of paper work. 26. In addition, introducing a crediting system within the "surtax" framework in lieu of the existing system of exemptions could actually make the surtax less complicated to administer. Although the ring system was sound in principle, in practice it was plagued by a number of problems. First, the system could only imperfectly remove the tax on business inputs since it could do nothing to prevent the re-entry of taxed products into the ring when a registered manufacturer purchased an item at either the wholesale or retail level. Second, to avoid revenue leakages and diversions of goods to non-taxed uses, considerable administrative resources were required to keep track of goods purchased by exempted producers. Audits of inventories and bonded warehouses were necessary to ensure that purchasers who were exempt from the surtax did not sell to parties whose sales were not taxed (for example to the services or the informal sectors). There was growing evidence of substantial leakages around the ring despite major efforts to police its boundaries. Out of a total of 226 registered firms, only 10 -- those also subject to excise taxation -- accounted for well over 80 percent of the revenues from the surtax, way out of proportion to their representation in the value added produced by the 226 firms. 27. It was in this setting that the study team recommended the replacement of the exemption approach with a tax crediting mechanism. Although it was not called a VAT, it made use of the principle of a consumption based VAT. All sales would become taxable, with the exception of export sales, and all sellers could claim credit for taxes paid on the purchase of inputs. This would ensure that production distortions due to  - 20 - the taxation of inputs would be minimized. The administrative advantage was that there would no longer be any need to distinguish between final and intermediate goods or between tax exempt and tax paying producers. The advantage in terms of revenue would be that taxes on final sales would be collected in increments at earlier stages rather than being collected in total at the end as in the ring system. 28. Thus, although the government initially balked at introducing a VAT on the grounds that it was administratively too complicated, the study team was successful in demonstrating the feasibility of a tax similar to the VAT. They were able to do this by showing that the amount of paperwork required for the current ring system was more or less the same as that which would be needed for a VAT and that the self enforcement aspects of the crediting system were superior to those of the ring system. 29. Realigning trade and domestic taxes: The next step was to map as much as possible the revenue function of the trade taxes into domestic taxes. This was to be accomplished by realigning the coverage and rate structure of the import surtax to be consistent with the domestic surtax so that the former could also be incorporated within the crediting network. The uplift factor of import surtax would be eliminated and the basic surtax rate would be reset to compensate for this elimination. All formal domestic production, as well as competing and non-competing imports, would be subject to the same basic surtax rate and crediting system. By this device, a proto-consumption tax would be created, at least through the manufacturing and import stage. Exports were to be zero rated and refunds made on the presentation of documentary evidence arising from the crediting process. This would ensure that, because the inputs used in exports would not be taxed, the tax structure would not interfere with export competitiveness.  - 21 - 30. Other efficiency enhancing measures included merging the import levy with import duties and the high tariff rates on luxury imports into a second (luxury) rate f or the surtax. The latter change would ensure that consumption of luxuries would be taxed without providing protection for their production. The protective structure of import duties could then be rationalized gradually. However, because of the existence of the foreign exchange allocation system, this could not be an immediate priority. In addition, it was recommended that a better and smoother functioning system of duty drawbacks for exports be established, and that the export duties that had been imposed in 1985 be removed. 31. Expanding the surtax base: The high surtax rate in Malawi reflected the narrowness of the tax base, even inclusive of imports. It was also a strong inducement for evasion. The study team, therefore, proposed that the size of the surtax base be expanded to make possible the eventual reduction of the tax rate. One recommendation was to extend the coverage of the surtax base to include the large establishments in the wholesale and retail sectors. It was anticipated that such a move would increase the share of total consumption subject to the surtax from 30 to 46 percent. Another recommendation was to include marginal manufacturing, in the form of packaging, bottling and assembly of components, within the definition of formal manufacturing activity. Mcreover, the sales value of manufacturing goods was to be defined on a comprehensive basis to include the sales of any ancillary services, such as warranties, transportation and the provision of finance. This broader definition of sales value would limit the scope of manufacturers to undervalue the sale of goods and overvalue the sale of services. Finally, in the case of manufacturers who operated their own wholesale and retail activities, the definition of  - 22 - manufacturing would be extended to include them and thus to remove any temptation to engage in tax minimizing transfer pricing. 32. The government was also urged to adopt a tax inclusive budget which would add equally to both government expenditures and revenues but which would eliminate any tax induced bias in purchases made by government agencies and bureaus. In the absence of a tax inclusive budget, the government and many public entities were exempted from paying import duties and other taxes with protective features such as the surtax. This exemption encouraged the public to purchase imports rather than domestic goods. This preference was not desirable. The adoption of a tax inclusive budget would also remove the emerging tendency on the part of some public agencies to engage in arbitrage by purchasing goods duty and tax free and reselling them to the private sector. 33. Equity and the rate structure of the reformed surtax: Table 4 displeys estimates of the incidence of both commodity and income taxes in 1983. The numbers suggest that, with the exception of the lowest income group, the incidence of commodity taxes6 was more or less proportional. Not only were commodity taxes proportional in their incidence at the upper end but rough calculations suggested that commodity tax rate increases after 1983 had hit the poor the hardest. This was in sharp contrast to the progressivity of income taxes. High income households paid nearly as much in income taxes as they did in commodity taxes. Low income households, however, paid very little in income taxes. Thus it was income taxes rather than indirect taxes that resulted in high income households having a total 6 Taxec on goods and services.  - 23 - tax burden nearly three times as large as that imposed on the poorest households in the economy.7 34. To address the lack of progressivity in indirect taxes, the study recommended that equity features be added to the embryonic consumption tax. This required that existing excises be redesigned to complement the restructured surtax. The redesign involved four basic changes: first, the valuation basis of excise (and import) duty rates was to bc~ shifted from specific (per unit of quantity) to ad valorem (per unit of value). Between 1978 and 19L'. the ad valorem equivalent rates on excisable products had dropped by half with a commensurate decline in potential revenue. This drop in ad valorem equivalent rates occurred desp:tte low rates of inflation and repeated increases in the specific rates. If ad valorem rates had been in force, revenue from excises could have grown without recourse to rate changes. The ad hoc increases in specific rates gave the appearance of increasing tax burdens which antagonized domestic producers without maintaining revenue yields. The shift to ad valorem rates on excisable products was judged to be feasible as the same products were already subject to the ad valorem domestic surtax. The second recommended change was that all ad valorem excises be collapsed into two or three rates that could be merged into the surtax structure as additional rates. These rates would be higher than the basic surtax rate. As they would apply primarily to final goods, the presence of differential rates would not complicate compliance with the surtax crediting system.8 Commodities would then be 7 It should be noted that within income taxes it was the exemption from taxation rather than the progressivity of the rate structare which contributed most to the progressivity of the incidence of the tax burdens. 8 If experience showed this not to be feasible, the higher surtax rates would be eliminated and replaced by equivalent ad valorem excises. This would remove goods taxed at a higher rate from the crediting system.  - 24 - classified into three or four groups according to their income elasticity. Those commodities consumed primarily by the poor would be exempted while those with the highest income elasticities would be subject to the higher surtax rates. Third, traditional excisable products (such as alcohol and tobacco) would also be reclassified and the lowest quality products would be subject to the basic surtax rate while the higher and highest quality products would be subject to the two progressive rates respectively. Fourth, the luxury import tariff rates would be replaced by the new surtax rates. If this proposal were to be adopted, the incidence of the new regime of commodity taxation would be more progressive as indicated in Table 4 (in parentheses). The higher surtax rates (or ad valorem excises) would play the primary role in providing greater tax equity within the indirect tax system.9 35. In summary, the proposed reforms would shift more of the indirect taxes from a production and trade base to a consumption base. A crediting mechanism vould be introduced into the operation of the surtax to remove the tax on business inputs and to eliminate tax cascading. Exports would be zero rated. Imports and domestic output would both be taxed at a common rate under the surtax. The issue of equity would be dealt with by adding one or two luxury rates to the surtax. There would be much less reliance on tariffs for revenue purposes. Lowering tariff rates on imported final goods rather than raising tariffs on imported intermediate and capital goods would reduce levels of both nominal and effective protection.10 9 Separately, the study also recommended the reinstatement of personal exemptions (or a zero rate threshold) and the removal of some loopholes in the personal income tax to address equity concerns in the direct tax system. 10 In the short run, the administered foreign exchange allocation system determined the domestic price of imports in the economy, as a result of whxch neithsr the level nor structure of tariff rates was binding and the prevailing tariffs only served to soak up the economic rents created by the rationing of foreign exchange.  - 25 - Table 4 The Incidence of Taxes in Malawi (1983) Percentage of Household Expenditures Commodity Income Total Income Group (Kwachas) Taxes a/ Taxes Taxes High (K > 4,800) 13.8 (17.9) 14.4 28.2 Medium (K 1,200-4,799) 14.3 (17.5) 4.5 18.8 Low (K 480-1,199) 13.1 (12.8) 1.0 14.1 Lowest (K < 480) 9.1 ( 8.8) 1.0 10.1 Source: Table 2.6, Tax Policy in Malawi, World Bank, November 1986. a/ The column in parentheses under Commodity Taxes indicates the anticipated pattern of tax burdens under the proposed commodity tax reforms.  - 26 - This overall approach in Malawi realigned the non-protective components of existing trade and production taxes to function as a rudimentary consumption tax, and merged the revenue function of the newly instituted taxes on trade with the new domestic tax. 4. Companv Income Taxes: 36. In the early 1980s company taxes generated a quarter of total tax revenues. The company tax in Malawi sensibly did not draw a distinction between corporations and other forms of business organization, thus avoiding a distortion that is found in many other countries, both developed and developing. Table 5 presents some of the main features of the company tax in Malawi before 1985. The statutory company tax rate was increased from 40 to 45 percent in 1975 and to 50 percent in 1981. Average effective tax rates, which reflect the revenue consequences of the tax structure, were significantly lower, although they too rose from 24 percent in 1975 to around 39 percent in 1984. Marginal effective tax rates, which reflect the incentives for allocating investment, were also lower than the nominal or statutory tax rate but higher than the average effective tax rate. The relationship between the average and marginal effective tax rates was opposite that normally desired. In general it is preferable to have a marginal effective tax rate which is lower than the average effective tax rate as the objective is not to lower the level of investment but to ensure that revenue is collected from economic profits (that is profits above that required to induce the marginal investor to invest). 37. The marginal effective tax rates were lower than the nominal rates for a couple of reasons. Prior to the tax reform, companies were allowed to pay taxes with a lag of at least one year. This had the result of lowering the marginal effective tax rate. Most of the disparity between  - 27 - the nominal and the marginal effective rates, however, was due to a generous set of investment incentives. 38. In addition, the investment incentives varied by sector. Manufacturing, unlike other sectors, was entitled to a 10 percent investment credit which did not reduce the depreciable basis. Manufacturers could also claim a 10 to 20 percent initial investment allowance which acted as a form of accelerated depreciation and reduced the depreciable basis. The uneven sectoral impact of the company tax provisions is indicated in Table 5 by the wide differences in marginal effective rates of taxation between the manufacturing and non-manufacturing sectors in Malawi.11 New investors in manufacturing faced an effective tax rate of 31 percent in 1974 and 43 percent in 1984 compared to rates of 44 percent and 58 percent in the non-manufacturing sectors for the comparable periods.12 39. In its appraisal of the company tax system in Malawi, the study team was concerned about the intersectoral and interasset biases present in the system. To remove these non-neutralities, the study initially recommended a move to a cash flow tax with no allowance for interest deductions. Such a tax would result in a zero marginal effective tax rate and a positive average effective tax rate. However, since revenue losses in the initial post reform period would be too high, the proposal was shelved despite some of its attractive features that would have simplified income tax administration. Instead the focus shifted to reducing the level 11 The calculations are based on the assumption of a debt-equity ratio of five percent, an inflation rate of eight percent and a nominal interest rate of 15 percent. 12 What is not shewn in the table is the tendency of the investment allowance to reduce even further the marginal effective tax rate on assets with a shorter life.  - 28 - Table 5 Companv Tax Rates Rate (Z) 1975 1984 Proposed Statutory tax rate: 45 50 50 Average Effective tax rate: 24 39 Marginal Effective tax rate: Manufacturing 31 43 39 (long-lived assets) Non-manufacturing 44 58 39 (long-lived assets) Source: Tax Policy for Malawi, World Bank, Washington, D.C., November 1985.  - 29 - and dispersion of marginal effective tax rates. It was not recommended in the study that Malawi follow the trend in some developed economies toward reducing statutory tax rates as it would not only lower the marginal effective tax rate but also the average effective tax rates. The latter reduction could creace a revenue loss (and a windfall to existing capital) even if nc new investment materialized. The loss of revenue would be too high to be offset by tax increases elsewhere in the system. Instead, since the intention of tax relief was to trade off revenue for new investment, it was recommended that the marginal effective tax rate be lowered by amalgamating existing investment credits and allowances into a single initial allowance at a higher rate (40 percent) to benefit new investments and that this initial allowance should be available to investments in all sectors.13 it was also recommended that the average effective tax rate remain high for revenue reasons. Although the final proposal fell short of complete expensing, it would have the desirable effect of encouraging aggregate investment by reducing the marginal effective tax rate without losing substantial amounts of revenue. 40. As a further step to avoid tax induced inefficiencies in resource allocation, the study urged the government to subject state-owned enterprises to the company tax. It also recommended that company income tax payments be moved forward to a current year basis in conformity with the treatment of personal income taxes. 41. Finally, it was felt that small busineeses required special measures. Because of inadequate record keeping, many small businesses were 13 Over time, agitation from investors in other sectors had already led to the broadening of the definition of "manufacturing" to include hotels and agricultural processing, but in an ad hoc manner. For example, non-manufacturing related buildings, such as warehouses and offices, were not eligible for tax allowances in any sector, even in manufacturing.  - 30 - able to shift a number of personal expenses into the category of business income deductions. As a result, businesses with significant amounts of turnover were able to report trivial amounts of taxable business income. To curb evasion on the part of small businesses, the study recommended that Malawi investigate the possibility of applying some form of presumptive tax to this group of taxpayers based on objective and publicly known criteria. Taxpayers subject to presumptive taxation were to be given the option of belonging to the regular tax system that applied to personal and company incomes. 5. Personal Income Taxes: 42. Prior to 1983 Malawi had a two tiered system for taxing personal incomes. Under the taxable income method, neither deductions nor exemptions were allowed, while under the chargeable income method both exemptions and deductions were permitted. Taxpayers were required to calculate their liabilities under both approaches, and then pay the larger of the two taxes. After 1983 only the chargeable income method was allowed, although at the same time personal exemptions were eliminated. Marginal personal income tax rates were progressive and ranged between three and 50 percent. Only full time workers employed in the formal sector of the economy and some self employed businessmen and professionals were subject to the personal income tax. For most workers, including government workers, the tax was withheld at the source through the PAYE system.14 This withholding arrangement reduced the cost of collection and ensured that taxes on current income were paid without a lag. An end year reconciliation of tax obligations was required only from those workers who also had non-wage income. 14 PAYE denotes pay-as-you-earn, a familiar withholding mechanism in many formcr British colonies.  - 31 - 43. Two other kinds of personal taxes were also assessed. Low wage workers in the urban economy were subject to a two tier graduated tax and males over the age of 18 who were not subject to any other type of income tax had to pay a nominal poll tax. These two taxes, however, were not important sources of revenue. Together they contributed only about 12 percent of total personal income tax collections which, in turn, constituted only 16 or 17 percent of total tax collections. As a result, it was estimated that as few as 10 percent of urban households were responsible .jr as much as 80 percent of the roenues collected through all types of personal income taxes. 44. The tax study made several recommendations aimed at broadening the tax base by closing loopholes and making the personal income tax fairer. To remove the burden of the tax from low income households, the study recommended either that the absence of personal exemptions be compensated for by introducing a zero rate income bracket or that the personal exemptions be reinstated. To broaden the base, the study recommended repealing the interest deduction on housing mortgage loans as housing did not generate any taxable income. The study also recommended repealing the exemption on interest income earned at Post Office Savings Banks even though in inflationary conditions taxing interest income is equivalent to taxing capital. The rationale for this recommendation was that this exemption was originally intended to encourage greater savings on the part of low income households but was increasingly being used by corporations for purposes of tax arbitrage. It was also recommended that greater use be made of creditable taxes on repatriated earnings from Malawi. Thus, a 10 percent withholding tax on all forms of remittances to foreigners, including interest, dividends, rents and royalties, was suggested to  - 32 - replace the prevailing non-creditable sop-up tax on dividends paid to foreigners. 45. Realizing that the valuation of fringe benefits for tax purposes was extremely tricky and that their non-taxation was leading to abuse, as was evident in the proliferation of non-taxed fringe benefits, the study team recommended that fringe benefits be taxed indirectly by disallowing them as a company tax deduction. This was to be made more effective by aligning the nominal company tax rate with the top bracket rate under the personal income tax.15 46. For a low income economy, Malawi had a surprisingly sophisticated method for integrating the taxation of company and personal incomes. An imputation or dividend gross up based on the nominal company tax rate was used. This approach, however, could result in a personal income tax rebate even if there was no company tax paid or if the tax paid was lower than the nominal tax rate. The tax study, therefore, recommended that the procedure for grossing up dividend income be modified so that any personal income tax refund would be based only on the amount of company taxes that had actually been paid. Although no capital gains tax existed, the tax study team suggested that a follow-up study be made of how such a tax might work satisfactorily in the context of Malawi's economy. 47. In summary, sn the area of direct taxation, the study suggested that the company income tax be reformed to reduce tax disincentives to new investment and to reduce the dispersion of marginal effective tax rates 15 The equivalence between taxing fringe benefits at the personal level and disallowing a deduction at the company level breaks down if effective marginal company tax rates are less than the nominal rate, or if the company is tax exempt.  - 33 - across different sectors in the economy. It was also suggested that non- neutralities in the persoinal income tax that arise from asset acquisition should be reduced by repealing the incentives for investing in housing over investing in other types of assets. Finally, a few changes in the personal income tax were recommended to broaden the base and ensure greater fairness. 6. Tax Administration 48. On -the whole, tax administration in Malawi, as exercised by the Department of Customs and Excises and the Department of Income Tax, was highly regarded by the tax study team. However, major weaknesses were identified in the area of generating and utilizing information on taxpayers and their tax payments. To overcome these deficiencies and thus improve management, the study made several recommendations. The most important of these was that the information system should be redesigned. In particular, it was recommended that a document control system be introduced and that data should be organized in such a way as to allow for the identification of trends and patterns in, for example, receipts, arrears and the settlement of delinquencies. This would increase the ease and speed with which management would be able to analyze changing conditions and to react to them appropriately. 49. Toward this end, it was recommended that a master tax file be created with numbers that uniquely identified all individual and company taxpayers. The master tax file would facilitate the cross checking of sales and income tax records (subject to confidentiality rules) and would provide an up to date account of any taxpayer as well as of the performance of the entire tax system. The size of the tax base could also be expanded if the possession of a taxpayer identification number were to become a  - 34 - prerequisite for receiving government contracts or for obtaining employment or access to credit. 50. To help the Malawian government to analyze the interaction of the tax system with the rest of the economy, the study team also recommended that a special tax analysis unit be established in the Ministry of Finance. This unit would not make tax policy but would provide policymakers with professional advice on how changes in other policy areas mignt affect the tax system and vice versa. It could also be used to monitor the performance of the current tax system and to evaluate proposals to change the system. 7. Revenue Impact 51. Despite the continuing need for more revenue, a decision was made early in the process not to strive for immediate increases in revenue. Instead it was required that the reformed system should generate at least as much revenue as the existing system by consolidating the revenue increases generated by the previous ad hoc changes, and, if possible, be more elastic in the long run. The reform proposals attempted to balance any revenue losses with offsetting revenue gains. Some measures, such as the more generous provision of initial allowances and the removal of taxes from business inputs, would result in significant reductions in total revenue. Offsetting these declines would be the higher revenues resulting from the broadening of the surtax and of income tax bases, from the adoption of ad valorem excise taxes (instead of per unit levies) and from the conversion of the company income tax to an estimated or current year payment basis. Moreover, while the revenues previously collected from imports of capital and intermediate goods would drop, this would be offset by a gain in revenue from the uncovered sectors of the economy where the  - 35 - average tax rate on many inputs would rise from about 14 percent to the basic surtax rate of 30 percent.16 The long run elasticity of the revenue system was also expected to increase with the shift to ad valorem excise and import duties and with the greater progressivity of the indirect tax system. Finally, once the surtax base was purged of many of its distortionary features, the surtax rate could be used as the principal means for addressing revenue requirements. In fact, the surtax has already begun to play this important revenue role. IV. IMPLEMENTATION OF THE TAX REFORM PROPOSALS 1. Handover from Study Team to Technical Assistance Team 52. To ensure support for the reforms, the Ministry of Finance had invited reactions from interested parties (both public and private) from the inception of the study. This was undertaken with a view to incorporating their concerns into the design of the new tax instruments. A preliminary report was presented to the Ministry of Finance in October 1985 and discussed with the Treasury and the two tax departments. A revised draft of the report was circulated more widely to other ministries and departments in November 1985. In January 1986, after extensive internal discussions, the government agreed in principle to most of the recommendations in the study. 53. Over time, the study team had managed to establish a solid working relationship with senior government officials and this partnership helped 16 Even though in principle the taxing of inputs is no more desirable for the informal sector than for the formal sector, in practice they could not be avoided given the existing administrative constraints. However, there was a silver lining. The higher tax rate imposed on inputs of the informal sector would not only expand the tax base immediately, by incorporating a group not previously subject to tax, but also in the long run if it encouraged more producers to enter the crediting network. The crediting would benefit them if they were in competition with other firms whose sales were credited in the reformed surtax, yet the tax on their output would generate more revenue than the tax on their inputs. For this to happen, however, barriers to registering and entering the crediting system would have to be removed or at least reduced.  - 36 - to create a sense of the reforms being essentially home-grown rather than something that was being foisted upon reluctant government officials. On the contrary, the fact that the government had initiated and participated actively in the exercise was a factor in its ultimate success. Confidence grew to the point where the government wanted the study team to carry out the implementation of the reforms instead of having to forge a new relationship with an unknown group. This was not feasible as a different set of technical skills (for example, legal, accounting and administrative) were required to implement the proposals. 54. In the end a compromise solution wae found. A member of the study team was appointed to direct the project because of his detailed familiarity with the intent and content of the proposals. The Harvard Institute of International Development won the contract to provide the institutional support for implementing the reform program. The implementation phase was to be financed jointly by the Bank and the Government of Malawi. The entire exercise was assisted by the continuity of key officials and by the fact that within the government a high level tax coordinating committee was set up to oversee the implementation of the reforms. The strong support of senior officials has played an invaluable role in putting the reform into operation. Perhaps the most crucial was the involvement and support of the Permanent Secretary of the Ministry of Finance who initiated the review and provided guidance and support through two thirds of the implementation phase. 55. Many of the central proposals for reform have already been implemented.17 Other proposals are either still under discussion or are in the process of being implemented. As in the study phase, the Government of Malawi has actively participated in working out the details for 17 Technical Assistance Team, Memoranda to Malawi Tax Reform Steering Group, 1987-1990.  - 37 - implementing the proposals. For the reform program to be implemanted in full and to be given a chance to work, it is important that supporters should stay the course and not give in to pressure to undermine the reforms. 2. Reform of Taxes on Goods and Services 56. Reform of the surtax: This is the linchpin of the reform of indirect taxes. It was expected that the most difficult part of this reform would be the introduction of the crediting mechanism. In the end, however, because the administrative features of the prevailing surtax had been analyzed so thoroughly during the study phase, it was possible to revise forms and procedures and to train staff in a relatively short period with assistance from former officials of the U.K. VAT administration. Within the first year of the implementation phase, it was possible, in the April 1988 budget, to announce the introduction of an operational surtax crediting mechanism. To cushion its adverse revenue effect, the government planned to phase in gradually the crediting of the tax on purchases of capital goods and spare parts. The crediting mechanism has performed well above expectations. Penalties and interest charges on late payments are being enforced. A danger has appeared on the horizon, however, with the courts overturning on appeal some of the decisions of the surtax office. If upheld, these judicial rulings could in time undermine taxpayers' willingness to pay their surtax obligations on time. 57. Reforms of other domestic and trade taxes: Other elements of the indirect tax reform that have been successfully completed are: (i) the merging of the non-protective aspects of the import duty and the import levy into the import surtax; (ii) the transfer of luxury import duties to a  - 38 - second rate for the surtax; (iii) the conversion of specific rates for import and excise duties to ad valorem rates; and (iv) the elimination of export taxes. 58. Some elements are still in process (though approved in principle). These include: (i) eliminating the uplift factor to create a single surtax rate; (ii) expanding the surtax base to cover the consumption of telephone and electricity services, as well as purchases made in restaurants and hotels; and (iii) the development of a feasible and effective duty drawback scheme. 3. Reform of Taxes on Income 59. Reform of company income taxes: Income tax reform has been implemented with assistance from the Technical Assistance advisory service of the U.S. Internal Revenue Service. The base of the company tax has been expanded in two ways. In 1986 a withholding tax on sales made by firms in the tobacco and transportation sector was introduced, and in 1987 state owned enterprises were included within the company tax framework. In 1988 all firms were put on an estimated payment system so that taxes become due immediately after income is earned rather than with a one or two year time lag as in the past. 60. The collapsing of all investment related allowances and credits into an initial allowance of 40 percent began in April 1988. The new provision applies to those assets previously eligible for the smaller initial allowance. It is to be extended in stages to other assets. A problem on the horizon is the possibility that a new investment code will be adopted with extensive provisions for fiscal incentives. There will be little justification for additional fiscal incentives to encourage investment or export once the surtax crediting system is expanded to all  - 39 _ imports and domestic products, the reformed duty drawback system is implemented and the 40 percent initial allowance is extended to other assets. 61. Some changes that still remain to be made are: (i) legislation to allow small businesses to be taxed on a presumptive basis; (ii) the adoption of more appropriate depreciation rates; and (iii) the revision of the imputation system. 62. Reform of personel income taxes: Implementation of the proposals for personal ncome tax reform has proceeded more slowly. So far, all that has been achieved in full is that the base has been expanded by including interest paid on government bonds, as well as dividends and certain other kinds of interest income, in the withholding system. 63. Most of the reforms are still being worked out, including: (i) the restoration of the personal exemption or, alternatively, the use of a zero bracket; (ii) the repeal of both the interest exemption on Post Office Savings Accounts and the interest deduction for home purchase loans, especially since the latter is used primarily by upper income households; (iii) the estimation of dividend gross ups to reflect actual t&axes paid; (iv) higher imputation values for fringe benefits; and (v) the conversion of the graduated tax into a flat three percent income levy. 4. Reform of Tax Administration 64. Many of the administrative components of the reform package have a'ready found their way into the tax system. The most important of these were the issuing of taxpayer identification numbers in December 1988, the construction of a master tax file (to be completed in 1990) and computerization, staff training, the redesign of forms and documents, and the introduction of a document control system (also to be completed in  - 40 - 1990). There are signs, however, that an important proposal -- the adoption of a tax inclusive budget -- may be shelved by the government as a result of numerous objections received from several government departments. 5. New Issues 65. After the dramatic rise in tax to GDP ratio in 1978 to 1984, it was unrealistic to expect taxes to increase revenue further in the immediate post-reform period. Nevertheless, tax revenue in this period has in fact risen (on average by 1.5 to two percentage points) in part due to inflation. However, this has not been enough to prevent the budget deficit from worsening. Current expenditures are at least four percentage points of GDP higher than before the study. As a result, with external flows still low, inflation rates have doubled to 25 percent per annum. This has the potential of undermining several of the earlier reforms. In the case of the company tax inflation is causing the effective tax rate to rise18 due to the erosion of the real value of depreciation allowances and the inclusion of capital gains from inventories in the measurement of taxable income. For the moment, the absence of indexation has been partially offset by provisions that are already in place for revaluing assets upon devaluation of the Kwacha. In the case of personal income taxes, inflation has contributed to significant bracket creep under the PAYE system. Bracket amounts have been fixed in nominal terms since 1985 while the consumer price index has risen well over 50 percent between 1985 and 1988. Pressures to index the rate brackets and other nominal magnitudes in the personal income tax are growing steadily but have been resisted so far for revenue reasons. The Technical Assistance team recently proposed19 that the statutory company income tax rates be returned to its earlier rate of 18 Given the relatively unleveraged capital structure of most companies in Malawi. 19 Technical Assistance Team, Memoranda, November 1989.  - 41 - 45 percent and that the top marginal personal income tax rate be lowered to 45 percent to avoid income shifting. Unfortunately, these proposals are being implemented without simultaneously implementing the concomitant proposals for base expansion. As a result, the revenue from the explicit tax system is being traded off against the inflation tax which can have an adverse effect on the efficiency and equity features of the tax system. V. SOME LESSONS FROM TAX REFORM IN MALAWI 66. Although the experiment with tax reform in Malawi is recent and unfinished, it points to some more general themes: 1. Reforms can sometimes be introduced more easily in a quasi- crisis situation as the shortcomings of the system may be more evident revealing important structural weaku4,*bses. For example, the existing tax system may not prove to be a good foundation for raising more revenue. The need for more revenue can thus act as an impetus for the rationalization of tax bases and rates by providing a justification for the political and administrative costs that have to be incurred to bring about change. 2. Successful tax reforms require a high level of government commitment and continuity. In Malawi, a crucial aspect of the success of the reform program so far has been that the government has been supportive of the project from the start. Senior officials and key departments have been deeply involved in the selection of reform options and their implementation. Clearly this consultative approach can be time consuming and resource intensive, but in the long run it is more effective.  - 42 - It fosters commitment to the program and creates the political vill to overcome obstacles as the reform unfolds. The participation of high level government officials may well be an indispensable element in carrying out a reform of the tax system, as they provide the continuity and thus stability without which the program might falter. By the same token the departure of key officials can sometimes stall the whole process of reform. 3. Tax structures can be rationalized even where data is limited. Making different tax instruments serve distinct and separate functions can add a clarity of conception to the reform process that makes it much easier to redesign instruments and to defend the proposed changes. In Malawi, the restructuring of the surtax to become the revenue workhorse in the tax system, the redesign of excise taxes to enhance the equity characteristics of taxation and the assignment of tariffs to serve the goal of protection were, and remain, important guiding principles of the tax reform process. 4. It is feasible to introduce tax reform in poor countries with limited administrative capacity and a limited tax base, even though the limitations can severely constrain reform oprions. Tax reforms may be easier to implement if they build on existing instruments rather than create entirely new ones. If an instrument has many desirable characteristics and has generated significant amounts of revenue in the past, it should be treated as the core of a reformed instrument as it has proven to be administrable. In Malawi, it was much easier  - 43 - to fashion a surtax that embodied many features of the VAT than to scrap the surtax and replace it with a bx_nd new VAT. This strategy takes advantage of the familiarity that tax administrators have with the current system and encourages a smoother transition from the existing tax system to a reformed one. 5. Implicit or hidden taxes may be as important, or even more so, as the overt features of the formal tax system. In Malawi, for instance, taxes on certain agricultural exports have been justified on the grounds of equalizing the burden of taxation between agriculture and other sectors. However, comparing formal taxes between sectors ignores the fact that agricultural pricing policies, particularly those administered through a marketing board, may already impose an implicitly high tax burden on agricultural producers. 6. It is possible to link trade and domestic tax reform successfully. The overall approach adopted in Malawi realigned the non-protective components of existing trade and production taxes to function as a rudimentary consumption tax. Though difficult to implement, these changes have either been completed or are nearly completed. 7. The exemption of low income earners from the personal income tax contributes significantly to the progressivity of the overall tax system (as only a very small percentage of households in a country are subject to the income tax). However, it is still necessary to ensure some additional progressivity in the design of indirect tax rates because  - 44 - these are the taxes that affect the poor. Progressivity in indirect taxes is also desirable to the extent that trnere are wealthy individuals who evade income tax. 8. In Malawi, the new tax system is now yielding more revenue than the old system (approximately two percentage points of GDP), in part because of the effect of inflation on tax payments, even though increasing revenue was not explicitly the goal of the reform program. However, since the budget deficit remains high and inflation continues to be a problem, further expansion of bases is necessary before the currently high statutory company tax rates and basic surtax rate can be reduced. The expansion of tax bases and improved administration are more important than rate restructuring. Both processes take time. Hence, in the short run tax reform is not a panacea for deficit reduction without substantial cutting and restructuring of expenditures.  _ 45 - Biblioar9phv Chamley, C., R. Conrad, Z. Shalizi, J. Skinner, L. Squire. "Tax Policy for Malawi", World Bank, 1985. Technical Assistance Team. 1987-1990. Memoranda to the Malawi Tax Reform Steering Group, Washington, D.C.: World Bank. Crown Agents Training Services, December 1987. "Sales Tax Reform in Malawi". United Kingdom. U.S. Internal Revenue Service, 1988. Assistance in Tax Administiration to the Government of Malawi. Conrad, Robert F. and Graham Glenday. 1989. "Malawi Tax Reform". Boston: Harvard Institute for International Development. "Aide Memoire". Prepared by the Mission to appraise the Harvard Institute for International Developmet Tax Reform Project, February 13-27. Washington, D.C.  PRE Working Paper Series Contact DAL AuIho Da for paper WPS469 Modeling Economic Behavior in J. Barry Smith August 1990 M. Abundo Peru's Informal Urban Retail Sector Morton Stelcner 36820 WPS470 Wh . Do Alternative Measures of Alexander J. Yeats August 1990 J. Epps Comparative Advantage Reveal 33710 About the Composition of Developing Countries' Exports? WPS471 The Determinants of Farm Gershon Feder August 1990 C. Spooner Investment and Residential Lawrence J. Lau 30464 Construction in Post-Retorm China Justin Lin Xiaapeng Luo WPS472 Gains in the Education of Peruvian Elizabeth M. King August 1990 C. Cristobal Women, 1940 to 1980 Rosemary Bellew 33640 WPS473 Adjustment, Investment, and the Riccardo Faini August 1990 R. Sugui Real Exchange Rate in Developing Jaime de Melo 37951 Countries WPS474 Methods for Measuring the Effect Anne Maasland August 1990 P. Dixon of Adjustment Policies on Income 39175 Distribution WPS475 Does Divestiture Matter? A Ahmed Galal August 1990 G. Orraca-Tetteh Framework for Learning from Experience 37646 WPS476 Health Insurance in Sub-Saharan Ronald J. Vogel August 1990 K. Brown Africa: A Survey and Analysis 35073 WPS477 Private Participation in the Delivery Thelma A. Triche August 1990 M. Dhokai of Guinea's Water Supply Services 33970 WPS478 Interrelations Among Child Mortality, John Marcotte August 1990 S. Cochrane Breastfeeding, and Fertility in John B. Casterline 33222 Egypt, 1975-80 WPS479 Conversion Factors: A Discussion Michael Hee August 1990 E. Zamora of Alternate Rates and Corresponding 33706 Weights WPS480 An Evaluation of Neutral Trade Policy Jaime de Melo August 1990 R. Sugui Incentives Under Increasing Returns David Roland-Holst 37951 to Scale WPS481 The Effects of Trade Reforms on James Tybout August 1990 R. Sugui Scale and Technical Efficiency: Jaime de Melo 37951 New Evidence from Chile Vittorio Corbo WPS482 Membership in the CFA Zone: Shantayanan Devarajan August 1990 R. Sugui Odyssean Journey or Trojan Horse? Jaime de Melo 37951  PRE Working Paper Series Contact 1Le Author for plper WPS483 An Evaluation of the Main Elements Refik Erzan August 1990 G. llogon in the Leading Proposals to Phase Paula Holmes 33732 Out the Multi-Fibre Arrangement WPS484 Stock Markets, Growth, and Policy Ross Levine August 1990 R. Levine 39175 WPS485 Do Labor Market Distortions Cause Ram6n Lopez August 1990 R. Luz Overvaluation and Rigidity of the Luis Riveros 34303 Real Exchange Rate? WPS486 A RMSM-X Model for Turkey Luc Everaert August 1990 S. Aggarwal Fernando Garcia-Pinto 39176 Jaume Ventura WPS487 Industrial Organization Implications Timothy Condon August 1990 R. Sugui of OR Trade Regimes: Evidence Jaime de Melo 37951 and Welfare Costs WPS488 Prepaid Financing of Primary Health Per Eklund August 1990 K. Brown Care in Guinea-Bissau: An Knut Stavem 35073 Assessment of 18 Village Health Posts WPS489 Health Insurance in Zaire Donald S. Shepard August 1990 K. Brown Taryn Vian 35073 Eckhard F. Kleinau WPS490 The Coordinated Reform of Tariffs Pradeep Mitra August 1990 A. Bhalia and Domestic Indirect Taxes 37699 WPS491 How Well Do India's Social Service Nirmala Murthy August 1990 E. Madrona Programs Serve the Poor? Indira Hirway 37483 P. R. Panchmukhi J. K. Satia WPS492 Automotive Air Pollution: Issues and Asif Faiz August 1990 P. Cook Options for Developing Countries Kumares Sinha 33462 Michael Walsh Amiy Varma WPS493 Tax Reform in Malawi Zmarak Shalizi August 1990 A. Bhalla Wayne Thirsk 37699