Tax Reform in Economies in Transition
A Brief Introduction to the Main Issues*

The transition from a command to a market economy requires profound reforms of the tax system. Such a transition will put downward pressures on the level of taxation at a time when public expenditure remains high. This paper outlines the main characteristics of the tax systems in centrally-planned economies. It describes recent changes in those tax systems. Finally, it discusses the major difficulties that will be faced, and the errors that must be avoided, during the transition.

Abstract

The transition from a command to a market economy requires profound reforms of the tax system. Such a transition will put downward pressures on the level of taxation at a time when public expenditure remains high. This paper outlines the main characteristics of the tax systems in centrally-planned economies. It describes recent changes in those tax systems. Finally, it discusses the major difficulties that will be faced, and the errors that must be avoided, during the transition.

I. Introduction

The transformation of the Central and Eastern European economies, from centrally-planned to market-oriented, will have profound implications for the tax systems of these countries. These implications are not yet fully understood. There is, thus, the danger that major mistakes will be made. In order to generate needed revenue and to perform functions which were not important in the previous environment the tax systems will need to be significantly reformed in both their legal and administrative aspects. In the process of transformation, the ratio of tax revenue to gross domestic product (GDP) is almost sure to fall and it may fall by more than the probable fall in the ratio of public expenditure to GDP. 1/ This trend will lead to, or will increase, fiscal imbalances. 2/ Given the difficulties that would be encountered in financing large fiscal deficits, inflation and external imbalances would be the almost inevitable outcome. It is, thus, important, to ensure that during the transition revenue be protected to the greatest extent possible.

The reform of the tax systems will take place under difficult economic, social, and political conditions. Successful tax reform is never easy but, given the circumstances, it is likely to be especially difficult in these countries and to take longer than many observers have assumed. In taxation there cannot be a “big bang” solution since the required changes, even when mistakes are avoided, cannot be made overnight. Because of the difficult environment in which the tax reform will be enacted, there is no certainty that the final outcome will be as good as one would desire.

The objective of this paper is to survey some of the major tax issues that will arise or have arisen in reforming these economies. 3/ These issues cannot be discussed in this paper in an exhaustive manner. Therefore, only a sketch will be provided. The paper will focus on major trends. Section II will provide a broad description of the main characteristics of these countries’ tax systems under central planning. This description will emphasize common and essential elements and will ignore the many specific aspects that made these tax systems more different from one another than is assumed here. Section III will describe some of the major changes that have occurred since these countries’ policymakers decided to take the road of market economies. Once again no attempt will be made to provide a comprehensive description of these changes. The objective of Section III is simply to identify basic and common trends in recent reforms. Section IV will discuss the remaining agenda for tax reform, emphasizing errors that should be avoided and major difficulties that must be faced. It will also contain a few concluding remarks.

II. Taxation Under Central Planning

Given the relatively low per capita incomes of the countries of Central and Eastern Europe, it is surprising that these countries have been able to collect tax revenues that, as shares of their GDPs, were as high as, or even higher than, those of the most advanced and most socially-minded countries of Western Europe. Table 1 provides, for 1986, some data for seven of these countries. 4/ The appendix tables give more details for the 1985/89 period.

Table 1.

General Government Tax Revenue of Central and East European Economies, 1986

(In percent of GDP)

article image
Sources: National sources for all countries. For OECD, Revenue Statistics of OECD Member Countries, (Paris, 1988).

1988.

Much of the tax revenue has come from three sources: (1) turnover taxes; (2) profit taxes; and (3) payroll taxes. Although the specific features of these taxes and their administration have differed among these countries, there have been important common elements. We will concentrate on these elements in order to isolate the potential effect on these taxes of the structural and policy changes taking place, or likely to take place, in these economies.

1. Turnover taxes

As traditionally applied in centrally-planned economies, turnover taxes can be thought of either as taxes charged on the sales from state enterprises to retailers, or as the difference between the prices charged to consumers (minus a small profit margin for the retailers) and the prices received by the producers. These rates were implicitly set by the planning offices when they fixed the prices for the retailers and the producers. There were no tax laws which legislated and thus fixed the level of these rates for periods of time. At times it was difficult or even impossible to obtain a schedule of relevant current rates. 5/ With few exceptions, the turnover taxes were not levied on sales made by producers to other producers. Therefore, the degree of cascading was limited. However, because of the wide variety of rates, the turnover taxes caused a lot of distortions in the price structure. 6/ Imports were also not taxed explicitly but a system of price differences was operated. The domestic price of imports would often be set, especially for consumer goods, at levels that far exceeded the landed cost of imports at the official exchange rate.

When prices are administered, both producer prices and retail prices are known. As a consequence, taxes resulting from the difference between these prices, acquire characteristics similar to those of ad rem taxes; that is, the tax revenue can be calculated once the sales volume is known. Because central planning assigned production quotas to state enterprises; because the state enterprises were relatively few in number; and because of the close controls on those enterprises, sales volumes were also easily ascertainable. When production costs changed and the government was unwilling to adjust the retail prices, discretionary price adjustments were allowed for the products produced by the state enterprises. Thus, the effective tax rates could and did change significantly over time. 7/ Turnover taxes were occasionally used to regulate demand in order to make it conform with the supply. For example, for some consumer goods in short supply, the turnover tax rates could be sharply increased. When this was not done, the lines of consumers to acquire these products became longer. For some necessities these taxes were low and in particular cases they could be negative as countries subsidized some products. Turnover taxes were traditionally applied with many rates; in some cases with, literally, thousands of rates. In Western countries even a small number of rates creates administrative headaches for value-added taxes. Thus, a question that arises is how the centrally-planned economies were able to administer these taxes.

Revenue levels much larger than those collected from value-added taxes in Western countries have been collected from these turnover taxes (see Table 1). In Western countries it has been very difficult to reach revenue from value-added taxes of 10 percent of GDP; in centrally-planned economies they have often reached, or exceeded, 15 percent. 8/ However, this comparison is in some respects misleading. Centrally-planned economies have for the most part not imposed separate excise taxes since the objectives pursued through the use of excise taxes could be achieved by directly fixing retail prices at much higher levels than the costs of production and, thus, collecting larger turnover taxes from particular products. If revenue from value-added taxes and from excises are added, then the total collection from all these indirect taxes in Western countries rises to levels closer to those of the turnover taxes in centrally-planned economies. 9/

Another aspect of the turnover taxes to which attention must be called is the method of collection. Under central planning the collection of turnover taxes was simple and highly effective. There was practically no evasion in this tax. Payments between enterprises and between enterprises and retailers were often made through credit transfers within state banks. Other forms of payment were not allowed. 10/ Since the prices of the products sold were administratively fixed and were, thus, known, and since the volume of sales was also closely monitored, the tax liability for the turnover taxes could be easily calculated by the state banks in which the enterprises had accounts on the basis of information on volume of sales. The taxes were largely transfers from the accounts of the retailers to the government account. In other words, the banks reduced the credit balances of the retailers by the required amount and credited part of that amount to the enterprises and part to the state as a tax payment. For services the collection mechanism was closer to that of Western countries but services accounted for a small share of total turnover taxes. 11/

This simplified description highlights the great dependence of the turnover taxes, as they existed under central planning, on three basic elements: (a) the price controls; (b) the method of payment; and (c) the role of state enterprises. The situation described above did not require large Western style tax administrations checking monthly declarations by taxpayers. These countries did not feel the need to develop them. Tax administrations, especially at the national level, were generally rather small and their role and power were somewhat limited. 12/ Control was exercised somewhat mechanically through field audits of the relatively small number of state enterprises.

Such a system could not survive (a) the liberalization of prices and of the methods of payments, (b) the privatization of state enterprises, and (c) a large increase in private sector activities especially in the service area. If prices were allowed to move freely, they would no longer be known with certainty. If the state enterprises were privatized or their management made more autonomous, their sales and financial accounts could not be controlled as well as under central planning. 13/ If the buyers were allowed to make payments through different channels (including cash), and through different banks, the control exercised through the payment system would break down. And, if many private activities came into existence especially in service activities, the traditional tax administrations would not be able to cope with them.

2. Profits taxes

Turning now briefly to a description of the taxes on profits, one element that must be emphasized is that these taxes, as they have been applied in centrally-planned economies, have had little in common with the taxes on enterprises in market economies. The main issues with these taxes relate to (a) the definition of profits; (b) the rate structure; and (c) efficiency implications.

In centrally-planned economies the definition of profits has been somewhat arbitrary, often administratively determined, and bearing little relation to an acceptable economic or objective definition. Lack of proper tax laws meant that there was no precise definition of taxable profit. A World Bank mission that visited nine state enterprises in Poland in early 1989 identified eight different definitions of profit. The measure of profits is further distorted by the existence of price controls, by the extremely low market interest rates paid by the enterprises, by the unrealistic amortization rules that often assume lives of assets and buildings much longer than their economic life (and, thus, lead to an overestimation of net earnings), by the discretionary and rather arbitrary nature of the expenses that can be claimed as costs, which, for example, exclude the deductibility of interest payments as a cost of doing business, by the inability to carry forward current losses, and by other problems. 14/ In general, the procedures used to determine profits often resulted in overestimation of profits and, thus, in excessive taxes.

If the definition of profits had been a distorted but still an objective one, and if the tax rates had been reasonable, the taxation of enterprises would have faced difficulties similar to those faced by enterprises in market economies especially when the rate of inflation is high and unstable and the tax system does not contain provisions for inflation adjustment. However, state enterprises have often been taxed with extremely high rates and on profit measures which were often negotiated and arbitrary. The profits tax was often the result of a bargaining process reflected in the annual financial plan, which comprised the amounts to be paid on account of taxes and payments to specified funds. The rates were often progressive with respect to some measure of “rentability” and could go as high as 100 percent when the rate of return (as measured) exceeded a given limit. There was, thus, a great disregard for efficiency considerations. In fact the tax was mainly an instrument of resource mobilization and a way of recovering past investments. More efficient enterprises were taxed more in order to subsidize less efficient ones. 15/

The state was both the tax collector and the only shareholder in the state enterprises so that it could determine what proportion of the total revenue (including amortization) of the enterprise it would appropriate. Often, it appropriated more than the objectively measured profits. Profits taxes were equivalent to a combination of taxes on enterprises and dividends distributed in a market economy. 16/ Profits taxes left very little incentives for the enterprises to become more efficient and raise profits. Efficiency was not rewarded by larger retained profits that could be used for additional investment or for increasing the incomes of workers or managers, but resulted in larger transfers to the government. 17/

Price liberalization and privatization will require profound changes in the taxation of profits. They would require that taxable profits be as close as possible to an economic definition, and, more importantly, that taxable profits be an objectively defined concept, and that tax rates are reasonable and determined in the law. Higher productivity, by generating higher profits, must convey economic advantages to the enterprises.

Once the change to a Western style type of profit taxation is made, the role of the tax administration will also have to change since it would be responsible for determining the accuracy of the tax declarations under situations where profits are measured on the basis of complex but objective accounting conventions and much information (e.g., on prices or sales) is no longer easily available. Moreover, just as managers will want to maximize profits on behalf of their enterprises, they will also want to minimize tax liabilities. The tax administration must have the skills to challenge the taxpayers when their declarations do not appear to be correct and must have the power to penalize them when they become delinquent. In short, an arm’s Length relationship with enterprises needs to be established.

3. Payroll taxes

There is not much to say about payroll taxes. These are not very different from the payroll taxes in Western countries. Unlike the profits and turnover taxes, payroll taxes have been imposed with criteria similar to those used by Western countries and the definition of payroll is similar. These taxes have been collected mostly from state enterprises often as a proportion of their payroll. 18/ Since payroll costs were often met by cash transfers from the state banks, the payroll tax could simply be withheld by the banks. Once again this was an efficient and simple system that did not require complex administrative procedures to administer. Evasion here was almost nonexistent. A market economy will bring the need to measure correctly the size of payrolls. Tax evasion may induce companies to underreport the size of their wage bill. 19/ Also, as real interest rates become positive, companies will have incentives to unduly retain the proceeds of the witholding tax. Thus, once again, tax administrations will need to be strengthened.

4. Personal income taxes

Under traditional central planning the desirable distribution of income was supposed to be achieved through policy decisions aimed at setting wages, at investing in priority sectors (as seen by the policymakers), and at setting consumer prices so as to benefit those with lower cash incomes. Private initiative was strongly discouraged through regulations, through the socialization of the means of production, and through prohibitive tax rates on incomes from certain activities. In some cases (for example, for interest incomes when real interest rates were negative), the taxation was implicit and could exceed 100 percent. 20/ In other cases, such as rents, the effective rates could also exceed 100 percent through a combination of high tax rates (at times exceeding 90 percent) and too limited deductions for expenses. Income from private business activity was legally taxed at very high rates though, in fact, the taxation was often based on presumptive criteria. These taxes were largely negotiated with the taxpayers. Wages were generally taxed either with proportional rates or with mildly progressive rates. Normally, the net result was a large number of schedules and highly uneven treatment of taxpayers, which was made worse by the discretion granted to regional and local authorities in the granting of tax exemptions.

For example, in Bulgaria the marginal rate on wages and salaries was 14 percent, on incomes for artists and scholars it was 50 percent, on “unearned income” (such as dividends) it was 81 percent, and on income from private business activities it was 85 percent. In fact, since the incomes of individuals came largely from wages and salaries much of the personal income tax revenue came from this source. For example, in Czechoslovakia, although the top rate on wage income was 20 percent while that on nonwage incomes was 80 percent, wage incomes contributed 95 percent of the total revenue from personal income taxes.

There were other taxes, for example, those on agricultural properties and on motor vehicles or, occasionally, on imports, but these generated little revenue so they are not discussed here. 21/

III. Main Features of Tax Reforms

The desire to restructure these economies toward a market system, and the realization that such a restructuring would have major implications for tax revenue, have prompted these countries to begin reforming their tax systems. All of them, to varying extents, have recently made significant changes. Hungary has been at the forefront of this movement, having started the process in the mid-1980s. This country now has a tax system which, in spite of some shortcomings, is broadly similar to the tax systems of Western and, particularly, EC member countries.

1. Changes in turnover taxes

The first and most important area for reform has been the turnover tax since this tax would be the one most affected, directly, by the restructuring of the economy. The changes in this tax have followed a fairly similar pattern. First of all, these countries have tried to reduce or eliminate the turnover taxes levied with negative rates by increasing the retail price of many products or, in some cases, by directly subsidizing state enterprises. Second, the nonparametric nature of these taxes has been reduced or discontinued since these taxes would no longer serve as a tool for the plan. In some of these countries, the rates of these taxes have been changed from de facto specific and variable, to ad valorem and, at least until the next legislative change, parametric. These rates are now determined by tax laws rather than resulting largely from the pricing decisions of planners. These new, ad valorem, rates are expressed either as shares of wholesale prices or of retail prices. In either case, they are collected at the production stage. The change from specific to ad valorem rates is important because it allows free movements of prices at the retail and/or the production level.

The third major change has been a drastic reduction in the number of rates while maintaining the general character of the turnover taxes. For example, Bulgaria has reduced the number of rates from about 2,000 to 43 and is planning to reduce them further to 4; Romania has reduced the rates to 20; and, Czechoslovakia to 4. Fourth, the spread among these rates is being progressively reduced although it still remains much too high. Finally, there is a move to extend the turnover tax to all enterprises rather than limiting it to the state enterprises. The administrative implications of this latter change are very important since, in a situation of rapid growth in the number of private sector establishments, it would considerably expand the number of taxpayers raising questions about the ability of the current tax administrations to cope with these increases. It would, thus, be necessary to exempt from these taxes enterprises with total sales lower than a given limit. This obviously requires controls to ensure that this limit does not become a porous or soft one.

Several of these countries did not have excise taxes so that particular products which, for health or other reasons, called for higher levels of taxation were subject to higher turnover tax rates. The traditional items generating large excise tax revenue in Western countries (alcohol, tobacco, and petroleum products) also contributed a very large share of total turnover tax revenue in these countries. In the Soviet Union and in Poland, for example, this share exceeded 50 percent.

Finally, another identifiable tendency is the widespread intention to introduce value-added taxes. This tax, together with the use of excise taxes, would replace the turnover taxes and would move the tax systems of the Eastern and Central European countries closer to those of the European Community thus facilitating the way for eventual membership. A value-added tax is already in existence in Hungary where it was introduced in 1988. 22/ Poland, Romania, Bulgaria, Czechoslovakia, and the Soviet Union are all planning to introduce value-added taxes in the next two to three years.

2. Changes in profits taxes

Significant changes are also being introduced in income taxation. Profit taxes represent a very high share of GDP in the countries of Central and Eastern Europe. 23/ As discussed earlier, difficulties with this tax have arisen from both the definition of profits and the non-parametric value of the tax rates. This nonparametric nature is in part the result of the fact that the government is also the sole shareholder and, thus, feels entitled to appropriate part of the “after-tax profits.” 24/ These countries are now paying a lot of attention to the possibility of making this a tax levied with parametric or stable tax rates on an objective definition of income. 25/ They are also more concerned with increasing efficiency and reducing distortions.

The recent reforms, while generally in the right direction, have not gone far enough in achieving these objectives. Accounting and conceptual difficulties are likely to remain and the concept of efficiency is still not well understood. Profits, especially if occurring in the private sector, are likely to be seen with a high degree of suspicion and the idea that higher profits are often a worthwhile objective (if they are not tied to monopoly) is not a popular one. For example, in the Soviet Union, although the normal tax on the profits of enterprises has been reduced and unified at a standard 45 percent rate, a 90 percent marginal “excess profits tax” rate has been imposed on profits above a notional average for the industry. 26/ Tins, of course, would reduce the incentive by the enterprises to achieve above average profits. In some of these countries there has been the view that profit taxes should be related to the rate of return to investment rather than to the size of the profit, the implication being that a higher return is not desirable.

In all of these countries there have been recent reductions in the rates at which company profits are taxed. For example, in Czechoslovakia, the rates were reduced in 1989 from the 75-85 percent range to 55 percent. In Poland and Hungary the rates were unified at 40 percent as of 1989 or 1990. 27/ However, because of rules related to depreciation, loss carryover, and the treatment of other items in the balance sheet of enterprises, the taxes on enterprise income remain very high. Because of this, and partly to encourage foreign investment, foreign enterprises and joint ventures have been provided tax holidays and other incentives and separate tax regimes (including tax rates). This has given rise to some abuses as the “foreign investment” may be domestic capital that goes abroad and returns as foreign investment. Special treatment continues to be provided to particular activities such as agriculture, food processing, etc. Future reform must aim at a better Leveling of the field at a lower level of taxation.

3. Payroll and personal income taxes

The foray into a market economy has raised questions about what kind of taxes should be levied on the incomes of individuals. Should these countries follow the popular Anglo-Saxon tradition of taxing personal incomes through global income taxes, or should they follow the older Latin or Mediterranean tradition of maintaining distinctions through a schedular approach? Another question is how high should the rates be?

As in other areas, Hungary has led the way in reforming the personal income taxes with the introduction, in January 1988, of a “global” personal income tax with rates ranging from 20-60 percent, on 21 brackets. 28/ The rates were reduced to the 17-56 percent range, on 8 brackets, in 1989. As of 1990, the top marginal rate has been lowered to 50 percent. There were lots of problems with this tax (problems of income aggregation, treatment of family, special preferences, etc.). These problems are likely to be corrected or, at least, reduced, in the near future. One criticism that has been advanced to this tax is that it represents an “invasion of privacy.” The distrust toward past governments has made some observers (for example, Kornai) question whether a progressive global income tax which requires individuals to provide much detailed information to the government can function in these countries’ present environment. 29/ In other countries current proposals would reduce the dispersion in the treatment of incomes from different sources and would reduce the top tax rates to the 40-60 percent range.

As of now major increases in payroll taxes have taken place in Czechoslovakia (from 20 percent to 50 percent in 1989) and in the Soviet Union (from 12.5 percent to 37 percent in 1991). In Hungary payroll taxes are now relatively high, at 53 percent. These increases are likely to be copied elsewhere. In both Czechoslovakia and the Soviet Union broad social security reforms were one of the government’s first reaction to popular discontent. Also in both countries, very large changes in administered prices were planned for the end of the current five-year plan. The high payroll taxes were set so as to take from the enterprises any gains that the price change would have given them, especially since a large part of the price increase implied falls in turnover taxes. These taxes are likely to attract attention in all countries in connection with badly needed revisions in social security and in pension schemes.

Another aspect worth calling attention to is the use by several of these countries of excess wage taxes. Enterprises have been discouraged from raising wages by more than allowed percentages by the imposition of very high taxes on enterprises granting excessive wage increase. One is reminded of similar tax-based incomes policy proposals discussed in the United States in the 1970s. It is not clear how effective these taxes have been in slowing down the increases in wages. They do indicate, however, that these governments have been losing control over the state enterprises. 30/

IV. Remaining Agenda

The previous discussion should have provided a feel for the need for fundamental reforms in the tax systems of the Central and Eastern European countries if these countries are to become full-fledged and efficient market economies. The fundamental reforms will be (a) in attitudes, (b) in the statutory characteristics of the tax systems, and (c) in their administrative set-ups. All of these will have to change if the reforms will be more than cosmetic. In the previous section recent changes have been briefly outlined. These indicate that some progress has been made but clearly much remains to be done. The reforms will be constrained by the need to maintain adequate levels of taxation and by the degree and pace of other economic reforms. The tax to GDP ratio will inevitably fall but it cannot be allowed to fall much below the public expenditure to GDP ratio. The latter is not likely to fall dramatically soon. 31/ The reforms will also be constrained by lack of particular skills (administrative, accounting, etc.).

Let us begin with the question of attitudes. The move to a market economy will require a strong faith on the part of the policymakers in the virtues of such an economy. This should be the main reason for the economic transition. Although such a faith need not extend to believing that the market is always right (clearly it is not), or that all government intervention is misguided (clearly much of it is necessary and useful), it does sharply limit the role of policymakers. It would be a major error to accept the movement to a market economy while continuing to believe that policymakers can consistently make better choices than the market. If this belief were prevalent, the policymakers would tend to replace planning with tax (or expenditure) instruments and to continue engaging in social engineering. It is, thus, important that they refrain from introducing value-added taxes with several rates 32/ (based presumably on the degree of “necessity” of a product); they should refrain from choosing winners and losers, or “essential” and “nonessential” investments, through tax rate differentials or incentives to enterprises engaging in some activities. They should limit the extent to which they discriminate among different types of personal incomes. And, they should refrain from making the rate structures highly progressive to, presumably, reduce inequalities.

One fundamental assumption of a market economy is that, to a large extent, profits and high incomes reflect positive contributions to the material well-being of society by individuals through better initiative, more risk taking, more effort, and better economic decisions. Thus, profits must be seen as deserved compensations for the pursuit of efficiency. This assumption requires that those who get high incomes must be able to retain at least a good part of them. How quickly the policymakers of these countries will fully grasp the importance of this assumption and accept it remains to be seen. It should be recalled that for decades the citizens of centrally-planned economies lived in an environment where private profits were considered as sins and where the tax system was used to kill these profits. Incidentally, the link between free market prices, efficiency, and profits must also be recognized.

One difficulty is that, during the period of transformation of the economy, many opportunities will arise for some individuals to take advantage of market imperfections (including temporary monopoly situations, insider information, etc.) to earn large incomes. These incomes do not have the same legitimacy as those which normally originate from the pursuit of efficiency in mature market economies. They will be associated with the concept of exploitation. How should the policymakers deal with these gains? Is it possible to distinguish in the tax laws between “good profits” and “bad profits”? Trying to penalize “bad profits” through the tax system is likely to justify tax reforms which may be highly deficient over the longer run. But accepting these profits through the tax system may make the population far less disposed toward, and more critical of, a market economy.

Particular developments will create difficulties for the tax system. Some of these have already been mentioned as, for example, the difficulties to the turnover taxes arising from free prices. Others are more subtle. For example, privatization and the removal of public planning will bring into existence many thousands of private activities. These activities will be difficult to tax because they will be mostly small and service-oriented and because the tax administrations will not be prepared to deal with them. At the same time, the share of total production contributed by the state enterprises is likely to fall. In other words, the share of GDP that is more difficult to tax will be the most dynamic one. This trend will make it all the more difficult to maintain a high tax to CDP ratio.

The early 1990s should see several important developments in the characteristics of the tax systems. First of all, there should be a major shift from taxes on production to taxes on consumption. The share of value added at the wholesale and retail level will likely grow. This factor as well as the need to keep indirect taxes fairly neutral will require the change from collecting taxes from a few thousand large state enterprises to collecting taxes from a far larger number of shops and enterprises. The relative role of excises is likely to grow, first, to exploit inelasticities in consumption and, thus, to minimize welfare losses; second, to generate some extra revenue; and, third, to achieve the same goals as in market economies (discourage consumption of tobacco and alcohol, make car owners pay for benefits received through the use of roads, etc).

Import duties can and, perhaps, should also contribute to total revenue especially in the period of transition. There must be scope for significant revenue-raising tariffs to replace quantitative restrictions as these are dismantled. Import duties could buy time for the government to reconstruct the domestic tax system. They could be dismantled gradually as the domestic tax effort strengthens and competitiveness improves.

Another important development will be the shift from taxes on the incomes of enterprises to those on the incomes of individuals. In the market economies (at least in industrial countries) a large proportion of all income taxes collected has come from taxes on individuals. In these economies the taxes on enterprises are, on the average, less than one fifth those on individuals. In centrally-planned economies a large proportion has come from few enterprises. This proportion will have to change since, once again, the large enterprises will lose ground and the cash incomes of individuals and small personal businesses will grow.

All of these developments will have very important implication for the statutory structure of the tax system and for the organization of the tax administrations. The tax administrations that exist in these countries have served them well in connection with their traditional tax systems. They would be inadequate to cope with some of the developments outlined above. New administrations should be created. This will not be easy because, unlike tax laws that can be copied from other countries, tax administrations do not lend themselves to reproduction through cloning. This is an area where so far progress has been very slow but it is the area which will largely determine whether the transformation of the tax systems of these countries will be successful. This is one basic reason why the statutory changes must be as simple as possible. The more complex these changes are, and the more objectives the policymakers try to promote through the tax system, the more likely it will be that the tax administrations will prove inadequate to cope with them.

Limited administrative capacity should make policymakers choose simpler alternatives—VATs with one rate, exemptions rather than zero rating, taxation of individuals rather than families, 33/ and so on. The role of banks in the collection of taxes in the new system must be studied and the design of adequate administrative mechanisms must be developed. The introduction of computers must follow the clear establishment of an adequate administrative structure rather than precede it. 34/

Finally, a word on foreign advice is in order. At this time work on Central and Eastern Europe is very much in fashion and many people are volunteering, or attempting to sell, their advice to the policymakers of those countries. Some of this advice is likely to be amateurish and damaging. It is important for these countries to be careful to screen the technical assistance they receive to ensure that it is in their best interest.