Vito Tanzi, Director of the IMF’s Fiscal Affairs Department, and Howell Zee, Chief of the Department’s Tax Coordination Division, met with the IMF Survey to talk about their Working Paper, Tax Policy for Emerging Markets: Developing Countries.


Vito Tanzi, Director of the IMF’s Fiscal Affairs Department, and Howell Zee, Chief of the Department’s Tax Coordination Division, met with the IMF Survey to talk about their Working Paper, Tax Policy for Emerging Markets: Developing Countries.

IMF Survey: What are the special characteristics and needs of developing countries that set them apart from other countries in terms of tax policy?

Tanzi: The most basic characteristic is the level of taxation, which in industrial countries is twice that in developing countries. Thus, the governments of industrial countries can do many things that the governments of developing countries cannot do. When developing countries attempt to do the same things, they get into trouble because they don’t have the resources. A second characteristic is the structure of taxation. In industrial countries, a large proportion of taxes comes from income taxes, especially on individuals. In developing countries, the share of personal income taxes is very small. A third characteristic is the quality of the tax administration, which is much better in industrial countries where the actual, or effective, tax system is not very different from the nominal, statutory one. The laws are broadly applied as they are intended to be applied. In developing countries, tax laws are passed, but the application is often very different.

Zee: The differences are not independent of each other. To a large extent, the differences in the level of sophistication of tax administration in developing countries influence the way these countries raise revenue.

Tanzi: Also, industrial country taxes are broadly mass taxes. In developing countries, the number of taxpayers is much smaller, because the distribution of income is much less even and the administration is not as good. So, the focus is on fewer individuals and fewer corporations.

IMF Survey: How critical is the level, as opposed to the composition, of taxation?

Tanzi: The level is very important for determining what a government can finance. Governments have certain basic needs, such as building roads, providing schools, and hiring soldiers and police. The structure is important for determining the incidence—who pays the taxes, how fair the tax system is, and so on. But the two are, of course, related.

IMF Survey: What would you counsel a developing country to do if it wanted to attract different types of investment? What is the role of tax incentives in advancing economic development?

Tanzi: This is a hot topic. The general view is that tax incentives are pretty useless and that the best incentives are low tax rates and a broad base. A tax system that is well designed, fairly enforced, and well administered is really the best incentive because investors want certainty; they don’t want too many changes.

Yet, for a variety of reasons, governments in some countries cannot face the idea of leaving allocation to the market. They want to play an active role, and sometimes this active role is an honest role. They truly believe they can influence investment decisions in a certain way. But incentives lend themselves to corruption. They are rarely totally objective. Somebody somewhere in the government has to make the decisions. Visualize a situation where a multinational corporation wants to invest millions of dollars in a country and is negotiating with a person making $200–$300 a month. You can see the possibilities for corruption. Sometimes tax incentives are given to favor friends or people of your party or religion, and sometimes they are given for defensive reasons. If I am a U.S. company and I want to invest in Central America, I can invest in any country. So I go to Costa Rica and say “Look, I want to invest this, but I’m indifferent between you and Nicaragua and Guatemala. If you give me a tax incentive, I’ll come here.” There are some quite prominent economists who still argue that tax incentives are a good thing. But they often do not take fully into account all the problems that arise when you try to make these incentives operational.

Zee: I find that no matter how many times we advise developing countries that they will run into problems if they use tax incentives, they respond that all the countries in the region have incentives, so why shouldn’t they? If they abandon them, all the investments will go to their neighbors. This problem requires a multilateral approach. Advising one country to abandon tax incentives will not be persuasive. You have to take at least a regional approach. Countries in a particular region have to agree to some sort of a policy. One country will not implement a policy that other countries in the same region do not follow. Also, some incentives are much better than others, and countries should focus on those that stand a better chance of being effective.

IMF Survey: What are the special issues developing countries must address in expanding their tax bases?

Tanzi: If you think of the income tax as being corporate and personal, the corporate income tax is about the same in industrial and developing countries, but the personal income tax has been a tremendous failure in most developing countries. Not too many developing countries collect more than 2 percent of GDP from the personal income tax, compared with 11 percent or more in industrial countries. There are several reasons for this tremendous difference. First, rich people are more powerful in developing countries than in industrial countries, and they don’t pay anything. Second, there are few large establishments where you can control the income that people earn, and there is a large informal economy. The informal sector, which accounts for 50–60 percent of the workforce, is difficult to tax. Administration is not very good, so governments often end up taxing public employees and the employees of large, especially foreign, corporations. These are the ones who contribute the most, with the result that much of the revenue from the personal income tax comes from wages and salaries and not from dividends or interest or profit. The number of taxpayers is relatively small, especially those who are subject to high marginal tax rates.

IMF Survey: Related to this, how does a country develop a tax culture—a sense of obligation, civic duty, an understanding of why they pay taxes?

Tanzi: A combination of factors would go into this. First, the government has to convince the people that government expenditure is productive and equitable, that there’s no corruption, and so forth. If people begin to feel that what they pay the government is wasted or ends up in somebody’s pocket, then the attitude toward taxes is not very good. Second, the tax system must be transparent, clear, and simple. Very often, tax systems are complex: people don’t understand them. Third, there is no culture of teaching taxpayers their obligations. Sometimes parliament passes a law, and that’s the end of it. There’s no attempt to explain the laws in simple terms. Along the same lines, services to the taxpayers are limited. In a country where the system works well, taxpayers can expect to get their questions answered quickly, get the forms they need, and make tax payments quickly and efficiently. In some developing countries, forms are often not available, questions cannot be answered, and taxpayers making payments have to stand in line for three days. This makes compliance costs extremely high. Finally, the administration must identify the taxpayers accurately and make sure they can follow simple rules. Penalties have to be reasonable, and they have to be applied. All these things help create an efficient tax system.

IMF Survey: How does a developing country with limited resources strengthen its tax administration?

Tanzi: We do a lot of work in many countries in this area. Sometimes, it feels like we are standing beside a lake and throwing a stone in the water. That causes a lot of movement initially, but a few minutes later, everything goes back to the way it was. Our mission team goes to a country, works very hard, writes a report, and advises people. The mission leaves, and everything goes back to normal.

Tax administration requires, first, a clear strategy. You cannot identify 100 different things that need to be done and go after all of them at once. Normally, there are two or three things that need to be done first. Sometimes we recommend what we call a large taxpayer unit to identify the largest 100 or 500 taxpayers in a country and develop techniques and programs to relate to them exclusively. Once this runs smoothly, the number of taxpayers can be expanded. Tax administration is one of the essential fiscal institutions for a country. If it doesn’t work, you can pass the best laws in the world, and they will not amount to much.

The level of wages is also very important. In many countries, wages are so ridiculously low that they are an invitation to corruption. The choice of tax administrators is also important too. Those who are chosen must have a good technical preparation, but there must be some indication that they are honest.

Another issue is whether the administration organizes its procedures by tax or by function. We have discovered that organization by function—assessment, collection, audit—is much more effective than organization by tax.

In some developing countries, every tax involves direct contact between taxpayers and tax administrators, and assessments are discretionary. The more opportunities there are for contact between the two, the more likely it is that they will establish a relationship, and corruption will develop. Thus, distance between taxpayers and tax administrators is very important.

Zee: Many developing countries are using scarce administrative resources to perform tax assessments for taxpayers during the filing phase and not enough for the audit and enforcement functions in the post-filing phase. It isn’t possible for a country to have complete control over taxpayers. There are too many of them. Many countries have moved to a self-assessment system followed by a profiling according to compliance risks. In this way, an effective audit strategy can be developed.

Tanzi: Without self-assessment, you cannot have mass taxation. You will be limited to relatively few individuals and enterprises. It is better to shift much of the work from the tax administration to the taxpayers.

IMF Survey: What do you see as the key tax priorities for developing countries in the near term?

Tanzi: The first priority would be to tax personal income more effectively, because income distribution is getting worse in the majority of developing countries. A major tax in the developing world is the value-added tax, and in many countries it needs adjustment—widening the base and, in some cases, not having multiple rates. I would also emphasize pruning the tax system. Over time, most tax systems develop unproductive branches. Often, these taxes remain on the books and cause confusion. I’ve always argued that one objective of tax systems should be to reduce the number of taxes to five, six, or seven. Another goal might be to get rid of foreign trade taxes or at least reduce the reliance on them.

Zee: Also, five to ten years down the road, these developing countries will probably face the same tax policy challenges as the industrial countries. Developing countries would be well advised to look ahead and anticipate those problems—for example, in taxing financial capital because capital can move very easily.

IMF Survey: Are the needs of transition countries different from those of both developing and developed countries?

Tanzi: The transition countries started the transition without having either a true tax system or a true tax administration. The arrangements that existed during the centrally planned period were very different from those of a market economy. Between industrial and developing countries, the differences are in the details and the quality of what is being done, but they are not fundamental. Between industrial countries and transition economies, there are enormous differences. The transition countries had to create a tax system and a tax administration without having the personnel, a taxpaying tradition, or accounting and legal skills. They had computer skills and could introduce computers very quickly, but they didn’t know what to do with them.

They had to do all these things while maintaining a level of taxation that is very high for the level of per capita income. Generally, rich countries tax more than poor countries. Most of the transition economies are relatively poor, but they started transition with a very high tax burden—sometimes 50 percent—that really could not be maintained. Taxpayers in these economies didn’t know what a tax was: they had never paid taxes. All the taxes had been transfers from state enterprises to the government.

It’s also difficult to convince policymakers that they should not have tax incentives. The tremendous proliferation of tax incentives has been one of the big fights in these countries. For example, when Hungary, one of the most successful countries, went into the transition, they wanted to give preference to writers and athletes. They held the view that some people were more deserving than others, regardless of the level of income.

IMF Survey: What are the implications of globalization for tax policy?

Tanzi: My view at the moment—which may not be totally shared by my colleagues—is that globalization will, over time, create problems in terms of tax revenues, especially for the high-tax industrial countries like Sweden, Denmark, and Canada. We know the direction of the problem, but we don’t yet know the magnitude. They will have difficulty maintaining the high level of taxation for a number of reasons, which are becoming more and more important: electronic commerce, use of hedge funds, offshore centers, travelers buying goods where they are cheapest and where tax rates are lowest, and tax-haven countries that impose very low or no taxes on imports of capital so that capital is channeled through them. In developing countries, the impact of globalization is far less certain. I’ve worked in Latin America, and many times I’ve told countries to tax interest income, dividends, and so forth, and they tell me they cannot because, if they do, the money they have will move to Miami. Overall, I would guess that the impact of globalization on industrial countries, especially on the welfare states, will be larger in absolute numbers than in the others.

Copies of IMF Working Paper 00/35, Tax Policy for Emerging Markets: Developing Countries, by Vito Tanzi and Howell H. Zee, are available for $10.00 each from IMF Publication Services. See page 223 for ordering information.