The goal of tax policy should be to raise revenue in ways that do the least possible damage to the economy. That means a tax system that is careful to preserve individual incentives for productive activity incentives for working, saving, investing, learning, inventing and creating. Taxes should also be as simple and "neutral" as possible, without special favors for some people, and extra taxes for others.

Like businesses, governments too have to compete in providing essential services at the lowest possible cost. The cost of government is often one of the largest costs of doing business. High tax rates often repel foreign investment and cause national savings to flee to other countries. This is called "capital flight." High tax rates may even cause many of the most skilled and industrious people to move to other countries. This is called "brain drain." High tax rates always cause much of the economy to disappear from the tax collector's view. This is called the "grey" or "underground" economy.

My first point to emphasize is that all taxes fall on individuals on people who supply labor and capital to the formal economy. A tax on corporate earnings must be paid by the individual workers, customers and owners of the corporations. A tax on what we consume is also a tax on what we produce: We have to produce in order to earn the income needed to consume. Consumption is the motive for production.

What is most important is the marginal rate of all taxes on activities that would add to income -- such as working harder, investing time and money in a better education, or saving and investing in a new business. When high marginal tax rates punish added income, they must also punish added production. That slows economic growth. When the economy grows slowly, tax receipts also grow slowly. To measure the total marginal burden, we cannot just look at one tax at a time. We have to look at the way they add up.

Neither Lithuania's payroll tax of 31% nor the individual income tax of 33-35% would be unbearable by itself. But the combination of these two taxes makes it extremely difficult to collect either one. If a worker who earns more than the tax-exempt amount makes some extra effort to produce an additional 130 litas per month, the employer must first pay 30 litas in Social Security tax, leaving 100. Then the worker must pay 35 % tax on the 100 litas, leaving 65. When the 65 litas are spent on consumer goods, the purchasing power of the 65 litas will be further reduced by VAT leaving 53 litas as the after-tax reward for earning 130. The marginal tax rate on added labor income is at least 59 percent, without even counting excise taxes, tariffs, the corporate tax, or the employee's payroll tax. When marginal tax rates are so high, they discourage work or drive it underground. A report on Lithuania by The Economist of London notes that "employers pay a 30% tax (social security contribution) on their workers' wages or salaries, a percentage which does nothing to encourage employers to reveal how much they are actually paying employees. This also explains why large numbers or Lithuanians have vanished from the job market but not shown up in official unemployment statistics."

When income and payroll tax rates are too high, they discourage working or saving in the formal economy. When excise taxes, value-added taxes and tariffs are too high, they discourage buying and selling in the formal economy. The IMF estimates that one-fourth of Lithuania's economy is already "underground" -- escaping taxes. The economist estimates 30 percent. This raises the tax burden on everyone else.

In Lithuania, the 1988-92 average of tax receipts was only 5% of measured GDP from individual income tax, but 7.6% of GDP from Social Security tax (estimates for 1994 are similar). Since the economy is probably at least 25% larger than the GDP figures show, due to underground work and sales, the Social Security tax collected only 6.1% of actual GDP, and the income tax raised about 4% of GDP. Labor income probably accounts for at least 60% of GDP. If so, a 30% payroll tax on 60% of GDP should yield revenue equal to 18% of GDP. The fact that the payroll tax yield is closer to 6% is a rough gauge of avoidance. It is possible that total elimination of the individual income tax might increase compliance with the Social Security tax by enough to raise 11% of actual GDP from that tax alone as much as is now collected from both taxes.

My second point is that extremely high tax rates often yield less revenue than lower tax rates, even in the short run. In Turkey, the lowest tax rate on individual income is 25%, and the highest rate is 55%. Revenues from these high tax rates amounted to only 5.3% of GDP in 1995. France, with tax rates as high as 57%, collects only 6.2% of GDP. The U.S., with much lower tax rates than France, collects 9.9% of GDP from this source. New Zealand, where the top tax rate is much lower than in the U.S. (33%), the individual income tax brings in 16.6% of GDP. Clearly, we cannot assume that high tax rates will produce high revenues.

The same story could be told about any other tax. Countries with VAT rates higher than 20%, for example, collect no larger a share of GDP from that tax than do countries with VAT rates of 13-14%. The problem is not simply that very high tax rates on income or sales yield little immediate revenue, as a share of today's GDP. The more important effect is that high tax rates slow the real growth of revenue over time by slowing economic growth.

Economic Miracles

To understand how tax policy can help or hinder economic growth, it is useful to examine the experience of many countries that emerged from serious economic trouble and began sustained periods of prosperity. These are called "economic miracles." Every economic "miracle," at least since Poincare cut France's top tax rate in half in 1927, has included a large reduction in the highest marginal tax rates. Some of the most famous miracles, such as Germany after 1948 or Chile after 1984, also included a huge reduction of taxes on imports ("tariffs" or "import duties"). Some of the most durable reforms, such as Hong Kong since 1983 and Argentina since 1990, also included a currency board or similar firm commitment to preserve the value of money. Today, I am going to focus taxes. But I would welcome questions about monetary policy, trade policy, or price controls.

Bolivia

In mid-1985, Bolivia's inflation rate reached 23,000%. Real GDP had fallen dramatically every year from 1978 through 1986. In 1987, however, inflation suddenly dropped below 15%. Since then, the Bolivian economy has grown by about 4% a year, with very little inflation.

How did they do it? The highest income-tax rate in Bolivia was slashed from 45% to a flat rate of only 10%. A 30% corporate tax was abolished and replaced with a 2% tax on corporate net worth. Average tariffs were cut from 80% to 20%. Over 400 sales taxes, some quite high, were combined into a 10% value-added tax. What was most unique about the Bolivian reform is this: If individuals can prove they paid VAT, they can deduct that entire amount from income subject to the income tax. As a result, Bolivians began demanding receipts, in order to claim the tax deduction for VAT. The underground economy suddenly became visible, and taxable. Tax receipts soared from 3.5% of GDP in 1984 to 14.7% in 1986. In effect, the income tax (which has since been increased to 13%) is now mainly an enforcement device for VAT.

Mauritius and Botswana

Another famous "miracle" is Mauritius, a large island off the coast of Africa. Mauritius suffered a major devaluation in 1980, accompanied by a 10% drop in real GDP, price controls, and 30% inflation.. By 1982, Mauritius had an unemployment rate of 22%, and one fifth of the people were attempting to emigrate.

Yet the economy suddenly began to take off, growing by 4.8% in 1984, then 6.9%, 9.7% and 10.2% by 1987. The budget deficit fell from 14% of GDP in 1982 to 1% by 1988, as revenues poured in from the rapidly expanding economy. Economic growth has averaged 5-6% ever since.

What caused this amazing switch from despair to ambition? At a 1992 World Bank conference, Paul Romer emphasized that "income and corporate tax rates were halved in 1983 (from about 70 to about 35 percent)." Income tax rates were recently reduced again, to no more than 30%.

Another African economic miracle is Botswana. For many years, economic growth in Botswana was the fastest in the world, averaging more than 10% per year for the first two decades after achieving independence in 1967, and still remaining above 4% in the 1990s. Botswana has no payroll tax or VAT, and has repeatedly reduced its highest income tax rates -- from 60% in 1979to 30% in recent years.

South Korea

South Korea briefly adopted a typical "austerity plan" in 1980, with a 1 7% devaluation and higher tax rates. Inflation jumped to 35% and real GDP fell by 5%. Later, a 1987 IMF report noted that, "During 1981-82, structural policies [included] a comprehensive tax reform and trade liberalization.". The highest income tax rates were immediately reduced by 20 percentage points in 1981, and by another 30 percentage points in later years. "Tax revenue . . .rose considerably." "Average tariff rates were also lowered from 35 percent in 1980 to 23.7 percent in 1983 and to 12.7 percent by 1988." Economic growth averaged 9.4% a year from 1980 to1990, slowing to 7.2% from 1990 to 1995.

Like much of Southeast Asia, South Korea is once again having problems with careless banking and company borrowing. But this setback will be quite minor when set against the amazing progress of the past 15 years. South Korea is now the 11th largest economy in the world.

Jamaica

In Jamaica, real output and income began falling in 1974 and continued to drop almost every year through 1986, by a total of 23%. Tax brackets were not adjusted for inflation, so the top tax rate of 57.5% eventually fell on incomes as low as $700 a year. More than a third of Jamaica's professionals and managers left the country.

In 1986, Jamaica cut the highest tax rate to 33% and reduced tariffs. Revenues immediately jumped from 26.8% of GDP to 31.9 % in 1986 -- a very high average burden, but nevertheless a less destructive set of marginal tax rates. Economic growth averaged 5.7% from 1987 to 1990, but slowed againafter that. So, the income tax rate was further reduced to a flat rate of 25% in 1994, and the economy again rebounded. Tariffs were also further reduced in 1993, yet the added trade volume meant that "receipts from import taxes grew by 2 percent of GDP despite lower tariffs."

Colombia

In Colombia, individual income tax rates were reduced from 56% in 1983 to 30% by 1986. The corporate tax was also cut from 40% to 30%. The maximum VAT was later reduced from 25% to 14%. Revenues rose from 7.8% of GDP in 1983 to 12-13% since the late 1980s. What is more important than the share of GDP, however, is that real tax receipts have increased year after year because of prolonged growth in real GDP -- almost 5% per year ever since 1986.

Chile

Chile is now regarded as an economic miracle. But in 1973, notes Sebastian Edwards, "Chile face initial conditions as close as they can possibly be to those of Eastern Europe. Chile had severe repressed inflation [1000%] . . an incredibly large fiscal deficit [27% of GDP], and a very large public sector in which all sorts of public and government-owned enterprises were operating at a very inefficient level." Even as late as 1983, just before most reforms began, Chile still had the highest per capita foreign debt in Latin America.

In this case, the taxes that were most dramatically cut at the beginning were not income tax rates, but the Social Security tax and tariffs. In the six years following May 1981, Social Security tax rates fell from 25% to zero. Chile privatized the social security system, replacing it with a system in which workers must pay 10% of their salaries into their own tax-free, private mutual funds. The private pension fund has resulted in a very high saving rate, and helped to develop modern capital markets for financing business expansion.

By 1984, average tariffs were also reduced from 94% to 10%. Income-tax rates were reduced at most income levels, although only from 65% to 45% at the top. By the mid-eighties, marginal tax rates on those earning twice the average family income were only 13%. Saving is favored because only"real" interest income is taxed. The corporate rate as deeply cut, from 47.5% to 32.5%. The VAT was reduced from 30% to 18%. Tax revenues rose from 7.8% of GDP to 10.6% right after the first wave of tax and tariff reform.

Singapore

Singapore found itself in recession in 1985, and responded by reducing the top income tax from 45% to 30%. They also reduced the percentage of payrolls required to be devoted to mandatory savings plans. Economic growth from 1987 to 1997 has averaged almost 9% a year. As each new Asian country copied the successful policies of its neighbors, economic miracles proliferated (despite unfortunate financial excesses in Thailand, Malaysia, Indonesia, and South Korea). More recently, a similar wave of tax rate reductions have been spreading across Latin America.

Argentina

In 1989, Argentina's economy shrank by 6.2%, and inflation exceeded 3000%. In 1991-92, Argentina reduced the highest income tax from 62% to 30%. Tariffs were completely eliminated on capital goods and reduced to no more than 20% on almost everything else. Government revenues rose from 13.1% of GDP in 1988 to 17.7% of GDP in 1992-93. That understates the revenue again, because real GDP rose by 25% in three years. Between the first quarters of 1991 and 1994, government revenues more than doubled, rising from 2 billion to 4.2 billion new pesos (equivalent to U.S. dollars).

Argentina soon began repatriating some of the huge investments that had escaped abroad to avoid taxes and inflation (about $60 billion), and the country attracted considerable foreign investment. The result was a large inflow of private capital -- $9.6 billion in 1992 and over $14 billion in 1993.

Unfortunately, the Mexican devaluation of late 1994 caused foreign investment to flee the entire region. But Argentina's currency board did not devalue, so (unlike Mexico) Argentina did not have to cope with both recession and inflation in 1995. For the twelve months ending in October 1997, industrial production was up 12.5%, and inflation was zero.

Peru

In Peru, real GDP fell by more than 25% from 1988 to 1990, and inflation reached 7,500%. In 1992, the highest income tax rates in Peru were cut from 50% to 30%. VAT rates of up to 55% were brought down to a single rate of 18%, and excise taxes were eliminated in 1993. Tariffs were reduced to 15% for almost allgoods, and 25% for a few others. Revenues increased from 6.5% of GDP in 1989 to 10.2% in 1992-93.

Hong Kong

In many ways, the most amazing tax system of them all was not copied from other successful reforms, but was created in 1947 by a little island country whose population has increased tenfold since then Hong Kong. Hong Kong never had a Social Security tax or value-added tax. The individual income tax has a generous tax-exempt amount of $8,300, and tax rates are 2- 15% above that. There is no personal income tax on dividends, interest or capital gains. The corporate tax is 16.5%, with rapid write-offs of equipment. There is a small inheritance tax and property tax.

The Hong Kong economy doubles in size every ten years or so, which is what happens when growth averages 6- 7%. Revenues have grown even faster. From 1983 to 1992, tax revenues increased by 409% in Hong Kong dollars, which are equivalent to U.S. dollars. In real terms, government consumption in Hong Kong has increased by 6% a year for many years. Yet government nevertheless remains small as a percentage of GDP, because GDP has grown just as rapidly as government. Everybody wins.

Lithuania: Toward a Constructive, Competitive Tax System

After this background in theory and experience, we can now apply the lessons to Lithuania.All of the major taxes in Lithuania have many "loopholes." They are far from neutral. Some sources of individual and corporate income are taxed at much different rates, even if the amount of income is the same. Purchases of some goods and services are taxed at very high rates, while others are not taxed at all. All of this political favoritism and social engineering distorts economic decisions. Labor and capital move into "priority sectors" or occupations because of special tax treatment, not because that is what consumers prefer. A tax system in which tax rates are highly uncertain, variable and dependent on the decisions of officials also creates opportunities for tax avoidance and corruption.

The tax on corporate income is 29% on businesses that employ more than 50 people, but only 8.7% to 14.5% on businesses that employ fewer than 50 people and also report low earnings. This encourages small companies to avoid expanding if that would involve hiring more than 50 workers. It also encourages such companies to keep costs higher and earnings lower. A foreign-owned business pays 30-50% as much tax as a Lithuanian business for five or six years, and sometimes zero for three years. Such "tax holidays" reduce revenue, requiring higher tax rates. If foreign companies are planning on remaining in Lithuania for more than a few years, they should be moreconcerned about the permanently higher tax than about the temporarily lower tax. Besides, why favor foreign business over domestic business? Both are equally useful, and neither can be taken for granted.

The individual income tax rate is 33-35% for employment income. But the tax is only 24% for income from an unincorporated enterprise, 20% on income from renting a house or lending money, 13% from an author's royalties, 10% from the sale of property and 5% from the sale of berries or nuts. If two people have the same income, why should one be taxed at a lower rate than another depending on how the income was earned? There is no legitimate reason to exempt income from pensions, alimony, scholarships, lottery winnings, blood donations, and so on. Income is income.

Little Steps

I will make a few specific proposals about how to improve the tax system in Lithuania. This is not meant to result in a perfect system, just some simple improvements.

1. Broaden the base of the VAT to include everything except financial transactions. The VAT currently exempts publishing, imported services, education, and medical goods, among other things.

2. Reduce the highest personal income tax rate to no more than 25%, regardless of the source of income. This would merely be comparable to Estonia and Latvia.

3. Impose a uniform 25% income tax on all business, foreign or domestic, large or small, corporate or not. This would also be about the same at Estonia and Latvia.

4. If businesses deduct interest payments but not dividends, then individuals should pay tax on interest income but not on dividends. Dividends have already been taxed at the corporate level. This is an easy way to eliminate double-taxation of dividends (which encourages business debt).

5. Put a ceiling on maximum tariffs, and announce that the ceiling will be lowered in stages, such as 40% in the first year, 30% in the second, and 20% in the third. Tariffs account for very little government revenue, and have nothing to do with the size of the current account deficit (which depends on the gap between investment and domestic savings, and is not necessarily a problem). High tariffs raise the cost of production and the cost of living, making a country poorer. High tariffs also artificially limit competition, and promote smuggling and corruption.

6. Develop alternatives to tax-financed pensions (social security), in order to avoid making older people too dependent on younger workers. This would involve allowing individuals to manage their own savings accounts - in which interest, dividends and capital gains could accumulate without paying tax. One plan could be similar to the U.S. 401k plans, in which employers voluntarily match employee contributions to their own portable retirement accounts. Something like America's new "Roth" retirement account would have to be available for self-employed workers. Such savings plans would also help to develop capital markets. And they would reduce the current account deficit, and the need for foreign capital, by increasing domestic savings.

Conclusion

Economists from the biggest industrial countries often advise other countries to pursue policies that are the opposite of what they would advise at home, such as sinking their exchange rates and raising their tax rates.

One reason may be the common opinion that the tax system can and should reduce high incomes. But taxes do not "redistribute" income, they just reduce income. An OECD survey found that, "the evidence for almost all countries suggests that the tax system overall has relatively minor effects on the income distribution." Efforts to levy high marginal tax rates on the returns to investments in capital or human capital just make capital and skills more scarce and therefore more valuable. That tax-induced scarcity raises before- tax returns for those who already own capital or professional college degrees.

The true incidence of graduated tax rates is shifted to workers (whose productivity is held back by a lower ratio of capital to labor). And the burden is also shifted to consumers (who must pay higher fees to professional people with scarce skills).

There are practical limits on how high taxes can go as a share of GDP except perhaps by keeping revenues the same and lowering GDP. Countries with the highest tax rates on incomes, payrolls, sales or imports, do not collect any more revenue than countries with much lower tax rates. In a dynamic analysis -- which looks at growth of real tax receipts over time rather than just the static percentage of GDP in one year -- the high-tax countries fare quite badly. What all high-tax countries have in common is little or no economic growth. So, their tax revenues do not grow either.

Since revenues cannot possibly keep rising each year as a percentage of GDP, the only way to keep revenues rising in real terms is to keep GDP rising in real terms. Tax policy's longest lasting impact on the budget is its effect on the economy.

The most successful "economic miracles" of this century always avoided wage and price controls, guaranteed to convert the currency to a more-credible currency at a fixed exchange rate, and reduced tariffs and marginal tax rates. Whenever combined marginal income, payroll and sales tax rates have been reduced to internationally competitive levels, this has resulted in (1) substantially increased net capital inflows and therefore a stronger currency and lower interest rates; (2) reduced emigration of skilled people ("brain drain") and increased personal investment in education; and (3) reduced tax evasion, more rapid economic growth and therefore increased real tax collections from all sources.

If you think small, you get small results. With the right policies which certainly must include a bold change in tax policy -- the Lithuanian economy should be able to grow by about 7% a year for quite a long time. And that is enough to double real output and income in ten years. It won't just happen. It takes a lot of work. But making work pay is a very good place to start.

This paper was delivered to the conference "Pro-liberal Reforms: Lifting the Tax Burden," 5 December, 1997