JEL Code: H2, K34, O23

Presented at the 14th Conference of the Business Association of Latin American Studies, Rio de Janeiro, Brazil, April, 1997

Any system of public finance is based on certain principles. Some principles are better than others and some principles conflict with others. Below is a brief discussion of some of the major principles of public finance from both a utilitarian and a rights perspective. Tax systems (more properly, systems of public finance) may be based on coercion or voluntary exchange. They may charge high or low rates. They may be based on the ability to pay principle or the cost-benefit principle. Rates may be uniform -- the same rate for everyone -- or discriminatory. The system may be easy or difficult to administer. Taxes may be hidden or visible. They may be easy or difficult to collect. Collections may go into the general fund or may be earmarked for specific uses. Collections may produce a steady flow of cash or a sporadic and unpredictable flow. The system chosen may produce minor or major distortions to the economy. The system chosen will inevitably have some effect on competitiveness and economic growth. Rules may be either simple or complex, clear or vague. And the rules may be stable over time or may change frequently. The tax base chosen may be either wide or narrow. And the system may result in major or minor effects on behavior and incentives and social harmony. The system may have high or low administrative costs. And the amount of revenue raised from a specific tax may, at times, be less than the cost of administering the costs of collection.

These characteristics and attributes are not mutually exclusive. Indeed, many of them coexist in the tax systems of all modern Western democracies. This section discusses these characteristics and attributes and, at times, attempts to determine which attributes are to be preferred to others.

In a liberal democracy, one goal is to maximize individual liberty and minimize the amount of encroachment on that liberty by the state. This philosophy can be applied to the means government uses to raise revenue. Government can raise revenue in one of two ways, either through coercion or voluntary exchange. Income, property, estate and gift taxes are examples of coercive forms of taxation because those paying have no choice about it. They must pay or face possible punishment. Voluntary forms of revenue raising include lotteries and user fees because coercion is absent.

It might be argued that taxation is not really coercive because voters, somewhere along the line, have consented to be taxed. But there are a number of flaws in this line of reasoning. For one thing, the voters who consented to be taxed did so sometime in the past. In the case of the individual income tax in the United States, for example, they gave their consent in 1913. Many of the people who gave their consent then are now dead. And many of those who were alive and of voting age back then did not give their consent.

It is a fundamental principle of both common law and basic justice that one person cannot be held for the contract of another, so even if consenting to be taxed is viewed as a contract between citizens and the state, the contract is null and void as far as those who did not consent are concerned. So taxation cannot be said to be noncoercive just because some group of voters agreed to be taxed sometime in the past (Jefferson 1984: 1280).

Another problem regarding the obligation of citizens to the state involves the concept of majoritarianism. Under majoritarianism, if 51% vote some item into law, the other 49% must go along with it whether they want to or not. Majoritarianism is a basic weakness of democracy, but it is endured so that democracy can function. It is a pragmatic compromise. If unanimous consent were required for everything, governments would not be able to pass many laws (which some say is not such a bad idea). But the fact that a law is passed by a majority does not mean that it is not coercive, because a large minority might disapprove of the law. Government is force, regardless of how many individuals might have been in favor of passing a particular law. And in cases where democracy is representational rather than direct, it sometimes happens that the peoples' representatives pass a law of which the majority of citizens does not approve.

The concept of majoritarianism is tied in with the concept of a social contract, the view that some group of people have somehow entered into a contract with the state and that they can bind the entire population. They give up a part of their freedom in exchange for certain benefits which the state can provide. Some social contract theorists (like Hobbes) would argue that individuals give up all of their rights in exchange for protection by the state. Others (like Locke) would argue that individuals give up some of their rights, and that they can reclaim these rights if the state fails to do its job (Hobbes; Locke; Rousseau)

Some commentators would go farther than Jefferson and Locke and assert that there is no such thing as a social contract, and that such agreements, even if they did exist, are not binding on anyone who did not agree to their terms (Spooner).

In a society where freedom and private property are valued, voluntary forms of revenue raising are superior to coercive forms. If government is viewed as the servant rather than the master of the people, then coercion is to be minimized and the possibilities for voluntary exchange maximized. Forms of revenue raising (like lotteries and user fees) that do not depend on coercion are to be preferred to forms of revenue raising that rely on coercion. And tax laws that expire after some period of time are to be preferred to tax laws that remain on the books until repealed, because of the consent factor. Tax laws should be subject to review after a few years, and the legislature should be required to decide, periodically, whether the tax laws should be renewed. While this requirement would not solve the consent problem completely, it at least would be a step in the right direction, since the degree of consent that any law enjoys declines with each passing day, as a portion of the citizenry dies and new citizens are born.

One argument that has been made against voluntarism in the area of government finance is that the government cannot raise all the funds it needs through voluntarism. Coercion is necessary for the state to function. While it would take a book to explore this issue, a few points can be made here. For one thing, this line of reasoning is pragmatic rather than philosophical or ethical. This pragmatic view basically holds that coercion is necessary to raise the necessary funds, therefore coercion must be used. Fairness, equity and property rights are totally absent from this line of reasoning.

Even if one concedes the point that coercion is needed to raise the "necessary" funds, one must still ask "how much is necessary?" If the goal of a free society is to minimize coercion and allow maximum room for individual choice, then government expenditures must be kept at a minimum, so that the amount of coercion needed to raise funds is minimized. Thus, the role of government must be minimized.

As it turns out, most of what government does can be done more efficiently and cheaper by the private sector, as numerous studies have pointed out (Poole; Fitzgerald; Bennett & Johnson; Block) While minimizing the role and size of government is a worthy goal, it is easier said than done. But regardless of the size of government, the tax system can be used to reduce the amount of coercion if it is decided to use voluntary means of raising revenue first, and resort to coercive methods only to make up the difference.

A tax system may impose either high or low tax rates. High tax rates are a major disincentive to production, savings and wealth accumulation. For example, if a painter in the 90% tax bracket is offered a commission to paint a portrait for $20,000, she will only be able to keep $2,000 for her efforts if she accepts the work. In effect, she has the option of working for below market prices or not working at all (Skousen 159). It is reasonable to expect that, in some cases, she will decide not to take the work. So the painting will never come into existence. Or perhaps she will take the commission but not declare the income, which increases the amount of tax evasion in the economy and decreases respect for the law.

High tax rates in the area of estate taxes discourage savings and investment. If 90% of one's assets are confiscated by the government at death, there will be a tendency to spend it all now and not leave much for the heirs. Capital accumulation suffers as a result, which has a negative effect on economic growth and the standard of living. Of course, one might give it to the children and grandchildren before death, but this route may not be possible if a gift tax takes a large chunk of the gift.

The rate of tax also has effects on the economy. Money taken out of the private sector cannot be used to invest in business expansion, cannot be saved, or spent on consumption items. A number of studies have concluded that high tax rates retard economic growth and that low rates encourage it (Reynolds 1985 & 1992; Rabushka & Bartlett; Wanniski; Marsden; Robbins & Robbins). Economic growth cannot come from raising taxes. All that raising taxes does is redistribute previously created wealth. So if maximizing economic growth is a goal, minimizing taxes must be part of the way to achieve that goal. Thus, it follows that lower rates are better than higher rates.

There is also a relationship between economic efficiency and the tax rate. If rates are high, businesses will stand to lose less if they make a bad investment decision because they can deduct the loss on their tax return. So there is a tendency to be less careful about making investments.

Low tax rates are to be preferred to high rates for a number of reasons. Low rates result in less distortion to the economy because less money is being taken from producers and redistributed to tax consumers. Resources are not being squandered on tax shelters that make no economic sense. Low rates aid in economic growth because more money is available for investment. Low rates are more fair than high rates because low rates involve less confiscation of personal property.

Ability To Pay or Cost-Benefit

There are two basic and diametrically opposed views of taxation. Those who favor the ability to pay approach view the state as a master, who extracts tribute from its subjects on the basis of how much they are able to pay. Those who take this first approach often also view the state as a benevolent father figure, who distributes tax benefits on the basis of need. In Karl Marx's words, "From each according to his abilities; to each according to his needs." (Marx)

Those who take the cost-benefit approach view the state as the servant of the people. Government provides services and taxpayers pay for the services. Those who benefit the most from the services should pay the most. And those who do not use a particular government service should not be forced to pay for it.

The ability to pay principle involves exploitation. The government exploits the producers -- the wealth creators -- and redistributes a portion of their income to wealth consumers -- those who use the various government programs. In that sense, it is a parasitical system. The cost-benefit principle, on the other hand, attempts to match costs with benefits. Those who use government services pay for them, and those who do not use the services are not forced to pay. Thus, the cost-benefit principle appears to be more fair than the ability to pay principle because it is based on principles of equity rather than exploitation. So a tax that is based on the cost-benefit principle is to be preferred to a tax that relies on the ability to pay principle, all other things being equal. However, it should be pointed out that both taxes raised under the ability to pay principle and under cost-benefit are generally raised by the use of force or the threat of force, except in the case where the method used is a user fee or lottery. So both methods suffers from defects, although the cost-benefit approach at least attempts to match costs and benefits equitably, whereas the ability to pay concept makes no such attempts.

Uniform or Discriminatory Rates

A tax system can be structured to have either uniform or discriminatory rates. There are basically three options. A tax that has discriminatory rates will take a larger percentage from some groups than others. An example of a discriminatory system would be a graduated personal income tax, which takes a larger percentage of marginal income from those who earn more, and a lesser percentage from those who earn less. For example, a graduated system might take 10% of taxable income between $0 and $10,000; 15% of taxable income between $10,001 and $20,000; 25% of taxable income between $20,001 and $35,000; and 40% of taxable income over $35,000. Such a system can be complex, with 10 or more different rates, depending on income level; or it can be relatively simple, with just two or three rates.

A graduated system discriminates against those who earn more. Such a system is based on the ability to pay philosophy rather than the cost-benefit principle. As was mentioned above, the cost-benefit principle is to be preferred to the ability to pay principle, so a graduated tax is less desirable than one that has the same rate for everyone. This is so not only because of fundamental fairness, but also because it is simpler, and simplicity is to be preferred to complexity.

Economists have given a number of utilitarian reasons against the graduated income tax as well (Blum & Kalven; Hayek). For one thing, it destroys the incentive of the most productive people. And since it is primarily the most productive people who save, invest and create jobs, a graduated tax will retard economic growth by reducing the amount of capital available for investment.

The second option is a tax that has uniform rates. An example of a tax having a uniform rate would be the flat rate income tax (Hall & Rabushka; Rothbard 1988). One advantage of the flat tax is that it is very simple to compute. It has only one rate, not several. Everyone pays the same rate. If the flat rate is 10%, then someone with $10,000 in taxable income will pay $1,000 and someone with $1,000,000 in taxable income will pay $100,000. Thus, the rich will pay more in total than the poor, although they will pay the same percentage.

One might argue that it is only fair that the rich pay more in total than the poor, because the rich have more property to protect than the poor. Since protecting property is a function of government, it seems only fair that those who have more property will pay more, in total, for government services than those who own less property. This line of reasoning seems reasonable on the surface. But if one digs beneath the surface, problems start to appear with this line of reasoning. For one thing, there may be little or no relationship between the cost of protecting property and the amount of property to be protected. It may cost more to protect 100 acres of land worth $10,000 than to protect a bank vault that contains $100 million (Rothbard 1970). But at least a flat tax makes some attempt at equity, whereas a graduated tax is grounded in the theory of exploitation.

Another advantage of having a single, flat rate is that it does not penalize those who are more productive, so it does not have as much of an adverse effect on incentives and capital formation (especially if the rate is low). It also does not have as much of an adverse effect on social harmony, since it does not pit one class against another -- the rich against the poor.

A third option would be a tax that charges everyone the same amount. An example would be a poll tax or a head tax (Cullis, Jones & Morrissey). For example, if the total cost of providing government services in a particular community is $1,000,000 and there are 10,000 people in the community, then every individual would be assessed a tax of $100. This form of tax is in keeping with the cost-benefit principle. Assuming everyone benefits equally from government services, then everyone should pay the same amount without regard to income. Otherwise, there are free riders and parasites.

This form of tax is the closest approximation to a market system. For example, if you go into the store to purchase a loaf of bread, the person behind the counter does not ask you what your income is before quoting a price for the bread. The price you pay for the bread is not based on a percentage of your income. The price is the same for everyone regardless of income level. And so it is with the poll tax. But the poll tax is not neutral because some people benefit from government services more than others. And if everyone received $100 worth of services in exchange for a $100 poll tax, why have a tax in the first place? Why not just abolish taxation and eliminate the middle man (government) by letting individuals pay for the services directly?

Easy or Difficult to Administer

A tax may be either easy or difficult to administer. Needless to say, a tax that is easy to administer is preferable to one that is difficult to administer. And while such an answer might seem obvious, in practice, many of the taxes now being imposed in the so-called advanced Western democracies are difficult, rather than easy to administer. For some reason, complexities have crept into the system, which make tax administration much more difficult than need be. Emerging democracies should be aware of this fact, so that they can avoid similar problems when they construct their own tax systems.

Visible or Hidden

A tax can be either visible or hidden. The two possibilities seem to be diametrically opposed to each other. Which possibility is better? That depends on the goals of the individual making the determination. If the goal is to raise as much revenue as possible with a minimum of protest, then a hidden tax is to be preferred to a visible tax. If the goal is to restrain the amount of tax the government takes, or to let taxpayers know exactly what they have to pay for government services, then a visible tax is more likely to help achieve that goal.

From a strictly moral or ethical point of view, a visible tax is to be preferred to a hidden tax. Disclosure should be as full and complete as possible. Disclosure is required on corporate financial statements so investors will know something about the stock they are buying. Product labeling laws force manufacturers to disclose the contents of the package. Bankers are required by law to reveal the full interest charges on the loans they make. And people should know the full cost of what they have to pay for government. If they don't, they will not be able to make intelligent decisions at election time, when they are called upon to decide who their political leaders will be.

An example of a hidden tax would be the value added tax. The value added tax is hidden because the people who ultimately pay it -- consumers -- do not know what portion of their purchase is for the product and what portion is for the tax. As a result, they may tend to blame the company that makes or sells the product for the high price, when in fact a significant portion of the price represents a tax.

Excise taxes on alcohol, tobacco and gasoline could be used for examples of hidden taxes. Although the tax may be stated on the package or gas pump, consumers seldom read these things. The tax they pay is part of the stated price.

An example of a visible tax would be a user fee. If people have to pay $4 to enter a national park, they are made aware immediately of the cost of that particular government service. It is not hidden. But if admission is free, the real cost is hidden. Someone has to pay to maintain the park, to pay the salaries of the employees who work there, and so forth. If the cost of maintaining the park is funded through general tax revenues, those who use the park for free are being subsidized at the expense of everyone else. The result is overuse of the park. If individuals had to pay the full cost of using the facility, fewer of them would use it, which would have beneficial effects on the environment, which is being harmed by overuse of government-owned recreational facilities.

Some taxes are partly visible and partly hidden. For example, the personal income tax in the USA is collected by having the employer take the tax directly from employee paychecks before they even receive their pay. The amount of the tax taken is disclosed on their pay stub, but they do not feel the full "bite" of the tax because they never actually saw the money. The tax would be more visible if the employer did not take it out of their checks, since they would then have to write out a check to the government for the amount they owe. Such a requirement would place more pressure on politicians to keep taxes low, since taxpayers would have a better feel for what they are actually paying for government services. But it would also increase collection problems, since not everyone would pay the tax when it is supposed to be paid. But where there is a tradeoff between disclosure and efficiency, the only ethical choice would be disclosure. If the government is the servant of the people, then the people should know exactly how much their servant costs. The full cost should not be hidden, even if the tax collection process would run more smoothly without disclosure. In the long-run, a full disclosure system would probably result in lower taxes, since taxpayers who are aware of the full cost of government will have a tendency to put pressure on their representatives to keep taxes low.

Easy or Difficult to Collect

A tax may be either easy or difficult to collect. From the collector's point of view, easy is better than difficult. The ease with which a tax may be collected should definitely be kept in mind by those who are trying to devise a tax system. In most Western democracies, corporate income taxes and individual income taxes that are withheld by the employer are relatively easy to collect because corporations and noncorporate employers do the collecting, and they merely turn over the proceeds to the government. The same is true of sales taxes.

General Fund or Earmarked

Taxes can either go into the general fund or be earmarked -- set aside -- for particular purposes. Income taxes are generally deposited into the general fund, to be used for a variety of purposes. Gasoline taxes are often used to maintain highways. They are set aside for a particular purpose. Some people are presently advocating that a portion of the tax on cigarettes be used for cancer research.

Whether revenues from a particular tax are placed into the general fund or earmarked for special purposes depends on a number of factors. Oftentimes, a tax increase will be more acceptable to the general population if it is earmarked for a particular purpose that is widely accepted (Buchanan 1963; 1967). Lottery proceeds that are used for education is an example. Social Security taxes that are set aside for taxpayers' retirement is another. Segregating tax revenue for specific purposes reduces the possibility that the money will be squandered on pork barrel projects. That is one of the advantages of earmarking tax collections, and it is one of the disadvantages of dumping tax revenues into one big pot. Earmarking tax dollars is a good control device that enhances the chances that the money will actually be spent for its intended purpose. But it is not a device that can be used for every kind of government expenditure. Dumping everything into a common pot gives the government more flexibility to shift spending to where it is needed. So there is a trade-off.

Minor or Major Distortion to the Economy

From an economic perspective, the best tax is a neutral tax -- a tax that does not divert the operation of the market from the lines it would take in the absence of taxation (Mises 737). The less distortion to the economy, the better the tax, all other things being equal. Economic resources are allocated most efficiently if they are allowed to gravitate to their most highly valued uses. Anything that causes economic resources to be diverted from their optimum use increases inefficiency and retards economic growth. Economic disruptions cause a number of problems, many of which are unforeseen. The best tax is a tax that causes no distortion. However, there is no such thing as a neutral tax (Rothbard 1981) so we must settle for something less than perfect.

Effect on Competitiveness and Economic Growth

Another attribute of tax policy is the effect it has on competitiveness and economic growth. This aspect of tax policy is especially important to emerging democracies, since the transformation from a centrally planned economy to a market economy entails major shocks to the economic system. As a general rule, it can fairly be stated that emerging democracies have difficulty competing with more developed market economies, either because they do not make the products that developed economies want, or because the quality is too low and/or the price is too high. These difficulties can be greatly relieved by knocking down trade barriers and importing modern technology. Many of the emerging democracies have skilled workforces that are willing to work at competitive wages, if only the system would allow them to use their skills and energies. It is important that these emerging democracies adopt tax rules that make it easier for them to compete rather than more difficult. The tax system they adopt should promote rather than retard economic growth and development.

Which tax policies would foster economic growth and development? The tax system should encourage investment, both foreign and domestic. That means that foreign companies should not be penalized for investing in the local country and domestic companies should not find their own country's tax system more burdensome than those of other countries.

There are a number of factors that make a tax system attractive or burdensome. Companies would rather invest in a country that has low tax rates than high ones. A number of studies have shown that high tax rates retard economic growth (Reynolds 1985, 1992; Rabushka & Bartlett; Wanniski). For example, one study concluded that the tax increase President Bush signed into law in 1990 retarded economic growth by 0.7%, destroyed 400,000 jobs, caused the unemployment rate to be 0.45 percentage points higher than it otherwise would have been, and caused stock prices to drop by 15%. It estimated that each 1.0% increase in the federal tax burden leads to a 1.8% reduction in economic growth (Dunkelberg & Skarburg). Another study of the U.S. tax system estimates that modifying payroll deductions, adopting neutral cost recovery, eliminating capital gains taxation, eliminating the tax on corporate dividends and eliminating the taxation and deductibility of interest would add 1.1 percentage points to the economy's growth rate; increase gross domestic product by nearly $6.2 trillion dollars between 1992 and 2000, which translates into a roughly 13% increase in the standard of living; create nearly 4 million jobs between 1992 and 2000; increase the stock of capital by $10.4 trillion between 1992 and 2000 (Robbins & Robbins). In a model of the U.S. economy that took federal, state and local taxes into account, researchers estimated that for every dollar of extra revenue raised, production declined by $0.332 (Ballard, Shoven & Whalley). A World Bank study of 20 countries found that those with lower tax rates had a more rapid expansion of investment, productivity, employment, and government services, and had better growth rates than countries that had higher rates (Marsden).

Simple or Complex Rules

A tax system can have simple or extremely complex rules. Simple rules are to be preferred to complex rules, all other things being equal. While this conclusion comes as no surprise, and may even seem not worthy of mention, it should be noted that many modern western democracies have rules that are far more complex than needed to raise the amount of revenue that is raised. Any emerging democracy that looks to a western tax system for guidance should keep this fact in mind.

In the USA, for example, the Internal Revenue Code and Regulations run to many thousands of pages, often in small print. The regulations for many tax code sections run to more than 100 pages. But the guidance does not end there. Tax practitioners and taxpayers who want to dig to the bottom of some tax issues must also consult with the thousands of tax cases that have been decided in three different federal court systems, plus Internal Revenue Service publications and privately published tax treatises. And even then, the answers are not always readily apparent, so they have to guess at what the correct answer might be. And to make matters even more unpredictable, when IRS issues a tax regulation, it sometimes does so retroactively, so a tax rule that is issued in 1993 might take effect as of 1981.

Tax rules in the USA are so complex that even IRS agents do not understand what the law says. A General Accounting Office report found that 44% of the penalties that IRS agents assessed under the payroll tax rules were wrong. And the payroll tax rules are so complex that one-third of all U.S. employers are assessed penalties each year (Payne) But complexity is not limited to the payroll tax rules. Many other areas of the tax code are also overly complex. IRS assessed 29 million penalties in 1990, many of which were later abated (Payne).

Emerging democracies must be very careful that they do not fall into this quagmire of complex and uncertain regulations. Any tax system adopted should have built-in safeguards that protect against unnecessary complexity. The Swiss method might be used as an example. In Switzerland, the number of words that can be used is limited. Emerging democracies must also guard against a proliferation of rules and regulations. It would not be too difficult to construct a tax code that is less than 50 pages in length. And there is no need to have three different and often contradictory court systems to settle tax cases. And any tax system that is adopted should have a provision that makes regulations prospective only, to reduce the amount of unfairness in the system and increase the predictability.

Clear or Vague Rules

Laws and regulations can be written in clear, easy to understand language. Or they can be written in language that is extremely vague and obscure. If laws are too vague and indecipherable, a court may rule that the law is void for vagueness. But in other cases, the law may be held to be valid even though it has a high degree of vagueness or obscurity. Obviously, clearly written laws are to be preferred to vaguely written ones. And while this conclusion might seem obvious, in practice, the laws are often vague, especially in the area of taxation.

A good illustration of this point is the survey that Money magazine does every year. For several tax seasons since 1987, Money asks 50 professional tax preparers to complete a federal tax return for a hypothetical family. And every year, the tax practitioners, who do this sort of thing for a living, come up with the wrong answers. For its 1988 study, all 50 tax preparers got different answers to the same fact situation (Anrig 1988). They arrived at answers that computed the tax liability to be anywhere from $7,202 to $11,881, a variance of about 65%. It conducted a similar test of 50 different tax practitioners in 1989 and got an even worse response (Anrig 1989). Again, it got 50 different answers, but this time, the answers ranged from a tax liability as low as $12,539 to as high as $35,813. The 1990 study was somewhat better. Two of the 50 practitioners got the right answer (Topolnicki 1990). Well, not quite. The 50 practitioners had 50 different answers again, but this time, the Money judges decided that two of the practitioners were close enough to the correct answer to be considered error-free. The practitioners computed tax liabilities ranging from a low of $9,806 to a high of $21,216, with an average of $13,915. The correct answer was $12,038, which was about half of what some practitioners said was owed.

The comedy of errors continued for the next two Money surveys. In the 1991 survey (Tritch & Lohse 1991), one practitioner actually got the correct answer, which was $16,786. The incorrect responses ranged from a low of $6,807 to a high of $73,247, which is quite a spread. In the 1992 survey, none of the practitioners got the correct answer (Tritch & Lohse 1992), although it was perhaps the most experienced group of the five groups that took the Money test over the past five years, since Money, for the first time, instituted a minimum experience requirement -- five years (Tritch & Lohse 1992). Answers ranged from $16,219 to $46,564. The correct response was $26,619.

Stable or Frequently Changing Rules

Governments may enact tax rules that remain on the books for decades. Or they may make rules that change constantly. While it is always a good idea to change a law that is bad, or to improve one where possible, many times, governments go beyond mere improvement and engage in finetuning. In the USA, for example, Congress has enacted major changes in the tax laws practically every year since 1981. And few tax practitioners would argue that the changes have made the tax law either better or easier to comprehend. The tax rules have, in many cases, become more complex and obscure over time. And the constant changes -- sometimes retroactive changes -- have made it difficult for tax planners to plan. The business decisions that businesses must make are often influenced by the tax laws, and when the laws change frequently, they must make their plans with a higher degree of uncertainty.

Frequent changes also increase compliance costs. Tax practitioners and business owners must spend time and money to learn the new rules. And businesses must adjust their practices to comply with new rules. The cost of complying with the tax legislation that was enacted in 1987 is estimated to be $6.2 billion for 1988 alone (Genetski). The cost of complying with the 1989 tax law changes is estimated to be $0.1 billion for 1989 and $5.6 billion for 1990. Compliance costs for the 1990 Budget Accord were estimated to be $23.2 billion in 1991 and $10 billion in 1992. And few practitioners would take the position that these tax law changes made compliance any easier. The point is, compliance costs must be taken into account when a government is thinking about enacting a new piece of tax legislation.

Another problem with frequent changes in the tax laws that has an especially devastating effect on newly emerging democracies is the effect that frequent changes have on foreign investment. It is difficult enough trying to convince foreign investors to invest in a new, unproven country, without making investment returns even more uncertain by frequently changing the tax laws. Russia is only one of several examples that could be given here to illustrate the point. Not only is the tax situation in Russia a rapidly changing one, but it is also uncertain. Not even Russian government officials always know what the tax laws are, or how they should be applied.

High or Low Administrative Costs

A tax system can be structured so that administrative costs are low. Or one can be structured so that administrative costs are quite high. When one is measuring the administrative cost of a particular tax policy, or of a tax system as a whole, not only the administrative costs to government should be considered. The cost to the taxpaying community must also be taken into account. In the case of the tax system in the USA, for example, the cost of administering the tax law falls more than 99% on the private sector. The budgets of the Internal Revenue Service and the parts of the Treasury and Justice Departments that administer the tax laws was about $6 billion in 1990, whereas the costs of administration that fell to the private sector for the same period were estimated to be $618 billion (Payne), which represents about 65% of the tax collected.

Cost and Benefit

Another factor to consider when constructing a tax scheme is the relative cost and benefit, from the tax collector's point of view. The tax collected should exceed the cost of administration. But the cost-benefit comparison should not be limited to just easily traceable, tax-related administrative costs, because the cost of administering a tax system sends ripple effects throughout the economy. These indirect, ripple-effect costs should also be considered (Weidenbaum & DeFina).

The tax law should also result in more tax being collected rather than less. While this seems obvious and hardly worthy of mention, the luxury tax rules in the USA provide an example of where such is not the case (USGAO). A few years ago, the tax code started to place an excise tax on "luxury" automobiles -- automobiles that cost over $30,000. Ever since this rule was passed, auto dealers that sell high priced cars have been complaining because people aren't buying their cars anymore. They are losing business, their profit margins have shrunk, and some auto dealers have had to go out of business. The employees who were employed by these dealers are no longer paying taxes but are collecting unemployment insurance or welfare. The auto dealers are paying less taxes because they make less profit. The amount of taxes lost by this rule has exceeded the amount of tax the government would have collected from the rule. Yet Congress has not yet repealed the tax because of inertia.

The same is true for the excise taxes that have been imposed on boats, airplanes, jewelry and furs. Many boat manufacturers have been driven out of business by the excise tax, as have a few small airplane manufacturers. Rather than raise tax revenues, the imposition of these excise taxes has reduced total tax revenues because the small amounts collected for excise taxes have been more than offset by the lower amounts of income taxes that have been collected. Yet these taxes were not immediately repealed because many members of Congress perceived that repealing luxury taxes would make it appear that they are coddling the rich. In fact, they are destroying working class jobs.

Emerging economies can adopt a system that meets the goal of raising revenue without the burden of excess complexity and administrative cost, not to mention coercion and inequity.

What kind of system should an emerging economy adopt? Here are some general rules.

• Noncoercive methods should be used first. Coercive methods should be used only as a last resort. In countries that do not need to support a large defense force, and where everything that is possible to privatize is privatized, there may be no need to resort to any form of coercive taxation. Lotteries and user fees -- the only two forms of noncoercive taxation -- will be sufficient. Aside from being equitable, these methods also do not distort incentives, so economic growth is not stifled. Noncoercive taxation methods are the only tax methods that do not violate property rights.

• Where coercion is needed to raise the necessary revenue, some taxes are better than others. Countries should never resort to a tax that is based on the ability to pay principle. Such forms of taxation are inherently unfair because they are based on exploitation. Taxes that are based on the cost-benefit principle at least have some degree of "rough justice," because there is some relationship between what people pay and what they receive in government benefits. That means a flat income tax is to be preferred to a graduated income tax. And there should be a "cap," some maximum beyond which individuals are not taxed. Based on the cost-benefit principle, it is difficult to justify forcing someone to pay $100 million in taxes, because it is impossible for someone to receive $100 million in benefits from government. Perhaps $10,000 or some other figure would be more equitable.

• Where coercive taxation is used, rates should be kept as low as possible, for a number of reasons. For one thing, low rates are not as inequitable as high rates. Taxation is basically just an acceptable word for confiscation, so the fairest tax is the one that confiscates the least amount of property. Also, lower rates have a less harmful effect on economic growth and do not cause as much distortion to the economic system as high rates.

• The system that is adopted should have low administrative costs and should be simple rather than complex. Most western tax systems are needlessly complex and administratively inefficient. There is no need for such a system to be adopted by an emerging democracy, although the possibility of adopting such a system is very real, especially if the emerging democracy wants to emulate some western tax system. This urge must be resisted.

• Taxes should be visible rather than hidden. This principle is based on fundamental honesty. Tax officials who construct a system of hidden taxes are derelict in their duties. Government officials have a duty to serve the public and not to hide anything from them. Public officials are the servants and the people are the masters. Public officials who hide the true cost of government from those who pay are acting dishonestly, just as a bookkeeper who hides costs from an employer is acting dishonestly.

• Where appropriate, taxes should be earmarked rather than poured into the general fund. Earmarking acts as a control or safety mechanism and is good accounting.

• Tax rules should be clear and easily understood by the average taxpayer. Vague tax rules cause unnecessary confusion, increase administration costs in both the public and private sector, and increase disrespect for tax laws and government in general.

• Once established, tax rules should not change frequently. The attempt to fine-tune should be resisted. Frequently changing rules increase administration costs and complexity and distort segments of the economy, as well as retard economic growth. And there is no need to change the rules frequently. If policymakers construct a system that is based on sound tax principles, there is little or no need to make frequent changes.

• A tax should not be adopted if the amount collected is going to be less than the cost of administration. While this point seems obvious, there are several examples of cases where Western democracies have violated this rule. As soon as it is seen that a tax will not collect more revenue than it costs to administer, the tax should be repealed.

• Tax laws (and perhaps many other laws as well) should have a definite life. They should not go on in perpetuity without the need for reenactment. It is easier to make a law than to repeal one, so laws that expire automatically after some period of time have a built-in safety mechanism that prevents a bad law from being on the books forever. Legislatures should be forced to reconsider the laws they have passed every so often. Such a requirement also makes legislators more responsible to the people.

A country that adopts a tax system based on these principles will enhance its economic growth and minimize inequity. Confiscation of property violates basic human rights, and the tax system advocated in this paper eliminates or minimizes the amount of confiscation that must occur in order to raise the funds needed to pay for government services.