Tax Riddle Cloaked In Enigma

Regardless of what proposal President Bush's tax reform commission comes up with, the principal debate about it inevitably will be how much more or less will each income class pay relative to current law. While a valid consideration, the overwhelming emphasis on distributional effects tends to push other important tax issues off the table.

To begin with, there are two different concepts of distributional equity: horizontal and vertical. But only the latter ever really comes up in congressional debate. Horizontal equity has to do with treating equals equally. In simple terms, if you and I have roughly the same income, we should each pay about the same taxes. Vertical equity has to do with ensuring that those with a greater ability to pay - i.e., the rich - pay a higher share of their income in taxes than the poor.

Other important principles of taxation are simplicity and efficiency. Ideally, we should have a tax system that the average person is capable of understanding and complying with, which does not involve excessive complexity or demand professional assistance. Obviously, our current tax system falls far short of this ideal. Growing numbers of moderate-income taxpayers require the aid of accountants and expensive tax software just to file their returns. And even tax professionals increasingly find themselves baffled by confusing and contradictory provisions of the law.

Tax efficiency involves taking money out of the economy in a way that discourages as little output as possible. Every tax discourages some production over and above the tax itself. Economists call this the deadweight cost of taxation. Some taxes are known to discourage more than a dollar of economic activity for every dollar raised. Others impose a cost of just a few cents. A common estimate is that the federal tax system as a whole has a deadweight cost of about 20 cents per tax dollar.

Economists know fairly well how to achieve reasonable equity, efficiency, and simplicity in the tax system. The problem is that it is extraordinarily difficult to achieve all these goals simultaneously, because they are inherently in conflict with each other. For example, any effort to increase efficiency will tend to be at the expense of vertical equity, because it will necessarily involve reducing taxes on capital, the income from which mostly goes to the wealthy.

Distributional problems are compounded by the practice of using snapshots of tax changes for individual years that assume everything stays the same except taxes. Distributional tables also use certain conventions about what constitutes "income" that are at odds with the way taxpayers themselves understand the term. In general, distributional methodology makes taxpayers look richer than they are by imputing to them income that they never really see, such as health and pension benefits, thus making tax cuts appear to be more tilted toward the wealthy than is actually the case.

A key way that benefits to the rich are exaggerated has to do with capital gains. Distribution tables always assume that the same capital gains would be realized regardless of how they are taxed. But obviously, people will realize more gains if the rate is reduced, thereby paying more taxes. According to a new Treasury Department study, between 1996 and 2000, capital gains revenues more than doubled even though the maximum tax rate fell to 21.19% from 29.19%. Yet even though the rich paid most of these higher taxes, distribution tables showed them getting a huge tax cut.

Finally, distribution tables exclude the higher incomes that may result from growth oriented tax changes and implicitly assume that people are in the same income class year after year. But we know that in the real world peoples' incomes tend to rise with age because of real economic growth and because workers gain knowledge, experience, and savings that yield promotions, higher wages, and investment returns. And many of those with low incomes are well-to-do elderly spending down their assets in retirement.

A new study by the Congressional Budget Office illuminates the high degree of income mobility in America by looking at the same group of individuals over a long time period, what economists call a panel or longitudinal study. This avoids the problems inherent in traditional snapshot analyses and lets us see how behavior changes in response to tax changes.

The CBO found that incomes are much more evenly distributed when the same people are examined over time than when their incomes are examined only in isolated years. One finding is that those in the second quintile (20%) of households had an income 26% higher when viewed over a 14-year period than when measured on an annual basis. In general, the study shows that the poor are richer, the rich are poorer, and tax cuts for the wealthy impact many more people than is commonly believed.

Mr. Bartlett is a senior fellow at the National Center for Policy Analysis.