Even though I often write as though I know something about taxes, I could not possibly finish my tax return by April 15. Begging the IRS for more time has become an annual certainty, although IRS auditors must be glad to put off looking at my oversized return as long as possible.

The problem began years ago, when I made the mistake of saving and investing. That means I must now report in intricate detail each sale or purchase of stock, and the precise source of every dollar of interest income and dividends. This bookkeeping requires a dozen extra pages, several weekends and many pots of coffee. Congress does not even let me deduct the cost of that wasted time.

With investment income, you have to make quarterly guesses about how much tax you will end up owing, which depends on how your investments end up doing. It is difficult to guess the net effect of gains and losses, and impossible to predict how the alternative minimum tax is going to bite. You can't guess right, but if you guess wrong there are penalties.

It would not be that difficult to make tax reporting much, much simpler. Unfortunately, tax simplification has lately become a matter of ritual rhetoric with little substance. The discussion has been bogged down by myths left over from the Tax Reform Act of 1986:

The first myth is that bringing the top tax down from 70 percent in 1981 to 28 percent by 1988 was paid for by eliminating itemized deductions. An editorial page writer for The Wall Street Journal Europe recently wrote "in the 1980s, Ronald Reagan accomplished his tax reform by trading lower rates for fewer loopholes."

That alleged link between tax rates and loopholes is widely believed but totally untrue. After the 1986 law was phased-in, deductions were still 23.1 percent of gross income from 1988 to 1993 -- exactly the same as the deductions from 1980 to 1985. Itemized deductions were somewhat smaller, but standard deductions were larger.

Loopholes did not have to be closed because revenues did not fall. Individual income taxes amounted to 9.3 percent of personal income tax in 1983-85 (when the top tax was 50 percent) and 9.4 percent in 1988-90 (when the top tax was 28 percent). The highest tax rates were subsequently increased in 1991 and 1993, but the income tax then dropped to 8.9 percent of income from 1991 to 1994. A top tax of 28 percent tax, while it lasted, paid for itself through increased labor force participation and reduced tax avoidance.

A related "Myth of '86" is that itemized deductions were a major source of tax complexity, so the extremely complicated 1986 law was thus said to be a big move toward simplification. From this, it is now said to follow that "true" tax reform is impossible simply because people prefer their complicated itemized deductions to the simplicity of standard deductions.

This is just politically convenient nonsense. It is not difficult to open an envelope and copy the amount spent on property tax and mortgage interest. What is difficult is reporting every tidbit of income and expense related to investments or a small business. We are required to "itemize" too many things as individual income, not too many deductions. Defining tax reform in terms of what is deducted -- rather than what is taxed -- is like being unduly fond of the wrong end of the horse.

The third myth was that a system with two tax rates would be inherently much simpler than a system with four or five. This notion apparently arose from a misunderstanding of the many complex benefits of a genuine flat tax. People simply jumped to the conclusion that if a single tax rate was better than two, then a system with two tax rates must likewise be much simpler than system with four.

There are many advantages of a single rate, but simplicity is the least of them. In his book Economics of the Public Sector, Joe Stiglitz notes that a flat tax "means that income can be taxed at it source: taxing income at is source will reduce compliance costs and increase compliance rates." Once a single rate is abandoned, Stiglitz adds, there is no reason to prefer two rates to three or three rates to four. A flat tax of 40 percent on salaries is not obviously preferable to a graduated tax of 15, 20 and 25 percent.

On the other hand, a single tax rate is absolutely essential, in terms of economic efficiency and simplicity, when it comes to returns on business investment. The United States has three corporate tax rates, with the lower 15-25 percent rates phased-out in a way that imposes the highest marginal rate on companies with puny profits.

Then we add all sorts of extra tax rates on the individual returns from corporate profits. Capital gains taxes vary from 18 percent to 35 percent, depending on how much time passed before it made sense to sell the asset and pay the tax. That is a policy no economist could possibly explain, much less defend.

Half of all dividends pay zero, the rest pay five different rates. Congress is currently wrangling over how much "the other half" should pay, as though that has nothing to do with the fact that half pay nothing. The amount of interest expense deducted far exceeds the amount of interest income taxed at various rates. Capital taxes are an incoherent mess, causing maximum damage for very little revenue.

The main reason the individual income tax is so complicated is that it tries to collect different tax rates on each dollar of investment income, depending on who earned it and how. If such income was taxed at a single rate, a reasonable tax could easily be collected at the source. With a flat tax of 15-20 percent on interest income, for example, banks and brokerages could simply withhold that much before sending you the check.

It is only because we try to charge Smith a higher tax than Jones on the same amount of investment income that individuals are required to report reams of information about investment returns.

It doesn't work. Taxpayers can often arrange their affairs to keep assets in tax-exempt funds and foundations, or in the hands of relatives in low tax brackets. Half of all dividends, three-fourths of all capital gains and a big share of interest income are untouched by the individual income tax.

A low, flat tax on investment income is the simple way to simplify. Zero sounds like a nice round number, but I might settle for something less simple than that.