Business
Navigation Bar
Navigation Bar



  Tax Statistics 'Prove' A Host of Positions

By Albert B. Crenshaw
Washington Post Staff Writer
Sunday, February 21, 1999; Page H08

"There are three kinds of lies: lies, damned lies, and statistics," British statesman Benjamin Disraeli is supposed to have said.

And he'd never even seen the U.S. income tax.

Had Disraeli been around for the recent debates over U.S. taxation, he would have seen his aphorism turned into a way of life, though in today's debates, lies and damn lies are actually rather rare. Statistics are the small-arms ammunition of the battle.

From the capital gains rates to individual retirement accounts to the current debate over a general tax cut, contestants are almost always able to marshal impressive arrays of numbers that support their positions. Sometimes they even use the same numbers, asserting they prove different things.

For example, Rep. Bill Archer (R-Tex.), chairman of the House Ways and Means Committee, issued a news release last week saying that 8 percent of taxpayers pay 62 percent of federal income taxes. This means, he said, that people who work hard and move up the economic ladder "are entitled to relief."

Opponents quickly rushed to the faxes and phones to point out that, therefore, in an across-the-board tax cut, the "richest" 8 percent of the taxpayers will get 62 percent of the benefit.

"A lot of this involves starting with your conclusion, which is usually a function of your political views, and then going out to find facts to support it," said one former congressional staff member.

Most tax debates involve two questions: Who is helped or hurt by a particular proposal, and how it should be paid for. And there are a number of ways to make it obvious your side is doing the most good for the most people at the lowest price. Among them:

Choosing the right time period. The impact of tax changes often shifts over time, especially as to whether they gain or lose revenue, but Congress requires revenue projections going out only five or 10 years. Thus a provision that may really be a revenue loser can be used to pay for another loser because the first one brings in more than it loses in the first few years.

For example, last spring when Senate Finance Committee Chairman William V. Roth Jr. (R-Del.) found himself short of revenue to pay for provisions he favored in the Internal Revenue Service restructuring bill, he proposed a change in the rules on the income limits for converting traditional individual retirement accounts to Roth IRAs.

The change greatly benefits older people seeking to use their IRAs as estate-planning devices, and it was expected to cause a rush of conversions and--because those conversions are taxable--a revenue surge. Roth made the provision effective in 2005 so the extra revenue would be in the budget "window."

Democrats howled that the change would lose $46 billion over the following 10 years and would remain a loser ever after. But Roth was within the rules; the provision was approved, and the bill passed.

Using individual taxpayers' situations or aggregate tax burdens. An unusual feature of the current debate is the general agreement on one fact: that total federal tax collections are the highest they have been since World War II.

Republicans and conservatives argue this shows that the tax burden is excessive.

"We do have a record-high tax burden right now. The reason is this is what you get when you have a progressive tax structure and you have a prolonged period of growth. People are getting kicked into higher brackets," said Stephen Moore of the Cato Institute.

The Tax Foundation, in a widely quoted figure, says a family with the median income for two-earner families pays 38 percent of its income in taxes--federal, state and local.

Not so, say others.

Whether the tax burden is too high is a matter of opinion, but the data show it has declined for most people in recent years, they say.

William G. Gale of the Brookings Institution has studied figures collected by both the Treasury Department and the Congressional Budget Office for families at the median income--about $55,000--and half the median income, and both have seen their federal taxes decrease during the past 20 years. By some measures, income taxes for families at half the median income are now negative, thanks to the earned-income and child credits.

Using percentages or dollar amounts, whichever looks better for your side. For example, Archer last week pointed to congressional Joint Tax Committee figures showing that the "average individual income taxes paid" by taxpayers earning $40,000 to $50,000 amount to $3,566, which he noted is about four times the $935 paid by people earning $20,000 to $30,000.

But critics grabbed their calculators and reported that on a percentage-of-income basis, the higher-income group pays about 8 percent and the lower about 4 percent--a ratio of 2 to 1 instead of 4 to 1.

Choosing what to count as income and what to count as taxes. Economists differ over the best measures of income, and there is even some disagreement over what counts as a tax or who pays it.

For example, the payroll taxes collected to pay for Social Security and Medicare are paid half and half by employer and employee. Interest groups seeking to show a high tax burden often assign both halves to the worker on the theory that it comes out of pay that would otherwise go to them. Many economists say that's defensible as long as a corresponding entry is made on the income side. But sometimes corporate taxes, estate taxes and even user fees turn up in the tax burden of the middle class in some calculations--a proposition critics find dubious.

Referring to income taxes alone or including other federal taxes in the discussion. The income tax isn't the only tax Uncle Sam levies. The payroll tax is levied on wages up to $72,600 for Social Security and all wages for Medicare, so that it represents a great share of total tax for lower-income workers. There are also various excise taxes--on gasoline and cigarettes, for example--that add up, especially for lower-income people, at least as a percentage of their tax bill.

So when Archer pointed to how much more income tax is paid by higher-income people, Citizens for Tax Justice issued a table showing that "for taxpayers in income groups below $100,000, most of the amount paid in federal taxes reflects non-income taxes."

Critics of that analysis reply that for the payroll tax, at least, taxpayers are buying a specific future benefit that in the case of lower-paid workers is sharply skewed in their favor (since Social Security accounts for a greater share of their earnings).

While much of the debate, especially the parts that involve specific provisions of the tax code, is driven by narrow self-interest, many of the theorists do try to arrive at a correct picture of the tax burden and what the best policy is for the country.

The problem is that "correct" is in the eye of the beholder, and with so much data so susceptible to different interpretations, even a place less cynical than Washington would likely come up with sharp differences.



© Copyright 1999 The Washington Post Company

Back to the top

Navigation Bar
Navigation Bar