Tax revenue mysteries

Alan Reynolds
|
Posted: Jul 13, 2006 12:01 AM

Federal tax receipts continue to soar, with the individual income tax now expected to rise by about 15 percent this year and the corporate tax by 20 percent.

A year ago, when people noticed that tax receipts were going to be up by 15 percent, New York Times columnist Paul Krugman tried to dismiss it as "a temporary blip." That card can't be played twice.

The Bush team, on the other hand, has an odd habit of describing reductions in marginal tax rates as "tax relief." But the phrase "tax relief" surely implies that taxpayers are paying less in taxes when, in fact, some are paying much more.

A White House press release boasts that "since the beginning of 2003, real GDP growth has averaged 4.0 percent per year, exceeding the post-World War II average of 3.4 percentage points a year." That 3.4 percent average includes recessions. But the new budget projections estimate that economic growth over the next three years will slow to 3.2 percent.

After seeing the estimated budget deficit for the current fiscal year slashed by 28 percent in just six months, how much confidence can we have in budget estimates that stretch out 10 years or more? In a paper presented at a U.S. Treasury conference in 2004, I noted that "budget forecasting errors follow a cyclical pattern, becoming too optimistic near cyclical peaks and too pessimistic in the early stages of recovery, such as 1984, 1994 and 2004."

In January 1993, the Congressional Budget Office (CBO) estimated the deficit would hit $653 billion in just six years. When we got to 1999, the actual figure was a $126 billion surplus. That wasn't because of higher tax rates in 1993. Individual tax receipts were a smaller share of GDP in 1993-95 than in 1988-90, and no higher than now. In the fall of 1994, the CBO continued to overestimate the actual deficit for the year 2000 by 5.3 percent of GDP -- a $520 billion exaggeration for a single year.

As time passed, such gloomy estimates had to be repeatedly revised. The 1993 CBO estimate of a $579 billion deficit by 2002 was reduced to $349 billion after two more years and to $188 billion after four. By August 2001, the CBO projected a surplus of $176 billion for fiscal 2002, oblivious to an ongoing recession.

Official bean counters frequently make huge errors in estimating tax revenues and budget deficits even six months ahead, much less six years ahead. But that should not excuse political failure to deal with an unsustainable 8.6 percent increase in federal spending here and now. Long-term budget projections are a relatively new innovation. All they appear to have accomplished is to excuse political procrastination about future problems and indifference to current spending. After all, what's another few billion for the latest trendy crisis, or a few hundred billion for another war, when compared with the huge entitlement crisis that lies ahead?

Revenues from the individual income tax are now estimated to be 8 percent of GDP -- exactly the same as during the first term of the Clinton presidency, 1993-1995. Tax receipts from stock options and capital gains were higher during the 1997-2000 stock market boom, of course. But that had nothing to do with tax policy, except that the capital gains tax was reduced.

The president's critics claim the rapid increase in individual tax receipts proves the rich are getting even richer -- simultaneously claiming the top 1 percent are paying a huge share of all taxes yet are not paying nearly enough. They invariably rely on vague allusions to estimates from economists Thomas Piketty and Emmanuel Saez which, in reality, just show the top 1 percent of tax returns (not households) received 14.8 percent of income reported on tax returns in 2002-2003 -- up insignificantly from an average of 12.9 percent from 1988 to 1995. The top 1 percent means 1.4 million, not just 500 corporate CEOs. Piketty and Saez recently offered a higher "preliminary" guesstimate for 2004, but the required IRS data are not yet available. That is why Congressional Budget Office estimates stop with 2003.

The CBO estimates show the top 1 percent receiving 11.5 percent of after-tax income in 2002 and 12.2 percent in 2003, compared with 13.2 percent in 1986 and 12 percent in 1988. The CBO also estimates that the top 1 percent devoted 22.3 percent of their income to the federal individual income tax in 1980, when the top tax rate was 70 percent on dividends, 50 percent on salaries and 28 percent on capital gains. The top 1 percent paid 20.8 percent in 2003, when the top tax rate was 35 percent on salaries and 15 percent on dividends and capital gains. And those top taxpayers reported much more of their income in taxable form in 2003 (as opposed to tax-exempt bonds and tax-free perks in 1980). As explained in my recent column about renaming the Laffer Curve, punitive tax rates yield nothing but rampant tax avoidance and a sick economy.

In terms of these new tax revenue estimates, the key point of the Piketty-Saez estimates is that business income accounted for 27.7 percent of the income attributed to the top 1 percent in 2003, compared with only 7.8 percent in 1981. As individual tax rates came down in 1982-88, more business chose to shift from the corporate income tax to the individual income tax, often as Subchapter S corporations or LLCs (limited liability companies). That was a mere bookkeeping change, yet it shows up in tax return data as a big increase in the highest individual incomes.

The newly anticipated 15 percent increase in tax revenue from the individual income tax is widely attributed to capital gains (despite a mediocre stock market) and to executive stock options and bonuses (despite the tiny number of top CEOs). Yet the larger part of it is more likely due to business profits from Subchapter S corporations (including banks), LLCs (including most hedge funds), law partnerships and the like. If the highest individual tax rates were pushed much higher, much of the income now attributed to the top 1 percent would quickly vanish -- moving back inside the corporate tax, where it was in the 1970s. That would make reported top 1 percent incomes appear much lower in the individual income tax data, yet that too would merely be another statistical illusion.