Last year, for the first time, the poor were more likely than the rich to have their tax returns audited, new Internal Revenue Service data compiled by Syracuse University researchers shows.
The I.R.S. audited 1.36 percent of all tax returns filed by people making less than $25,000 last year, compared with 1.15 percent of returns filed by those making $100,000 or more. Since 1988, audit rates for the poor have increased by a third, from 1.03 percent, while falling 90 percent for the wealthiest Americans, from 11.4 percent.
The need to audit the highest income taxpayers is somewhat less today than in 1988, I.R.S. officials say, because more of these individuals are wage earners whose pay is fully reported by employers and because Congress has eliminated some of the deductions most likely to be abused.
But there has been no reason to reduce audits of corporations and the self-employed, which fell to record low levels last year because Congress did not authorize funds to keep up with the increased number of such taxpayers. These taxpayers received less scrutiny even though the General Accounting Office, the investigative arm of Congress, said in 1997 that they are more likely than the working poor to pay less in taxes than they owe.
The focus of I.R.S. audits on the bottom of the income ladder extended to businesses as well.
In 1999, unincorporated businesses with less than $25,000 in sales, classified by the I.R.S. as Schedule C enterprises, were more likely to be audited than larger unincorporated businesses. The I.R.S. audited 2.7 percent of the tiny ventures, more than double the 1.3 percent rate for those with $25,000 to $100,000 in sales and more than the 2.4 percent rate for those with more than $100,000 in sales.
The intensified focus on low-income taxpayers resulted from pressure on the I.R.S. beginning in 1995. Newt Gingrich, who was then House speaker, and other Republican Congressional leaders were concerned about misuse of the earned income tax credit, a program that allows the working poor, especially those with children, to receive money from the government through a form of negative income tax. They proposed to sharply reduce the credit, prompting Pesident Clinton to counter with a plan to bolster audits to reduce fraud and mistakes.
The I.R.S. was also far less likely to take action last year against those who did not pay their income taxes, the new data shows. Levies on paychecks and bank accounts were down 85 percent from 1997, and liens to secure the government's interest were down 69 percent.
Seizures of property to pay back taxes, the most severe enforcement action, were down 98 percent, to 161, from about 10,000 annually the previous 10years. Auditors recommended $4.5 billion in taxes and penalties last year,down from $6.3 billion in 1993.
Much of the decline in enforcement is because of shrinking staff and complex new rules enacted by Congress in the I.R.S. Restructuring and Reform Act of 1998. Seizing property, for example, now requires a 54-step process that more than two dozen revenue officers have described as virtually impossible to navigate. And the permanent I.R.S. staff is the same size now as in 1983,even though the total number of tax returns has increased one-third and the number of complex returns by high-income individuals has grown even more.
Over all, one of every 66 corporations of all sizes was audited, a level of scrutiny significantly lower than the one in 37 audit rate for Schedule C businesses with less than $25,000 of revenue. Among the largest corporations,those with more than $250 million of assets, the audit rate was 34.5 percent last year, down from 54.6 percent in 1992.
These findings emerge from an exhaustive compilation of tax data posted on the Internet by the Syracuse researchers. The data, covering I.R.S. audit and enforcement actions through last Sept. 30, the end of the government's 1999 fiscal year, was provided to the university's Transactional Records Access Clearinghouse by the I.R.S.
Charles O. Rossotti, the commissioner of internal revenue, said that overall audit rates had fallen so far that he feared the tax system could not bear any more cuts.
Last week, asking a House subcommittee to support a 9 percent increase in his budget, Mr. Rossotti said that "we're really risking the entire tax system" by continually slashing audit rates.
In an interview, Mr. Rossotti said: "I am not going to make the claim that at this moment the system is threatened because I don't know at what point it is threatened by falling audit rates. But longer term, the threat is there if a belief begins to develop that, 'Hey, I am paying my taxes but the guy next door or the business across the street isn't, and the I.R.S. is not, or cannot, do anything about it.' "
Mr. Rossotti said that the only reason audit rates for the working poor had risen while the rates for wealthier taxpayers had declined was a mandate from the White House and Congress for close monitoring of the earned income credit.
A couple without children can use the credit to eliminate all income taxliability and get back an additional $347, but working parents can wipe out their liability and collect as much as $3,816 more. Representative Bill Archer,the Texas Republican who heads the House Ways and Means Committee, said studies had showed that more than 20 percent of returns claiming the credit had misused it.
But John Karl Scholz, a University of Wisconsin economist who studied the earned income credit in 1995 when he was a Clinton administration tax policy adviser, said that while fraud was a continuing problem, most of the issues uncovered in audits tended to be disputes over who got the credit when a couple with children had separated or divorced. Mr. Rossotti expressed the same view in an interview.
Mr. Scholz said that "if you consider the amount of noncompliant behavior uncovered for a dollar of I.R.S. enforcement resources, the amount of attention being given to the earned income tax credit is much too large."
Excluding the special audits of the working poor and people who fail to file a tax return, the audit rate for those making less than $25,000 would fall from 1 in 74 to 1 in 300. For businesses with sales under $25,000 the rate would fall from one in 37 to one in 588.
Mr. Rossotti said, "The issue of whether you are more likely to be audited if you make less than $25,000 than if you make more than $100,000 is totally distorted by the presence of the special earned income tax credit program."
Senator William V. Roth Jr., the Delaware Republican who heads the Senate Finance Committee and who was the sponsor of the restructuring act, said he was troubled by many of these statistics. "The tax laws must be applied with integrity and fairness," he said, adding that "I am hopeful that with adequate resources, better training, the leadership of Charles Rossotti and continuous modernization of the agency, the I.R.S. will improve."
Mr. Archer, the chief tax writer in the House, said he had found no flaw in the I.R.S. audit priorities and criticized the Syracuse researchers.
"This entire analysis is terribly misleading and cynical," Mr. Archer said in a statement. "Ninety-seven percent of all the so-called 'audits' relating to the earned income credit are correspondence audits, which is a relatively simple matter of mailing additional information to the I.R.S."
For now, though, the I.R.S. is scrutinizing the earned income credit with such wariness that it is sometimes denying the credit to people who are legitimately owed it on nothing more than suspicion, according to several low-income taxpayer clinics run by law schools.
Last week in Los Angeles, a tax court judge heard the case of Maritza Reyes, a cleaning woman who earns $7,000 annually and who was denied the credit after an audit.
She had separated from her husband, and each had taken one child and applied for the earned income tax credit. The I.R.S. rejected Ms. Reyes's application for the credit but at trial produced no evidence to support its position that she and her husband had not actually separated. A ruling is expected in three months.
Professor Frank J. Doti of the Chapman University law school in Orange, Calif., who with Pallavi Shah, a law student, represented Ms. Reyes, said that many low-income people audited by the I.R.S. wound up in worse straits than did Ms. Reyes. Few have access to low-income legal clinics, and they give up when they are wrongly denied the credit. Then their cases are added to the statistics on fraud and errors on which Congress relied when it financed the increased audits of the working poor.
"There is fraud in this program, but the handling of this case is bizarre," Professor Doti said, adding that the situation was not unique. The I.R.S. declined to comment on the Reyes case, as it does in all cases involving individual taxpayers. "Our policy is to rely on the facts and the law in determining eligibility," said Frank Keith, an I.R.S. spokesman.
To identify tax cheating by high-income taxpayers and those with complex returns the I.R.S. relies on face-to-face audits by highly trained revenue agents. The number of such audits fell sharply again last year.
In 1981, the earliest year for which data is available, one in 63 tax returns received such an audit, but last year only one in 322 did, down from one in 217 in 1998.
This decline partly reflects the shrunken I.R.S. audit staff, which has less time to examine complex returns.
But a second factor, perhaps more significant, is also at work, said David B. Burnham, who with Professor Susan Long is co-director of the Syracuse research organization. The I.R.S. relies on data at least 11 years old to select returns for audits based on patterns of income and deductions, Mr.Burnham said.
Changes in the economy, including companies offering more stock options in lieu of cash compensation and taxpayers deducting interest on home equity loans and margin debt, make much of this data useless for identifying cheating.
But since 1995 Congress has barred new studies to measure taxpayer compliance by the I.R.S.
The Syracuse University Web site, trac.syr.edu also lists data from the Administrative Office of the United States Courts that shows that criminal prosecutions for tax crimes, long a minor government activity, continue to decline. Last year, there were just 722 such prosecutions in the nation, half the figure in 1981, when there were one-third fewer taxpayers.
The I.R.S. data also showed an area of increased activity that may foreshadow the government's ability to collect taxes in the years ahead. The number of cases opened on people who have stopped filing returns has soared. In 1997, the I.R.S. initiated 822,898 taxpayer delinquency investigations. Last year, that number nearly doubled to 1.58 million.