Sun, Jul 05, 2009

Washington

High-income tax cheats scrutinized

Bloomberg News
Tucson, Arizona | Published: 10.24.2008
People with incomes above $500,000 a year are more likely to cheat on their taxes than those with lower incomes, according to a study of unpublished Internal Revenue Service data.
The study, by University of Michigan business professor Joel Slemrod and IRS economist Andrew Johns, concluded that people who earned between $500,000 and $1 million hid 21 percent of their money from the tax agency, while people with incomes under $100,000 misreported 8 percent or less. Those who earned more than $2 million hid about 11 percent of their income.
The study analyzes tax-cheating patterns by "true income," a method that adds income known to have been hidden from the IRS to what was actually reported.
"We find that, when taxpayers are arrayed by their 'true' income, defined as reported income adjusted for the underreporting estimated by the IRS tax-gap methodology, the ratio of aggregate misreported income to true income generally increases with income," Slemrod and Johns wrote. The paper was first published Sept. 12 and first reported Tuesday by Forbes Magazine.
The study, based on 2001 data, has implications for the way the IRS tackles the so-called Tax Gap, or the estimated $345 billion in taxes that go unpaid annually. Some $290 billion of that is never collected, the IRS says.
Noncompliance rates
Earlier IRS studies found that noncompliance rates were low among wage-earners whose employers withhold taxes from paychecks and higher among people who receive capital gains, partnership and business income that doesn't have to be reported to the IRS in as much detail as salaries.
Earlier IRS studies found as much as 57 percent of small- business income, or about $68 billion, goes unreported annually, while 1 percent of salaries go unreported.
Slemrod and Johns said richer Americans may not fear getting caught cheating or the consequences if they are.
"Lower noncompliance would result from a higher perceived probability of detection, a higher perceived effect of the level of noncompliance on the perceived probability of detection, and the accompanying penalty on detected evasion," the authors said.
Slemrod said it isn't clear why evidence of cheating appears to peak in the $500,000-$1 million range. One possibility, he said, is that the IRS's statistical methodology doesn't accurately detect more sophisticated tax-avoidance activity by multi-millionaires.