Lower income people cheating on cap gains?

[Demosthenes]

Surprising Finding in Capital Gains Study
By DAVID CAY JOHNSTON
Published: May 5, 2007

Capital gains are the most highly concentrated form of income, with just 10,000 or so taxpayers earning a third of all profits from selling investments and only 3 percent of gains going to people who make less than about $50,000.

So could it be that most of the tax cheating on capital gains income, as measured in dollars, is by people at the lower end of the income ladder?

That was the observation made Thursday in a letter by Thomas A. Barthold, acting chief of staff for the Congressional Joint Committee on Taxation. The letter was addressed to Representative Rahm Emmanuel, a Chicago Democrat who wants to crack down on tax cheating, especially by investors. Mr. Emmanuel said he did not believe the conclusion.

Mr. Barthold wrote in the letter, “The majority of the dollar amounts of underreported capital gains income from securities transactions” were among people in the 10 percent and 15 percent income tax brackets.

That would mean people with taxable incomes of no more than $30,650, which in most cases means people with total incomes, before deductions, of less than $50,000.

Taxpayers with incomes of $50,000 or less earned about $15 billion of the total of $471 billion in long-term capital gains reported in 2004, Internal Revenue Service data shows.

Mr. Barthold’s letter cited an Internal Revenue Service study of tax cheating based on data from 2001. The detail that a majority of the cheating is by low-income people had not been previously known.

The committee said that its own internal computer model and other data supported the statements in the letter attributing most of the cheating to lower-income Americans.

Mr. Emmanuel made available a copy of the letter and other documents yesterday and vowed to press an inquiry into which investors are cheating and what can be done to tax investment income as accurately as wages.

Jay A. Soled, a professor of business at Rutgers University who has made several studies of tax cheating, said he doubted it was true that most of the cheating was by lower-income people. “It seems mathematically impossible.”

“So the culprit in capital gains cheating is Joe Sixpack?” Professor Soled asked.

He said the finding was at odds with his own studies and a 1979 I.R.S. study that he considers much more thorough than the much smaller study cited by Mr. Barthold.

He called for full disclosure of the details of the I.R.S. study and any other documentation supporting what Mr. Barthold wrote.

The I.R.S. has told Congress that it taxes 99 percent of wage income, but only about 70 percent of nonwage income. The disparity is largely because of different reporting rules. Employers tell the I.R.S. how much workers make, but Congress trusts business owners and investors to report their income with little or no verification.

In the case of securities, the only reporting is of the proceeds from the sale of stocks or bonds. The price paid to acquire the securities is not reported and thus taxpayers can cheat by inflating the price paid.

Professor Soled said there is significant evidence that many taxpayers make up the purchase price they report.

Mr. Emmanuel said he was disturbed that the Treasury Department estimated that requiring brokerages to report the price paid for securities would reduce $12 billion of tax cheating by only about $300 million annually in the early years.

He said that he spoke by telephone with Henry M. Paulson Jr., the Treasury secretary, about how this allowed too much cheating. Both men are former investment bankers, Mr. Paulson most recently as chairman of Goldman Sachs.

“You would fire any first-year associate at Goldman who left $11.7 billion on the table,” Mr. Emmanuel said he told Mr. Paulson.

The Treasury Department declined to comment.

[Quixote]

“The majority of the dollar amounts of underreported capital gains income from securities transactions” were among people in the 10 percent and 15 percent income tax brackets.

Jay A. Soled, a professor of business at Rutgers University who has made several studies of tax cheating, said he doubted it was true that most of the cheating was by lower-income people. “It seems mathematically impossible.”

In the case of securities, the only reporting is of the proceeds from the sale of stocks or bonds. The price paid to acquire the securities is not reported and thus taxpayers can cheat by inflating the price paid.

Professor Soled said there is significant evidence that many taxpayers make up the purchase price they report.

My experience with low income taxpayers suggests that when they underreport capital gains they do so by not reporting the sales at all, not by overstating their basis. They do so due to unfamiliarity with capital gains on their part, and occasionally on the part of their return preparers, who are more often than not unenrolled preparers.

[jg]

It would not be surprising to me, at all, that lower income taxpayers omit a greater percentage of their capital gains. (It was already pointed out the raw dollar amount is not likely greater for lower income tapayers.)
Presumably the report includes capital gain distributions (CGD) from mutual funds in this computation. It would be quite easy to understand how this CGD entry on the 1099-DIV is missed or omitted in many cases. As many readers know, the CGD must be reported on Schedule D or directly on the line for capital gains in addition to the dividend amounts on Schedule B or the line for dividends. That there are multiple line entries or different forms needed from the one information report seems to be easy to miss for those not familiar with tax preparation.

I would not, however, characterize such an unintentional omission, even one that underreports income, as cheating.