by Denise M. Topolnicki Reporter associate: Isaac Rosen

Money’s third annual test for professional tax preparers has established an ominous trend: every year the contestants have done worse. When we asked 50 tax pros to complete a 1040 return for a hypothetical family back in 1988, we weren’t surprised that no two preparers computed the same tax due — and their bottom lines varied from $7,202 to $11,881. We figured that the pros would improve as they became more familiar with the outrageously complicated Tax Reform Act of 1986.

We were wrong. Another group of 50 volunteers fared even worse on last year’s test. Their calculations of taxes due ranged from $12,539 to $35,813 — and only 10 of 50 preparers submitted returns free of significant errors.

This year’s results were even more depressing: our test stumped all but two of the 50 tax pros who were game enough to grapple with it. (A separate Money survey found that while Internal Revenue Service employees are also prone to proffer bad advice, their performance has improved appreciably. See page 97.)

For the third year in a row, no two preparers came up with the same tax due for our hypothetical family of four with annual income of $132,000. Answers ranged from $9,806 to $21,216; the average was $1 3,915. The certified public accountants who designed the test for Money gave the correct tax as $12,038, which means that our unfortunate family could have paid nearly twice as much as necessary.

Also, as in the past, there was no logical connection between fees and performance. The pros charged $271 to $4,000 and spent from 4% hours to more than 50 completing the 23-form return. The average fee was $1,012; the average number of hours spent was 15, for an hourly fee of about $67. Of the two preparers who fared best, one charged about $275 above average, the other around $260 below. Two who strayed the farthest charged $22 and $492 below the norm.

More bad news: Over the three years Money has given the tests, all of similar complexity, the average preparation fee has risen by 30% from $779 in ’88 to $865 in ’89 to $1,012 this year.

Many of the participants erred on routine problems, including computation of self-employment tax and the child-care expense credit.

Why, then, such a disappointing — and deteriorating — showing from a cross-section of pros who had the self-confidence to accept our challenge? “I had hoped that more than two people would get everything right,” said C.P.A. Steven W. Caldwell, a partner with Blum Shapiro in Farmington, Conn., who designed the test with a colleague, tax manager Edmond L. DiClemente. “Granted, few clients have as many tax problems as our hypothetical family did, but the problems weren’t unusual. A professional would see them at least once or twice in a tax season.” Caldwell felt that so many did poorly because taxes are just one part of their practice and tax law has become too complex to keep up with on a part-time basis. But he also found that “some of the mistakes that people made were just plain careless.”

Money selected Caldwell and DiClemente to draft this year’s exam and evaluate the results because their firm made no mistakes and exhibited a fine eye for detail on last year’s test.

Caldwell’s $12,038 wasn’t the only possible correct tax due. One of the two pros who turned in an A-plus 1040, C.P.A. Hugh K. Campbell of Cincinnati, calculated a tax due of $11,968 because he computed depreciation on our family’s rental properties differently from Caldwell. Campbell, who spent 25 hours on the project but charges a flat annual fee for tax advice and preparation, would have billed the family $750.

A-PLUS PERFORMER HUGH K. CAMPBELL

The other crack preparer, C.P.A. Richard H. Wetter of Semanchin Wetter & Schieb in Amherst, N.Y., figured a tax of $ 11,861 because he also calculated depreciation differently and offered alternative advice to our hypothetical taxpayer regarding an excess contribution to a rollover IRA. Most of the preparers who solved that problem correctly called for withdrawing the entire excess contribution. But Wetter moved $2,000 of it to a nondeductible IRA, thereby reducing the penalty for early withdrawals. His reasoning: Money had told the preparers the client wanted to save as much retirement money as possible. Wetter would have charged our family $ 1,285 for 16 hours of work.

The two extremes in taxes due were calculated by C.P.A. Beth B. Mayer of Greensboro, N.C., who underestimated the tax due by about $2,200: and C.P.A. Raymond E. Agins of Trenton, N.J., who would have made the family send the IRS an extra $9,200. Says Mayer: “When you use software, sometimes you just go on automatic and forget what you should know.” Agins, while disagreeing with some of Money’s instructions, admitted that some of his assumptions may have been shaky.

HIGH-TAX PREPARER RAYMOND E. AGINS

Of the errant majority, most were quick to admit their mistakes and eager to learn from them. More than one said that the exam was good practice for the tax season. A few participants, however, argued that their returns differed from Caldwell’s chiefly because they made assumptions about details in our 4 1/2-page problem that seemed unclear to them. Our response: they hadn’t heeded our instructions, which stressed that participants should call us with any additional questions. In fact, we answered dozens of queries with Caldwell’s assistance. One of the keys to Campbell’s success was his willingness to follow this guideline — he called four times to clarify 44 tax points.

LOW-TAX PREPARER BETH B. MAYER

Some pros felt that they would have avoided at least the mindless errors had the software they normally use for completing returns been available when weconducted our test in December and January. But no one claimed that software would have guaranteed a perfect return. Says C.P.A. Jonathan Grobani, who worked on the test with Robert M. Olshan of Washington. D.C. (together they came in about 55,500 too high and charged 54,000, the highest fee): “Software would not have solved about half of the test problems because a human being has to analyze a lot of issues before entering anything into the program.”

This year’s participants included 35 certified public accountants from local firms, eight independent enrolled agents who are allowed to practice before the IRS, two non-C.P.A.s and five tax preparers from chains, including H&R Block, Jackson Hewitt, Triple Check Income Tax Service, General Business Services and Tax Man. (Seven C.P.A.s and one enrolled agent agreed to take part but changed their minds after receiving the test.)

None of the Big Six accounting firms was willing to try. Most said they were too busy to fill out a hypothetical return; a spokesman for Arthur Andersen was more candid, concluding that his firm had a lot to lose by performing poorly and little to gain by doing well. Indeed, contestants from national firms in both earlier tests were generally beaten for accuracy and affordability by the competition.

Money’s hypothetical family consisted of John Smith. 40, his wife Jane, 38, their daughter Jill, 4, and son Jack, 3. Their problems were the kind we felt Money readers might well share — and learn by. The family moved from Phoenix to Los Angeles last year because John, a salesman, changed jobs. Jane closed her money-losing retail toy business and became a marketing consultant. The Smiths earned a total of $ 132.000 in salaries and self-employment income. They also owned a rental cottage in South Carolina, shares in two publicly traded limited partnerships and a real estate general partnership, plus mutual funds and stocks.

Here are the specific situations that flummoxed the pros, listed in order of the greatest effect on the Smiths’ tax bill:

Mutual fund redemption.

Between 1985 and 1989, the Smiths purchased 3.500 shares in a mutual fund at a total cost of $151,500. Last year they instructed the fund’s custodian to redeem 1.000 shares, but they didn’t specify which ones. They pocketed $42.000 from the redemption. The preparers were free to use one of three methods to calculate the cost of the shares for the purpose of computing a capital gain or loss: double-category averaging, single-category averaging, or first in, first out.

Twenty-one pros saved the Smiths a bundle with double-category averaging. Under this method, shares held long term (more than a year) are separated from those held short term. The cost of each share is determined by dividing the total cost of the shares in a category by the number of shares in that category. If you don’t specify which shares should be sold first, as the Smiths didn’t, the long-term shares are considered to have been sold first. Thus Caldwell computed the cost of the Smiths’ shares at $43,800. Since they realized only $42,000 on the shares’ redemption, they could claim a $ 1.800 capital loss.

The preparers who used single-category averaging came up with a smaller loss; those employing the first-in. first-out method figured that the Smiths’ shares cost less than $42,000, resulting in a gain. In all, answers ranged from a $1,280 loss to a $12,000 gain. (One C.P.A. declared the sale a wash.) The difference in tax due on a $12,000 gain vs. a $1,800 loss is a painful $3,864.

Worthless stock.

Jane met all of the requirements under Section 1244 of the tax code to claim an $8,000 ordinary loss on worthless stock that she held in her ill-fated toy company. Under Section 1244, Jane was eligible to take an ordinary loss of up to $100,000 ($50,000 if she filed separately). Had her losses topped those limits, she could have deducted the excess as capital losses.

Twenty-three preparers ignored the provisions of Section 1244 and claimed a capital loss instead of an ordinary loss. (Two participants took an ordinary loss but calculated it incorrectly.) The preparers who took a capital loss instead of an ordinary loss forced the Smiths to pay an additional $2,044 in taxes. Reason: only $3,000 of capital losses are deductible dollar for dollar from ordinary income each year. The ordinary loss would have been completely deductible for 1989. Leftover capital losses can be carried forward and used to offset capital gains or income in future years.

Retirement savings.

We told the pros that the Smiths had no employer-sponsored qualified retirement plans last year but were eager to sock away as much money as possible in tax-deferred retirement accounts. Our instructions also made it clear that the Smiths didn’t meet with their tax preparer until 1990. which meant that Jane missed the Dec. 31 deadline for setting up a Keogh plan, where she could have slashed up to 20% of her self-employment income or $30,000, whichever is less. Yet 19 practitioners funded a Keogh for Jane anyway. They should instead have recommended a simplified employee pension plan (SEP) for Jane; SEPs needn’t be established or funded until the tax-filing deadline, and you can put away 13.04% of your self-employment income or $30,000, whichever is less. Caldwell figured that Jane was eligible to put $7,621 in a SEP.

Worse yet, six participants failed to recommend a SEP for Jane and instead advised the Smiths to put $2.000 each into tax-deductible IRAs. One C.P.A. who gave this advice later opined that he wasn’t wrong, just less aggressive than Caldwell. The IRS wouldn’t consider his answer wrong. But he caused the Smiths to contribute and deduct only $4,000. vs. the $7,621 they could have put into a SEP, so we venture to guess that a taxpayer would call it a blunder.

Temporary rental of principal residence.

The Smiths had difficulty selling their house in Phoenix, so they rented it out at a loss after moving to Los Angeles last March. They finally sold the property last December. The IRS and the tax court both have held that renting your principal residence for a few months while it’s for sale doesn’t make it a rental property. Sodeductions for rental expenses are limited to the amount of rental income; you cannot claim a loss.

Unfortunately for six of our contestants, the Smiths live in California, which along with Arizona is under the jurisdiction of the Ninth Circuit Court of Appeals. In 1985, that court allowed a rental loss deduction in a case similar to that of the Smiths. Since that ruling governs taxpayers in Arizona and California. Caldwell felt justified in claiming a $5,440 passive rental loss. Most of the preparers did the same, but the six played it more conservatively and didn’t take the loss, which added $1,523 to the Smiths’ tax bill. We didn’t penalize the less aggressive preparers because of the regional nature of the issue, but the Smiths had a good case for claiming the loss.

Moving expenses.

Five practitioners didn’t treat selling expenses — commissions and legal fees, for instance — on the Phoenix house as deductible moving expenses. Five other preparers didn’t take all of the allowable deductions. Those who mistakenly claimed no write-offs hiked the Smiths’ tax bill by $736.

Inherited lump sum.

As the sole beneficiary of her uncle’s employer-sponsored pension. Jane received a lump-sum distribution of 100 shares of company stock from a qualified pension plan after her uncle’s death. She sold the securities for $12,500 last December. Fifteen pros correctly computed a lax of $170 on Jane’s windfall using favorable 10-year averaging; six others figured slightly lower taxes by rounding numbers. Jane was eligible to use the 10-year method because her uncle turned 50 before Jan. 1, 1986.
Everyone else erred because they miscalculated the taxable portion of the lump sum, used less favorable five-year averaging to figure the tax or forgot to exclude a $5,000 death benefit from the taxable amount. Jane had a right to the full exclusion because she was the sole beneficiary of her uncle’s pension.

Excess IRA contribution.

John received a $20,000 lump-sum distribution from his former employer’s profit-sharing plan and rolled over the entire amount into an IRA. Problem was, $6,000 of that amount represented John’s after-tax contributions, which he wasn’t permitted to roll over. Indeed, the IRS would consider the $6,000 an excess contribution and slap a 6% excise tax on it.

John could have avoided the penalty by withdrawing $6,000 from his IRA before filing his return. The portion of the account’s earnings attributable to his excess contribution ($750) should have been reported on his return and, though it seems unfair, subjected to the 10% tax penalty for early withdrawals from retirement plans. Unfortunately, just seven preparers did that. (Two others who were also correct told John to keep $2,000 of his excess contribution in a non-deductible IRA and withdraw only $4,000.) The rest either ignored John’s excess contribution, thereby exposing the Smiths to a possible IRS audit and a 6% excise tax, or increased the couple’s tax bill by leaving John’s excess contribution in his IRA and applying the 6% tax. Oddly, two preparers who completed otherwise sterling returns, C.P.A. Donald J. Hougardy of Grangeville, Idaho and Donald G. Straw of the General Business Services chain’s office in Littleton, N.H., didn’t apply the 10% early-withdrawal penalty.

Kiddie tax.

The Smiths’ four-year-old daughter Jill earned $1,200 in income from a custodial account in her name. Her parents could have filed a separate return for her or reported her income on their 1040, using the new Form 8814. Though we instructed the preparers to complete all necessary tax forms for the family, 13 neglected to report Jill’s income anywhere. C.P.A. Nicholas Mazzone of Cerio Mazzone Kaminsky & Co. in Cleveland argued that our instructions were fuzzy, but he managed to retain his sense of humor. Said he: “I assumed that the kid had her own accountant.”

Mutual fund dividends.

In January, the Smiths received $200 in dividends that had been declared by a mutual fund during the fourth quarter of 1989. Since the date of declaration is the one that counts as long as the check shows up before Feb. I, the pros should have reported the income. Alas, 21 preparers failed to do so. One C.P.A. jokingly insisted that he always ignores the tax law that applies in this situation.

Self-employment tax.

It’s hard to believe, but four preparers missed this one. Two got the wrong answer by calculating Jane’s business profits incorrectly. The other entrants goofed because they couldn’t get their hands on the 1989 tax forms before our deadline. We permitted them to use 1988 forms but told them to make any necessary changes. They forgot to note that the largest amount of combined wages and self-employment earnings subject to Social Security tax for 1989 is $48,030, up from $45,000 in 1988.

Child-core expenses.

The form used to report such costs states clearly that you cannot claim expenses of more than $4,800 for the care of two or more dependents. Nevertheless, one C.P.A. listed the full $8,000 that the Smiths spent on child care. He calculated a credit of $1,600; the right answer was $960.

What lessons can the average taxpayer draw from the pros’ bad showing? Should you consider giving up on the entire profession and doing your own return? Before you take such a plunge, ponder your answer to this daunting question: if our insanely complicated tax code can trip up even highly trained C.P.A.s. what hope would a mere layman have unless his return was relatively simple or he was an avid amateur accountant willing to put in long hours of study?

Money’s test results should convince you that at least two steps are in order.

First, become generally informed so that you can bring at least some independent intelligence to your meeting with your preparer. Start by investing in a solid lax guide and reading the sections that apply to you. A fine choice: J.K. Lesser’s Your 1990 Income Tax (Simon & Schuster. SI 1.95); other picks appear on page 17 of the February issue of Money.

Second, keep asking your preparer questions about any items on your return that you don’t understand. The more questions you ask. the more you’ll force your pro to think again about the issues — and maybe even spot an error that might otherwise bring down the wrath of the revenuers.

Your Tax Dollars

Surprise! The IRS gets more helpful

And now for some relatively comforting news. Our third annual survey of the folks at the IRS who answer your queries by phone showed they did passably well — a big improvement over past years. In Money’s 1988 survey, these employees, officially known as taxpayer service representatives, dispensed erroneous advice nearly half the time. Last year, after the agency reformed its training program, the figure dropped slightly to 41%. This year, Uncle Sam’s helpers answered only 28% of our questions incorrectly.

What gives? IRS spokesman Henry Holmes credits the training program. The 11-week regimen now puts more emphasis on problem areas. But he modestly adds: “There is no doubt that our accuracy rate needs improvement. We’re trying to find ways to ensure better-quality service.” Efforts include keeping more seasonal temps full-time and getting more temps to return annually. All service reps also now have what the IRS calls “probe and response” booklets guiding them through follow-up questions on thorny issues.

As we did in 1988 and 1989, this past January we placed calls to 100 IRS service reps across the country. Each respondent was asked one of 10 questions, and each question was posed to 10 different reps. The questions, typical of what reps are likely to encounter, were prepared with the help of C.P.A. Martin Nergaard of  Hansen Jorgenson & Co. in Minneapolis.

We discovered that most service reps aced the questions if they could look up the answers in a 1040 instruction booklet or in a basic IRS publication. But they tended to give misleading advice if they had to dig. For example, nine of 10 gave incomplete replies when asked whether we could deduct a deceased person’s medical expenses on his or her final income tax return. All of the reps correctly said yes, but only one told us that we had to include a statement waiving the right to deduct the expenses on the deceased’s estate tax return.

Service reps were generally genial, with a few ugly exceptions. When we asked a Boston rep to send us written confirmation of his advice, he gruffly replied: “You have to call Washington for that. If you don’t think the answer is good enough, you can come into the IRS in Boston and discuss it with someone on the first floor.” He was wrong as well as rude: he should have told us that we could get a written reply within 10 days if we sent our question in writing to a local IRS office.

—D.M.T. with Andrea Coccia and Deborah Lohse