Not All Tax Software Is Created EqualDean Foust
Last year, the makers of tax software programs taxed the loyalty of their customers. A spate of software bugs that created calculation errors or missed deductions, coupled with perpetually jammed phone help lines, left many users of leading programs such as TurboTax and Kiplinger TaxCut wondering if the benefits outweigh the hassles.
Should you trust your taxes to a software program? For many individuals, the answer is still yes. Despite its flaws, income-tax software remains a better alternative to using a pencil, calculator, and scratch pad-- which can add hours to your work and increase the chances of a math error. And for their part, the makers of the leading programs vow they have taken steps to minimize the risk of last year's problems recurring.
DISCLOSURE. Instead of ripping up their programs to build in more features, most of the companies are concentrating on making sure their programs work right. Intuit also says it has doubled the size of its technical-support staff, to 500, to field questions about TurboTax and MacInTax. And all the software publishers pledge they'll be more diligent in reporting bugs. In years past, they tended to wait until late in tax season to issue an all-inclusive list. But after last year's fiascoes, they have vowed to get the word out sooner: Intuit and Parsons through their Internet Web sites and others through online services, fax, and phone.
Given the relative lack of new features to compare and discuss, BUSINESS WEEK decided to subject this year's crop of tax programs to a practical test: a hypothetical tax return that was so complex it bordered on the diabolical. This return from hell was the handiwork of Alfred C. Giovetti, a Catonsville (Md.) certified public accountant and accounting instructor whose students include Internal Revenue Service agents. Giovetti created a return for a fiftysomething couple earning a combined $200,000 from a variety of sources including salaries, employee stock options, and rental property. He tested five popular programs--TurboTax, Kiplinger TaxCut, Simply Tax, Personal Tax Edge, and Perfect Tax--then stacked up their results against a $3,500 program popular among professional tax practitioners.
His finding? While all of the programs--with the notable exception of Perfect Tax--interpreted the tax code accurately, they nonetheless pose many critical questions so ambiguously that any user lacking a CPA license could answer some inaccurately--and risk incurring not just penalties but an IRS audit. "These programs lull us into a sense of false security," says Giovetti. "We assume the computer knows more than it often does."
TROUBLE SPOT. Particularly troubling was how most programs breezily ask users up front whether they think they are a candidate for the alternative minimum tax, which some people must pay if they have a large number of deductions. Answer no, and chances are the software won't determine afterwards, based on your subsequent entries, whether your initial judgment was right. When Giovetti added a generous dollop of deductions and exemptions into his formula--which would have subjected his make-believe couple to the AMT--the programs underestimated the correct tax liability by as much as $40,000. The moral: For high-income households with extensive deductions, exemptions, and other shelters, tax software is no alternative to a good accountant.
For the rest of us, the best program might be TaxCut, which allows you to complete your return using an interview format that's more intuitive and logically arranged than those of rivals. TaxCut's "audit" feature--which flags answers that could trigger a visit from the IRS--is so thorough it catches discrepancies that only a CPA might be expected to find. And in Giovetti's test, only TaxCut detected that a prior employer made a premature retirement payout without withholding the mandatory 10%--a move that, if not caught, can subject the recipient to an IRS penalty.
NET WORTH. Not that TaxCut is without flaws. Like other programs, it doesn't always automatically transfer answers to other relevant sections of the return: Users, for instance, must manually carry the gain from Form 4684--which handles casualty losses--to Schedule D. And while the program's audit feature will take you to the point in your return where a problem may lie, it then returns you to the beginning of the audit function.
Giving TaxCut a run for its money is Personal Tax Edge. Don't be fooled by its $19 price tag--Tax Edge's interview and audit features are as strong as any of its higher-priced rivals. Admittedly, Tax Edge lacks the extensive help features of TaxCut and TurboTax. But the deluxe CD-ROM version of Tax Edge, which was acquired two years ago by Intuit, also comes with access to the company's Web site--which provides extensive online help. Another bonus: Tax Edge includes a set of nifty investment calculators that can be used long after tax season and are worth the $19 price by themselves.
Tax Edge fared well in Giovetti's test: The program takes more care than others to prevent users from taking the same deduction twice by alerting you when an entry has been transferred to another form--although, irritatingly, it won't always handle the transfer automatically. Tax Edge will also appeal to last year's users of Novell's Tax Saver, which was discontinued, because both programs share the same underlying computer code.
Mildly disappointing was the performance of TurboTax, which despite being the best-selling tax program, still hasn't worked out some annoying kinks in its interface. Don't skip around various sections of the interview process, or between the interview and the underlying tax forms, because TurboTax won't always take you back to the question you answered last--forcing you to waste a lot of time relocating your place.
Giovetti also found that TurboTax isn't always good at remembering how you responded to relevant questions earlier in the Q&A, either. When you answer questions about Form 1099-R, for instance, it advises you that you'll also have to complete the related Form 5329--but doesn't provide a similar reminder later on, when you actually need it. To its credit, TurboTax reviews each form as you complete it--rather than just when you finish the whole process--giving you a chance to correct inconsistencies while the information is fresh in your mind and you have the documentation at your fingertips.
TWO DOGS. Then there's Simply Tax, which is, well, simply awful. At first, it appears that Computer Associates International finally got it right with the three-year-old program. The interview function begins impressively, with entries transferred automatically to other relevant portions of the return. But you quickly get the impression that the designers of Simply Tax ran out of time because the Q&A function abruptly becomes less intuitive and simply starts throwing forms at the user to fill in. Even then, some of the forms are displayed in a small typeface that's almost illegible on a 15-inch monitor. What's more, Simply Tax's handling of depreciation is too simplistic--not allowing you to choose the number of years or the method.
Simply Tax only looks good when compared with Perfect Tax, a $79 program available by direct mail from Dallas-based Financial Services Marketing. Perfect Tax's poor performance is baffling, given that its parent also sells more expensive versions to professional tax practitioners. But in Giovetti's hypothetical return, Perfect Tax miscalculated the child-care credit by a hefty $480 and didn't provide the proper deduction for margin-interest expense or for rental property. Against this program, a pencil, pad, and calculator almost seem preferable.