THE BUSH ADMINISTRATION HAS COME UP WITH STILL another outrageous tax-giveaway scheme, this time not by legislation but by administrative fiat. In mid-December, the administration announced that it would soon send out billions of dollars in tax refunds to companies that have flouted the tax laws--and to the major accounting firms that helped them do so.
At issue is something called the "research and experimentation" tax credit, which dates back to Ronald Reagan's 1981 tax-cut law. According to the congressional reports on the original legislation, the free market doesn't do an adequate job of encouraging scientific advances. Not only is research inherently risky, it was argued, but businesses that engage in successful research often find that the fruits of their efforts generate public benefits well beyond the profits accruing to the companies. Consequently, "many businesses have been reluctant to allocate scarce investment funds for uncertain [or inadequate] rewards."
Whatever you may think of this argument--I find it pretty dubious--solving this "market failure" problem was clearly what Congress had in mind in 1981. But once the tax break was adopted, companies and their tax advisers quickly set out to pervert its purpose. What, after all, is "research"? Soon, horror stories emerged about tax credits being successfully claimed for such scientific breakthroughs as McNuggets, Gillette's Lemon-Lime shaving cream, and new fashions in clothing. More generically, as one wag put it, "if you send the janitor down to fix the boiler and he succeeds, it's repairs; if he fails, it's R&D."
In the 1986 Tax Reform Act, Congress tried to curb abuses of the credit by defining "research and experimentation" more narrowly. To address the McResearch issue, tax breaks were specifically denied for work on "style, taste, cosmetic, or seasonal design factors." More important, the 1986 law tightened the general definition of "qualified research."
Besides satisfying the vague 1981 rule (which refers to an-other tax-code section that allows a write-off for research expenses), the 1986 law required tax-credit-eligible research to meet two additional conditions. Since Congress wanted to subsidize only "risky" research that entails some chance of failure, there must be "a process of experimentation." And since the key goal is to promote science that generates public as well as private benefits, research must be "undertaken for the purpose of discovering information ... which is technological in nature. …