IRS releases R&D credit audit plans.

In November 1995, the IRS Research Credit Issue Specialist released drafts of audit plans to be used in examination of research credit claims. There are two plans: one for general research credit claims, and one for research credits based on internal-use software development. The plans are not part of the IRS Manual; therefore, while they are recommended, the plans are not mandatory and an agent is not required to use them.

The general research credit audit plan recommends that agents take a three-step approach. First, the agent is to request all available information (including all workpapers) that the taxpayer used to determine the credit. Second, the agent is to review, in detail, all expenses that the taxpayer has used in its credit calculation to determine if each expense is a qualifying expense. Third, the agent is to review the projects or activities to determine if each project is a research activity. The unusual part of the recommended approach is that the Service is examining expenses claimed prior to determining whether the project qualifies as a research project. It seems that this approach means that the IRS will determine whether a project is a research activity in part on the basis of the quality of the taxpayer's bookkeeping. For example, if a project has a large amount of expenses charged for nonresearch activities (such as training or administrative costs), the Service might then conclude that the whole project was not research. Conversely, if the vast majority of the expenses charged to a project code look like research expenses (e.g., engineering costs), the IRS might accept the project as a research project.

The audit plan for internal-use software development credits follows the same general approach. The tone of this audit plan, however, is much more aggressive and instructs agents to "challenge the software development activity" and make the taxpayer prove why each software development activity meets the definition of research. In this regard, the plan states that a footnote in the Conference Committee Report, which the Service reads as requiring that software development expand or refine the existing principles of computer science, is a key point. Clearly, this audit plan envisions that the taxpayer has an extremely high burden of proof to meet in order to claim credits for software development. It should be noted that this audit plan's approach arguably contradicts the generally favorable and much more inclusive language in the body of the Conference Committee Report; see H. Rep. No. 99-841, pages 71-76.

While the use of these audit plans is not mandatory, the plans themselves do signal the approach that the IRS is likely to take in auditing claims for research credits. Clearly, it is crucial that taxpayers spend time and effort preparing clear documentation of the expenses qualifying for the research credit. The audit plans stress that contemporaneous documentation of expenses is important. Careful documentation of both expenses and projects will greatly increase the probability of having credit claims sustained if audited.