Your client is a mid-sized taxpayer who has received an IRS audit letter for a recent tax year. The IRS agent schedules a visit during which he advises you of your rights as a taxpayer and explains IRS Publication 1, which describes the process for examination, appeal, collection, and refunds. The agent explains that he will personally review your books and records to fulfill mandatory compliance checks and will try to get an IRS engineer to address some of the more technical issues, such as the R&D Tax Credit. He will also need to review the personal tax returns of all corporate officers.

Naturally, you may be concerned and have questions such as:

• Have you reserved enough to cover potential adjustments and penalties?
• Will audit issues discovered this cycle cause the IRS to open up previous years still under statute?
• Was your R&D Tax Credit significant enough for the IRS to assign an engineer to deal with the technical merits of the claim?
• Is the assignment of an IRS engineer a sign that the R&D tax credit is considered significantly risky?
• Should you consider hiring a freelance R&D engineer who understands what the IRS engineer is looking for?
• How does an IRS engineer read and evaluate an R&D study?
• Do you need someone with experience steering audits to a reasonable settlement rather than a trip to appeals or tax court?
• Who is this so-called “IRS engineer”?

IRS Engineers

IRS engineers are generally STEM-trained professionals with engineering degrees, many of whom hold a Master’s degree in finance or accounting. They are subject matter experts on IRC sections 41 and 174. They determine whether the work performed meets the 4-part test and whether business component expenses have been properly assigned. The engineer prepares the final 886a part of the NOPA, which serves as an official record of the audit. Engineers prefer detailed facts about the projects and the qualified activities. They prefer not to spend their time forcing “a round peg into a square hole” and would rather easily fit taxpayer-supplied narratives into IRC requirements.

Typical R&D Tax Credit Arguments

What types of activities are qualified as R&D activities? A better way to approach this problem is to identify what activities are specifically prohibited. One area of ongoing conflict is “routine engineering.” This is due to differences in terminology and perspective between engineers and the accountants who penned the tax rules. Engineers rarely call what they do “routine.” In general, engineers consider each activity they perform specific to the situation at hand.

Another type of activity often misunderstood by non-engineers is debugging. Debugging a manufacturing assembly line is different from debugging software code. When developing software, it is common to run a debugger, a separate application, and run small tests or trials while writing the code. This is different from debugging an assembly line. In software, debugging is a process of experimentation. In the second case, debugging is to fix an issue in an already completed business component. There are rules against reworking of existing processes, which are likely historical.

Adaptation is taking an existing business component and modifying it for use by another customer. This type of activity needs to be described very carefully to avoid running afoul of regs that specifically forbid changing an existing business component for another customer’s use.

Documentation issues are also quite common. Many IRS engineers will demand contemporaneous time tracking data. However, rarely do IRS engineers have the time to examine this data closely. The best bet is to have the data and also provide a summary of the information it contains. In the Kyocera vs. Commissioner case, the timing of the documentation collection was considered in disallowing the large R&D credit. Collecting narratives later, after R&D has been completed, is risky.

All expenses must be tied back to specific business components. Whether they are wages, supplies, or contract research, each must be associated with a specific business component. Testing is another area of contention. There are at least 24 different types of testing. Some qualify, but most don’t and are considered a quality control activity. Care needs to be taken in this area. An experienced R&D engineer will zoom in on these potential issues.

Contemporaneous documentation is created concurrently with the performance of research and is required under IRC section 41. But by the time you have received an audit letter, the time to prepare proper documentation has passed. We suggest that you organize your documentation by business component. Your engineer will want to trace at least a few documents back to the Form 6765. High-level summaries are useful as an introduction, but not as the majority of your substantiation. Time records are useful as backup, and it is likely your engineer will want to trace a few of those as a spot check. Reconstructions of activities and verbal interviews are less useful than actual written documentation, such as emails and design reports. General statements such as “all engineering hours qualify” will raise your engineers' hackles and should be avoided, as should boilerplate text.

Understanding the IRS Engineers' Constraints

IRS engineers must be able to justify any concessions made during the audit to technical reviewers and defend any compromises. This means they will need to dig into the details to prove they fully reviewed the tax credit claim. Often, they suggest sampling business components or tax years and then extrapolating the results to make the data easier to digest. This type of sampling is called judgment sampling, as opposed to statistical sampling. Rev. Proc. 2011-42 is the primary IRS guidance detailing the use and evaluation of statistical samples.

Be prepared to concede business components with weak facts in exchange for substantiation for projects that are more clearly qualified. The audit is often a give-and-take based on technical issues. Overall, your IRS engineer’s personality is often the most influential factor in your audit's results. This is because the law governing the R&D tax credit, IRC 41, is complex, and multiple practitioners often disagree on its implications.

Three things that CPAs should consider once receiving an audit letter:

• Organize R&D files by business component
• Pre-identify weak projects to concede if needed
• Consider using an experienced R&D engineer as part of the defense team

Addressing these items early can significantly improve outcomes in R&D exams with IRS engineers.