The incentive problem | Research Credit Integrity Alliance

Key points

A federal research credit is calculated from qualified research expenses, so every expense moved into or out of the claim changes the credit.1
Deciding what counts involves real judgment: whether each activity meets the statutory tests, how much of each person's time was spent on qualified work, and how to treat contractors.123
The taxpayer carries the burden of showing that the requirements are met, and must keep records that substantiate the claim.42
When an adviser is paid a share of the credit, each judgment call that adds to the claim also adds to the fee. This is an argument about incentives, not a finding about any adviser or any claim.

Why this is about incentives, not people

This article sets out a structural argument. It asks what a fee arrangement rewards, and where in the preparation of a research credit claim those rewards can make a difference. It does not claim that any adviser has inflated any claim, and it does not estimate how often that happens. Nothing here suggests that advisers who charge contingent fees do not prepare careful claims. The point is that the fee design itself pulls in one direction, and that a business should understand which direction before it signs.

How the credit is calculated

The federal credit for increasing research activities is in Section 41 of the Internal Revenue Code. Under the regular method, the credit is 20 percent of the amount by which a taxpayer's qualified research expenses for the year exceed a base amount.1 A taxpayer may instead elect the alternative simplified credit, which is 14 percent of the qualified research expenses that exceed 50 percent of the average qualified research expenses for the three preceding years.1

Qualified research expenses are mainly of two kinds. In-house research expenses are wages for qualified services, supplies used in qualified research, and certain amounts paid for the right to use computers. Contract research expenses are, in general, 65 percent of amounts paid to a person other than an employee for qualified research.1

The practical effect is simple. Once a business is above its base amount, each extra dollar of qualified research expenses adds to the credit at a fixed rate. Each dollar removed takes it away. Under a fee calculated as a percentage of the credit, the same is true of the adviser's fee.

Where judgment enters

None of the tests below is a simple yes or no. Each requires facts to be gathered and a view to be formed. That is where fee incentives can matter.

Is the activity qualified research?

Section 41 defines "qualified research" through a series of requirements, and the IRS's audit techniques guide for the credit describes them as four tests.4 In the guide's terms:

The section 174 test. The expenditures must be ones that may be treated as research or experimental expenses under section 174.4
The discovering technological information test. The research must be undertaken to discover information that is technological in nature. The regulations say research is undertaken to discover information if it is intended to eliminate uncertainty about the development or improvement of a business component, and that uncertainty exists if the information available does not establish the capability or method for developing or improving it, or its appropriate design.2
The business component test. The information must be intended to be useful in developing a new or improved business component, which the statute defines as a product, process, computer software, technique, formula or invention to be held for sale, lease or license, or used in the taxpayer's trade or business.1
The process of experimentation test. Substantially all of the activities must be elements of a process of experimentation for a qualified purpose.1

The IRS guide says the taxpayer must be able to show that an activity meets all four tests, and that the tests are applied separately to each business component.4

Several features of these rules leave room for different conclusions on the same facts:

"Substantially all". The regulations treat the process of experimentation requirement as met only if 80 percent or more of the research activities for a business component, measured on a cost or other consistently applied reasonable basis, are elements of a process of experimentation for a qualified purpose.2 Measuring that share requires choices about what counts as an activity and how to weigh it.
What a process of experimentation is. The regulations describe it as a process designed to evaluate one or more alternatives where the capability, method or appropriate design of the result is uncertain at the start, and they say it must be an evaluative process.2 They also say that uncertainty about a business component's design "does not establish that all activities undertaken" to achieve it are a process of experimentation.2
The shrinking-back rule. If the requirements are not met for a whole product or process, they are applied to the most significant subset of its elements, and so on down until a subset qualifies or the most basic element fails.2 Where to draw the line is a judgment.
Exclusions. The statute excludes, among other things, research after commercial production begins, adapting an existing business component to a particular customer's need, duplicating an existing component, routine data collection, routine quality control testing, and research related to style, taste, cosmetic or seasonal design factors.1 Each exclusion turns on facts that can be described in more than one way.

Whose wages count, and how much?

Wages count only for "qualified services": engaging in qualified research, or the direct supervision or direct support of it.1

The regulations draw these lines closely. Direct supervision means first-line management of qualified research, and does not include a higher-level manager to whom first-line managers report.3 Direct support includes, for example, a clerk compiling research data, but does not include general administrative services such as payroll, accounting for research expenses, or general cleaning.3

Where an employee does both qualified and other work, only the wages allocated to qualified services count. Unless another method can be shown to be more appropriate, the allocation is by the share of the employee's time actually spent on qualified services.3 If at least 80 percent of an employee's wages are allocated to qualified services, all of that employee's wages for the year are treated as qualified.3 Where time records are incomplete, an allocation may rest on interviews and estimates. Each estimate is a judgment call, and the 80 percent threshold means a small change in an estimate can move a large amount of wages.

Contractors and supplies

Payments to contractors count at 65 percent in most cases, and only for qualified research.1 Supplies count only if they are tangible property used in the conduct of qualified research, other than land, improvements to land and depreciable property.1 Both require a view on what the money was actually spent on.

How a percentage fee interacts with these calls

The argument here is general economic reasoning about incentives, sometimes called the principal and agent problem. It is stated as reasoning, not as an observed fact.

A business (the principal) hires an adviser (the agent) because the adviser knows the rules and the business does not. The business wants a claim that is as large as the rules support and no larger, because an overstated claim can come back as tax, interest and penalties. The adviser's pay depends on how the arrangement is written.

Under a percentage fee, each judgment call resolved in favor of including an activity, a person or an expense raises both the credit and the fee. Each call resolved the other way lowers both. The adviser's financial interest therefore points toward inclusion at every margin.
Under a fee for the work, the adviser's pay does not change with the answer to each call. The financial pull toward inclusion is removed, although other pressures remain, such as the wish to keep a client satisfied.

Three features of research credit work make this pull matter more than it might in other settings.

The consequences fall unevenly. If a claim is later reduced, the tax, the interest and any accuracy-related penalty are the taxpayer's.56 A paid preparer can face a separate penalty, but it is measured by reference to the preparer's own income from the return or claim, and in general applies only to positions without substantial authority (or, for disclosed positions, without a reasonable basis) that the preparer knew or reasonably should have known about.7 How these rules fall is explained in Who bears the risk of an overclaim.

The burden of proof sits with the business. The IRS audit guide states that the burden is on the taxpayer to establish that all of the section 41(d)(1) requirements have been met.4 The regulations require a taxpayer claiming the credit to keep records "in sufficiently usable form and detail to substantiate that the expenditures claimed are eligible".2 A business that relies on an adviser to decide what goes into the claim is still the party that must prove it.

The business may not be able to check the work. A business that hires an adviser to apply these tests may not be able to check how they were applied. That makes it hard for the business to see where judgment calls were made and which way they went. Information gaps for businesses looks at this in more detail.

What pulls the other way

An honest account of incentives has to include the forces that push against overstatement. These are general observations.

Preparer penalties. The preparer penalty described above gives a paid preparer a direct reason to avoid unreasonable positions.7
Professional rules. Treasury's rules for practitioners generally prohibit contingent fees for matters before the IRS, with exceptions, which limits these arrangements for practitioners those rules reach.8 The limits of those rules are covered in Professional standards and their limits.
IRS information requirements. For research credit refund claims, the IRS requires specific information about business components, research activities and qualified expenses at the time the claim is filed.9 This is explained in Amended and refund claims.
Reputation and client relationships. An adviser whose claims do not hold up may lose clients. This is a real pressure, but it acts slowly and depends on the business learning what went wrong.

Fixed fees have their own incentive problems. A fixed fee can reward doing less work than a claim needs, and a business should ask what a fixed fee covers. The difference is that a fixed fee does not tie the adviser's pay to the size of the number on the return.

What Treasury and the IRS have said about fee arrangements

When Treasury adopted the current contingent fee rule in 2007, it said the rule supports voluntary compliance "by discouraging return positions that exploit the audit selection process".10 In announcing its requirements for research credit refund claims, the IRS said that such claims "consume significant resources for both the IRS and taxpayers".11 Neither statement is a finding about any adviser or any claim, and we do not present them as one.

What this argument does not show

The incentive argument shows a direction of pull. It does not show how strong the pull is, how often it changes a claim, or which claims it affects. We are not aware of an official published measure of how fee structure relates to claim outcomes in the federal research credit, and we do not offer one. Our analysis of the economics of claim preparation, with its method set out, is in the white paper The economics of claim preparation.

Our position

Swanson Reed, the publisher of this site, is a fixed-fee R&D tax adviser and has a direct commercial interest in this question. Our position is that Congress should prohibit contingent fees in federal research credit engagements under Section 41. The credit depends on many judgment calls, and the taxpayer carries the burden of supporting each of them. In our view, the adviser who makes those calls should not be paid more for resolving them in favor of a larger claim. See Our position.

Sources

Office of the Law Revision Counsel, 26 U.S.C. 41, Credit for increasing research activities, subsections (a), (b)(1) to (b)(3), (c)(4)(A), (d)(1) to (d)(4), United States Code 2024 edition published by the Government Publishing Office. https://www.govinfo.gov/content/pkg/USCODE-2024-title26/html/USCODE-2024-title26-subtitleA-chap1-subchapA-partIV-subpartD-sec41.htm (accessed 11 October 2026). Back
Treasury Department and IRS, 26 CFR 1.41-4, Qualified research for expenditures paid or incurred in taxable years ending on or after December 31, 2003, paragraphs (a)(3)(i), (a)(5)(i), (a)(6), (b)(2) and (d). https://www.ecfr.gov/current/title-26/section-1.41-4 (accessed 11 October 2026). Back
Treasury Department and IRS, 26 CFR 1.41-2, Qualified research expenses, paragraphs (c)(2), (c)(3), (d)(1) and (d)(2). https://www.ecfr.gov/current/title-26/section-1.41-2 (accessed 11 October 2026). Back
Internal Revenue Service, Audit techniques guide: Credit for Increasing Research Activities (i.e. Research Tax Credit) IRC 41, Qualified research activities, chapter 5, part a and the section 174 test (publication date June 2005). https://www.irs.gov/businesses/audit-techniques-guide-credit-for-increasing-research-activities-ie-research-tax-credit-irc-ss-41-qualified-research-activities (accessed 11 October 2026). Back
Office of the Law Revision Counsel, 26 U.S.C. 6601, Interest on underpayment, nonpayment, or extensions of time for payment, of tax, subsection (a), United States Code 2024 edition published by the Government Publishing Office. https://www.govinfo.gov/content/pkg/USCODE-2024-title26/html/USCODE-2024-title26-subtitleF-chap67-subchapA-sec6601.htm (accessed 11 October 2026). Back
Office of the Law Revision Counsel, 26 U.S.C. 6662, Imposition of accuracy-related penalty on underpayments, subsections (a) and (b), United States Code 2024 edition published by the Government Publishing Office. https://www.govinfo.gov/content/pkg/USCODE-2024-title26/html/USCODE-2024-title26-subtitleF-chap68-subchapA-partII-sec6662.htm (accessed 11 October 2026). Back
Office of the Law Revision Counsel, 26 U.S.C. 6694, Understatement of taxpayer's liability by tax return preparer, subsections (a)(1) to (a)(3), United States Code 2024 edition published by the Government Publishing Office. https://www.govinfo.gov/content/pkg/USCODE-2024-title26/html/USCODE-2024-title26-subtitleF-chap68-subchapB-partI-sec6694.htm (accessed 11 October 2026). Back
Treasury Department, Circular No. 230, 31 CFR 10.27, Fees, paragraph (b). https://www.ecfr.gov/current/title-31/subtitle-A/part-10/subpart-B/section-10.27 (accessed 11 October 2026). Back
Internal Revenue Service, Research credit claims (Section 41) on amended returns frequently asked questions, introduction and question 21 (page last reviewed or updated 1 October 2026). https://www.irs.gov/businesses/corporations/research-credit-claims-section-41-on-amended-returns-frequently-asked-questions (accessed 11 October 2026). Back
Treasury Department and Internal Revenue Service, Regulations Governing Practice Before the Internal Revenue Service, final regulations, T.D. 9359, 72 FR 54540 (26 September 2007), preamble, "Contingent Fees", pages 54541 to 54542. https://www.federalregister.gov/documents/2007/09/26/E7-18918/regulations-governing-practice-before-the-internal-revenue-service (accessed 11 October 2026). Back
Internal Revenue Service, IRS sets forth required information for a valid research credit claim for refund, News release IR-2021-203, 15 October 2021. https://www.irs.gov/newsroom/irs-sets-forth-required-information-for-a-valid-research-credit-claim-for-refund (accessed 11 October 2026). Back

This article discusses fee practices in general terms. It is not a statement about any particular adviser.