The statute in one sentence
26 U.S.C. 41(d)(1)The term "qualified research" means research:
Internal Revenue Code, section 41(d)(1)
- with respect to which expenditures are treated as domestic research or experimental expenditures under section 174A,
- which is undertaken for the purpose of discovering information (i) which is technological in nature, and (ii) the application of which is intended to be useful in the development of a new or improved business component of the taxpayer, and
- substantially all of the activities of which constitute elements of a process of experimentation for a purpose described in paragraph (3).
Practitioners read that sentence as four questions, asked of every project. Is there a new or improved business component, for a permitted purpose? Does the work rely on hard science? Was there technical uncertainty at the start? Did a process of experimentation resolve it? The sections below take them in that order. For the plain-English version and where the work shows up by industry, read what qualifies for the credit.
A new or improved business component
Part oneA business component is "any product, process, computer software, technique, formula, or invention which is to be (i) held for sale, lease, or license, or (ii) used by the taxpayer in a trade or business of the taxpayer." The research must be intended to produce "a new or improved function, performance, or reliability or quality," and it is never treated as qualifying "if it relates to style, taste, cosmetic, or seasonal design factors."
Section 41(d)(2)(B) and 41(d)(3)
One purpose is enough
Function, performance, reliability, or quality. A project that improves any one of them clears this part. Style and taste do not count, which is why a redesign for looks alone stays out of a study.
The component is often your product
For an architect it is the drawing set, the manual for building the design. For an integrator it is the engineered system for one client's space. For a practice it can be a process: the protocol that folds a new device into care.
Each component stands alone
The test is applied to every component separately. A firm with twenty projects in a year has twenty questions, and the study claims the ones that pass.
The law shrinks back to what passes
When a whole product fails the test, the regulations look at its most significant subsystem, then the next, until they reach a part that passes or the most basic element that fails. A production line is treated as its own component, separate from what it makes.
Technological in nature
Part two"Information is technological in nature if the process of experimentation used to discover such information fundamentally relies on principles of the physical or biological sciences, engineering, or computer science."
Treasury Regulation 1.41-4(a)(4)
Rooted in hard science
The work has to rest on a science whose results are measurable and repeatable. Architecture rests on engineering and mathematics. Medicine rests on biological science. Control programming rests on computer science.
The soft sciences stay out
The statute excludes research in the social sciences, arts, and humanities. Marketing studies and management theory do real work, and neither satisfies this part.
Usually the easiest part
For most of the industries we serve the science is already in the work. The study's job is to name it, and to tie each project's uncertainty to it.
Common knowledge is fine
The regulations say the research need not seek information that "exceeds, expands or refines the common knowledge of skilled professionals" in the field. A patent is evidence, never a precondition.
Technical uncertainty at the start
Part three"Uncertainty exists if the information available to the taxpayer does not establish the capability or method for developing or improving the business component, or the appropriate design of the business component."
Treasury Regulation 1.41-4(a)(3)
Three kinds, any one will do
Capability: can it be done at all. Method: how it should be done. Appropriate design: what it should look like when finished. Design uncertainty is the one most projects carry, because nobody knows the final form at the start.
Administrative uncertainty is not technical
A permit that will not come or a budget that keeps moving. Those make a project hard for reasons the test ignores. The questions it counts are technical ones, like the load path of a structure or the contour of a custom orthotic.
The first one carries the most
Design a custom home and the uncertainty is real. Design three more like it and each carries less. The study sorts a project list on that line, and the unique work is what it claims.
Edge cases are where the time lives
Common cases are settled work. The patients and the sites that the standard approach does not fit are where the uncertainty sits, and if they are a fifth of the year, roughly a fifth of the time is eligible.
A process of experimentation
Part four"A process of experimentation is a process designed to evaluate one or more alternatives to achieve a result where the capability or the method of achieving that result, or the appropriate design of that result, is uncertain as of the beginning of the taxpayer's research activities." The requirement is met when "80 percent or more of a taxpayer's research activities, measured on a cost or other consistently applied reasonable basis," are elements of that process, and it "does not require that the taxpayer succeed."
Treasury Regulation 1.41-4(a)(5) and section 41(d)(1)(C)
Evaluating alternatives
Redlines and calculations for an architect. Bench testing and on-site problem solving for an integrator. Fitting and refitting for a practice. Each is a systematic way of trying one answer against another.
Failure still qualifies
A prototype that did not work or a design that was abandoned is evidence of experimentation. The regulations say so in as many words.
Substantially all means most of it
At least four fifths of the activity on a component has to be experimentation, measured by cost or another consistent basis. Routine production folded into a project leaves the project eligible; the production time itself is simply left unclaimed.
Commercialization is the cutoff
Once a component is ready for sale or regular use, the experiment is over. Work after that point is production, and the statute excludes it by name.
What the law rules out
Section 41(d)(4)- Research after commercial productionOnce the component is ready for sale or use, the work that follows is production.
- Adaptation of an existing componentFitting an existing product or process to one customer's needs, without technical change, is service rather than research.
- DuplicationReproducing an existing component from its physical examination or from plans is not development.
- Surveys, studies, and routine testingEfficiency surveys, management activities, market research, routine data collection, and routine quality control testing stay out. Testing that evaluates alternatives during development is a different thing, and it counts.
- Software built only for internal useSoftware your business runs on, rather than sells, faces an added set of requirements before it qualifies.
- Foreign researchOnly work performed in the United States, Puerto Rico, or a United States possession counts. Where the work happens governs, not where the business is registered.
- Social sciences, arts, and humanitiesThe second part of the test, restated as an exclusion.
- Funded researchWork that another party paid for, judged by who bears the financial risk and who keeps the rights. The next section takes this one apart.
Who holds the credit on contract work
Funded researchRisk decides it
The question is who would lose money if the work failed. Under a fixed fee, you deliver or you are not paid, so the risk and the credit are yours. Under an hourly arrangement, the client pays whether or not the work succeeds, so the risk sits with them.
Rights matter too
The party that keeps substantial rights to the results is the party that can claim. Most professional-services contracts leave those rights with the firm that did the work, and the study confirms it contract by contract.
Once the test is met, what counts
Section 41(b)Qualified research expenses are "in-house research expenses" and "contract research expenses." In-house expenses are "any wages paid or incurred to an employee for qualified services," "any amount paid or incurred for supplies used in the conduct of qualified research," and "any amount paid or incurred to another person for the right to use computers in the conduct of qualified research." Qualified services are "engaging in qualified research, or engaging in the direct supervision or direct support of research activities which constitute qualified research."
Section 41(b)(1) and 41(b)(2)
Wages, three ways
The people doing the work, plus the people directly supporting or supervising it. A principal reviewing redlines and a physician assistant recording a procedure both count. Administrative staff do not.
The four-fifths rule for people
An employee who spends 80 percent or more of the year on qualified services is treated as spending all of it. For a team dedicated to development, the credit grows quickly.
Supplies and computing
Materials consumed in the experiment, including prototypes that never shipped, and cloud computing used to run it. Equipment that is depreciated is not a supply.
Contractors at 65 percent
The statute counts 65 percent of what you pay an outside party for qualified research performed in the United States, provided you bear the risk and keep the rights.
What the record has to show
SubstantiationUsable form and detail
The regulations require records kept "in sufficiently usable form and detail to substantiate" the expenses claimed. They do not prescribe a format, and they do not require hour-by-hour time tracking.
The evidence you already keep
Drawings, redlines, change orders, test results, fabrication logs, and follow-up notes tie each project to the four parts. Interviews with the people who did the work produce reasonable estimates of time, checked against what a firm like yours would plausibly spend.
The study is the record: each component, each part of the test, and each expense, written for the examiner before one ever calls.
Common questions
Does the work have to succeed?
No. The regulations state that qualifying "does not require that the taxpayer succeed in developing a new or improved business component." The attempt is what the test measures.
Does it have to be new to the world, or patentable?
No. The standard is new to your company. The regulations say the research need not exceed the common knowledge of skilled professionals in the field, and a patent is evidence rather than a requirement.
What does "substantially all" mean?
At least 80 percent of the activities on a business component, measured by cost or another consistently applied basis, have to be elements of a process of experimentation.
What if only part of a project qualifies?
The regulations shrink the test back to the most significant subsystem that passes, so a qualifying part of a larger product can be claimed on its own.
Do permits, approvals, and budgets count as uncertainty?
No. They are administrative uncertainty. The test asks about capability and method, and about the appropriate design. Those are technical questions.
Does routine testing count?
Routine quality control and routine data collection are excluded by statute. Testing that evaluates one alternative against another during development is experimentation, and it counts.
Our customer paid for the project. Whose credit is it?
It depends on the contract. If you bear the financial risk of the work and keep substantial rights to the result, the credit is yours. Fixed-fee arrangements usually put it with you; hourly arrangements usually do not.
Does the test change for a small practice or a one-person firm?
No. The test is applied to each business component regardless of the size of the business. Smaller firms produce smaller credits, and the credit grows as the work does.
Where does the four-part test come from?
Section 41(d) of the Internal Revenue Code and the Treasury regulations under it, chiefly regulation 1.41-4. The statute names the requirements; the regulations define uncertainty, technological in nature, and the process of experimentation.
Thirty minutes shows whether your work passes.
Book a call and bring the person who knows the work. We will walk the four questions against your projects and tell you plainly whether a study is worth doing.
Book a call Or write to inquiries@thefoundryfirm.com