Every year we take on a client who is genuinely surprised to learn that Section 174 applies to them. They are not a laboratory. They do not employ scientists. In one memorable case the company manufactured commercial refrigeration units and had spent two years refining a defrost cycle, which is research whether or not anybody in the building would have used that word.
The confusion is understandable. Section 174 governs the treatment of specified research or experimental expenditure, and since the 2022 tax year that expenditure must be capitalized and amortized rather than deducted in the year incurred. Five years for domestic activity, fifteen for foreign. The name suggests white coats. The statute does not.
The test is about uncertainty, not job titles
Expenditure falls under Section 174 when it is incurred in connection with a trade or business and represents research and development costs in the experimental or laboratory sense, meaning activities intended to discover information that would eliminate uncertainty about the development or improvement of a product.
The operative word is uncertainty. If your team started a project not knowing whether the capability could be achieved, or not knowing how it would be achieved, that is the uncertainty the statute is describing. Software development is explicitly included. So is process improvement, formulation work, and the engineering effort behind a physical product.
If a project began with a genuine open question about whether or how something could be done, you are probably inside Section 174 whether or not anybody called it research.
Why this hurts more than it looks
The cash effect is the whole story. A company spending two million dollars a year on engineering salaries previously deducted two million against current income. Under the capitalization rules it deducts a fraction of that in year one, with the balance amortized across five years and a half-year convention on top.
Taxable income rises sharply while nothing about the business has changed. For a profitable company this is a timing problem that eventually evens out. For a company operating close to break-even, it can produce a tax liability in a year with no cash to pay it, which is how an accounting rule becomes a going-concern conversation.
- Wages of employees performing, supervising or directly supporting qualifying activity
- Supplies and materials consumed in the development effort
- Contract research paid to third parties, subject to specific allocation rules
- Software development costs, whether the software is sold or used internally
The interaction with the R&D credit
These are two separate provisions and companies routinely conflate them. Section 41 is a credit; Section 174 is a treatment. You can be required to capitalize under 174 while claiming nothing under 41, which is the worst of both positions and more common than you would expect.
The sensible response is to run both analyses together. Where a company must capitalize anyway, the incremental work to substantiate a credit claim is modest, and the credit frequently offsets a meaningful part of the cash cost the capitalization created. Handled separately, by different advisors, the numbers stop agreeing and both positions weaken.
A short test
Ask three questions about the last two years. Did any project begin without certainty that the objective was achievable? Did anyone iterate: build, test, discard, rebuild? Was the uncertainty technical rather than commercial? Three yeses and you should assume Section 174 applies until someone demonstrates otherwise.
If that describes your company and the last two returns treated engineering salaries as a current deduction, the position is worth reviewing before it compounds into a third year.
This note is general information, not advice. Tax and accounting outcomes turn on specific facts and the rules change. Speak to us, or to another qualified advisor, before acting on anything here.