Every company claiming the research credit makes a choice, whether or not anyone frames it as one. Take the full credit and reduce your domestic research deduction or capitalized amount by the same figure, or elect under section 280C(c)(2) to take a credit worth 79% of the full amount and leave the deduction intact. The election is made by checking a box on Form 6765, it cannot be changed later, and with immediate expensing restored under section 174A the arithmetic now favors the election for most profitable claimants.
Why the Credit Cannot Be Claimed and Deducted Twice
Section 280C prevents a double benefit. Research spending generates a deduction. The same spending generates a credit. Without a coordinating rule, a company would deduct a dollar of wages and also receive a credit computed on that dollar, recovering more than it spent.
The default rule reduces the related tax benefit. Domestic research or experimental expenditures otherwise taken as a deduction or charged to a capital account are reduced by the amount of the credit allowed. Where the expenditure would otherwise be currently deducted, that reduction increases current taxable income. Where it is capitalized, the capitalized amount is reduced instead.
The alternative reduces the credit instead. Electing under section 280C(c)(2) leaves the research deduction untouched and reduces the credit. The choice is between reducing the related deduction or capitalized amount and accepting a smaller credit.
The Two Choices and What Each Costs You
The reduced credit is the gross credit less the gross credit multiplied by the maximum corporate rate. At a 21% corporate rate, the election delivers 79% of the credit.
In the simplest current-year case, the comparison starts with your marginal rate.
| No election (full credit) | Election under 280C(c)(2) | |
| Credit received | 100% of the gross credit | 79% of the gross credit |
| Research deduction | Reduced by the gross credit | Unchanged |
| Taxable income | Increases by the gross credit | Unchanged |
| Net value on a $100,000 credit, 21% rate | $79,000 | $79,000 |
| Net value on a $100,000 credit, 37% rate | $63,000 | $79,000 |
| Qualified small business monetizing via the payroll offset | $100,000 | $79,000 |
The pattern is straightforward once it is set out. The reduction inside the election is fixed at 21% regardless of who you are, while the cost of declining is your own marginal rate. In a simplified federal current-year comparison, where both benefits are immediately usable, 21% is the break-even rate.
That is a starting point rather than the answer. The research credit is a general business credit and its use can be deferred, so a company without enough liability to absorb it does not receive full value in the year it is earned. Net operating losses, the position of pass-through owners and state treatment all move the comparison as well.
The last row is the clearest case, and it is narrow. A qualified small business actually monetizing the credit against payroll taxes receives cash now, so electing gives up 21% of a benefit it can currently use, in exchange for protecting a deduction with no current value. Even there the foregone deduction retains future value through the loss carryforward.
How the Election Interacts With Section 174A Expensing
This is where guidance published during the amortization years has gone stale. For tax years 2022 through 2024, section 280C(c)(1) used a different coordination rule: if the research credit exceeded the amount allowable as a deduction for qualified research or basic research expenses, the excess reduced the amount charged to capital account. That treatment often made the reduced credit election less attractive than it is under the restored section 174A regime.
Section 174A changed the comparison. With domestic research costs immediately deductible again for tax years beginning after December 31, 2024, the original 280C mechanics are back in force. A claimant either reduces the 174A deduction by the credit or elects the reduced credit. The deduction being protected is now a full current-year deduction rather than a fifth of one, which restores the election’s value for anyone paying tax above 21%. Our post on R&D tax credits and section 174 covers the expensing change itself.
The retroactive window has closed. Eligible small businesses that made the retroactive 174A election for 2022 through 2024 could also make a late 280C(c)(2) election or revoke a prior one for those years under Rev. Proc. 2025-28. The outside date was generally July 6, 2026, although the section 6511 limitations period could produce an earlier deadline for some 2022 returns. That special relief has expired, though ordinary amendment and administrative adjustment rules still apply to other corrections.
The Election Must Be Made on an Original Return
The election is made by checking the box on Form 6765 with an original, timely filed return, including extensions. It cannot be made or changed on an amended return, and once made it is irrevocable for that tax year.
Form 6765 has to be completed even if no credit is claimed on the original return. A company that expects to identify a credit later, and wants the reduced credit election available when it does, needs the form filed and the box checked with the original return. Our Form 6765 guide covers the mechanics.
The practical failure is one of sequencing. The election is often the last box anyone thinks about, checked or left unchecked while a return is being finalized, with the modeling never performed. Because it is irrevocable and cannot be corrected by amendment, that box deserves a decision rather than a default.
Modeling the Decision for Your Entity Type
C corporations taxed at 21% are close to indifferent on the federal arithmetic alone, which pushes the decision onto state conformity and the company’s loss position.
Passthrough entities are where the election most often wins. The credit and the add-back flow through to owners taxed well above 21%, so declining the election costs more than the 21% reduction inside it. The entity makes the election, and the owners bear the consequences. Most companies settle this alongside the credit computation itself as part of an R&D tax credit study.
Companies using the payroll tax offset are the clearest case for declining. The 21% reduction is a real reduction in cash received against payroll taxes, while the deduction being protected has no immediate value to a company with no taxable income.
Loss companies generally should think in terms of when the loss will be used rather than whether there is income this year, since an add-back that increases a net operating loss is not costless if that loss is carried forward and used.
State conformity is the item most often missed. Some states run their own version of the election and some do not follow the federal choice at all, so a federal election that improves the position can be neutral or unhelpful at state level.
If you are preparing a return with a research credit on it and the 280C box has not been modeled, that is worth resolving before filing. You can request a free analysis or estimate the credit with our R&D tax credit calculator.
Frequently Asked Questions
How much smaller is the reduced credit?
It is the gross credit reduced by the maximum corporate rate, which produces 79% of the gross credit at a 21% rate. The reduction is fixed at that rate whatever the taxpayer’s own marginal rate happens to be.
Can I change the election after filing?
No. The election must be made on an original, timely filed return including extensions, and it is irrevocable for that tax year. It cannot be made or changed on an amended return.
Should a startup using the payroll offset make the election?
Usually not. Electing reduces the credit by 21%, which is a direct reduction in the cash applied against payroll taxes. The deduction the election protects has little current value to a company with no taxable income to shelter.
Did section 174A change the answer?
For most profitable claimants, yes. During the amortization years the deduction being protected was spread over five years, which weakened the case for electing. With immediate expensing restored, the original 280C mechanics apply and the election recovers its value for taxpayers above the 21% break-even, subject to credit utilization and loss position.
Do states follow the federal election?
Not uniformly. Some states operate their own version of the reduced credit election and others do not conform to the federal choice. The federal and state positions should be modeled together rather than assuming the federal election carries through.
This article provides general information about federal tax rules and is not tax advice. Speak with your tax adviser about your specific circumstances.
