Section 174A Election Calculator: Expense or Capitalise R&E

Why would anyone elect to capitalise R&E when expensing is available?

Because of what the two deductions do to adjusted taxable income, and it runs opposite to intuition.

The One Big Beautiful Bill Act created section 174A, permanently restoring immediate expensing of domestic research and experimental costs for tax years beginning after 31 December 2024. It also permanently restored the add-back of depreciation, amortisation and depletion when computing ATI for the section 163(j) interest limitation — the EBITDA-style measure that applied before 2022.

Put those two together and a trap appears. An immediate deduction under 174A(a) reduces ATI, because it isn't amortisation and gets no add-back. An amortisation deduction under the 174A(c) election does not reduce ATI, because it is added straight back. Grant Thornton states it plainly: electing to capitalise and amortise domestic R&E would result in a higher business interest limitation compared with immediate expensing.

The calculator's defaults show the gap. On $8,450,000 of deductible R&E, expensing drops ATI from $30,000,000 to $21,550,000 and the interest limitation from $9,180,000 to $6,645,000. Disallowed interest goes from $620,000 to $3,155,000.

The limitation you give up is exactly 30% of the deduction — $2,535,000 here — which is the whole mechanism in one line.

How do I actually choose between the two elections?

Compare present values, not first-year tax. First-year tax will almost always favour expensing and it is the wrong test.

At the defaults, expensing saves $1,064,700 of tax this year. That looks decisive. But $2,535,000 of interest deduction has been pushed into an indefinite carryforward, and its value depends entirely on when you can use it.

Discount both paths and the picture tightens:

Expensing still wins — by $176,723. But a model that ignored the interest limitation would have reported $331,943. The limitation quietly consumed $155,220, or 46.8% of the headline advantage.

The decision rule the calculator produces is the useful output: the two elections tie at 11.34 years of recovery delay. If you expect to absorb the disallowed interest sooner than that, expense. Longer, capitalise.

And if the carryforward is never usable — a plausible state for a company that stays leveraged and research-heavy — the advantage flips to negative $200,407 and capitalising is the better election outright.

What does the catch-up deduction do to all this?

Nothing, and that is the good news — but only because of a specific procedural ruling.

Taxpayers who capitalised domestic R&E across 2022 to 2024 under the TCJA version of section 174 can recover the unamortised balance, either entirely in the first tax year beginning after 2024 or rateably across that year and the next.

The open question was whether that recovery counted as amortisation for ATI purposes. If it didn't, a large catch-up would have crushed the interest limitation in the year it was claimed. Rev. Proc. 2025-28 resolved it: the deduction of unamortised domestic R&E expenditures is amortisation for federal income tax purposes, including for computing the business interest limitation.

At the defaults, the $7,200,000 catch-up is added straight back to ATI, preserving $2,160,000 of interest limitation that would otherwise have vanished. It delivers $1,512,000 of tax benefit with no interest cost at all.

The practical consequence is a clean split. The catch-up is free of the ATI problem; only your current-year expensing election carries it. Those are separate decisions and should be modelled separately — which is why they sit in separate sections here.

One thing to watch: if the catch-up is large enough to drive taxable income negative, the excess becomes a loss carryforward rather than a current benefit. The calculator flags that amount.

What else is different about section 174A beyond the ATI question?

Three things, and two of them cost money if missed.

Foreign R&E did not change. Research conducted outside the United States and its territories remains on 15-year amortisation under section 174. There is no election and no immediate deduction. Foreign software development costs stay there too. The calculator takes foreign amortisation as a separate input for exactly this reason.

Section 280C was amended. The deduction under 174A must be reduced by the amount of the section 41 research credit, unless you take the reduced-credit election. At the defaults, a $9,600,000 spend with a $1,150,000 credit produces an $8,450,000 deduction, not $9,600,000. Modelling the gross figure overstates every downstream number.

The election is annual and permanent in form. Section 174A(c) capitalisation can be elected for any tax year over a period of not less than 60 months, and a separate 10-year election is available under section 59(e). This is a recurring planning decision, not a one-off transition choice — which is the main reason this calculator is worth keeping rather than running once.

Also worth knowing: for alternative minimum tax purposes, R&E deducted under 174A generally must be amortised over 10 years regardless of the regular-tax treatment. That is outside this model.

When does this model stop being the right tool?

Four boundaries, and the first two matter most.

One year, one entity. The model prices a single tax year against a single amortisation schedule. A multi-year projection where ATI, interest expense and R&E spend all move together needs a full forecast, and the answer can differ — particularly if a future year has spare limitation capacity that absorbs the carryforward cheaply.

The recovery-year assumption is yours. Everything here turns on when disallowed interest becomes usable, and the calculator cannot know that. It takes it as an input and solves for the breakeven so you can see how sensitive the answer is. If you cannot defend the recovery year, the honest reading is the bracket: expensing is worth between negative $200,407 and positive $331,943 depending on it.

No CAMT, BEAT, FDII or state conformity. All four interact with this election and several push in opposite directions. Immediate expensing can affect base erosion percentage for BEAT purposes, and state conformity to 174A is uneven. A large taxpayer needs those modelled alongside.

Not a substitute for the method change. Adopting immediate expensing is an accounting method change reported with the return, and the transition elections have their own procedural requirements under Rev. Proc. 2025-28. This tells you which answer is worth more; it does not file anything.