Incorporator The Journal of U.S. Formation
Start Free Quiz
Editorial 8 MIN READ

The Section 174 fix, finally (maybe): what OBBBA would do

Four tax years of capitalized R&D, a failed Senate cloture vote in 2024, and a June 2025 reconciliation bill that could make it go away

The Editors
Staff, Incorporator.org
8 min read Updated
Contents 6 sections
  1. What Section 174 actually requires, in operational terms
  2. The failed fixes
  3. OBBBA, as drafted on June 10
  4. What a retroactive fix actually does to a 2022 return
  5. Second-order effects, and what to actually do
  6. Sources

or four tax years, a software startup that spent a dollar on an engineer's salary could deduct roughly twenty cents of it in the year the dollar left the bank account. The Section 174 R&D capitalization rule, enacted by TCJA § 13206 in 2017 and effective for tax years beginning after December 31, 2021, is the single largest unforced error in recent federal tax policy, and the One Big Beautiful Bill Act moving through reconciliation this month is the closest Washington has come to undoing it.

Whether OBBBA actually reaches the President's desk before the extended filing deadline in October is the question running every startup tax-planning meeting in the country right now.

What Section 174 actually requires, in operational terms

Before TCJA, a taxpayer could deduct research or experimental expenditures in the year paid or incurred under the pre-2022 version of IRC § 174(a). TCJA struck that and replaced it. The version of § 174 in force for tax years beginning after December 31, 2021 requires "specified research or experimental expenditures" to be charged to a capital account and amortized ratably over five years for domestic research, or fifteen years for research attributable to activities conducted outside the United States. The amortization begins at the midpoint of the tax year in which the expenditure is paid or incurred, which means a first-year deduction of 10% for domestic research (a half year of a five-year life), 20% in each of years two through five, and a final 10% in year six.

The practical effect on a bootstrapped or lightly funded software company is severe. If the company paid $2,000,000 in engineering salaries in 2024, and those salaries are treated as software-development costs (which § 174(c)(3), added by TCJA, explicitly makes them), then $200,000 is deductible in 2024 and the remaining $1,800,000 sits on the balance sheet amortizing out over the next five years. A company that broke even on a cash basis in 2024 therefore has $1,800,000 of taxable income and a federal tax bill of around $378,000 at the 21% corporate rate, on cash it does not have.

Notice 2023-63, published by the IRS in September 2023, remains the interim guidance taxpayers have been relying on for scope and method questions, and the IRS stated that taxpayers could rely on the Notice for tax years ending after September 8, 2023 pending issuance of proposed regulations. Those proposed regulations have not issued as of this dateline. Notice 2024-12, issued in December 2023, clarified one scope question around contract research and modified a portion of the earlier Notice. Everything else is still interim.

The failed fixes

The rule was never supposed to survive. The 174 amortization was scored, on paper, as a revenue raiser to offset other provisions of TCJA, on the assumption that a future Congress would repeal it before it took effect. Congress did not.

The first serious attempt was H.R. 7024, the Tax Relief for American Families and Workers Act, introduced by Ways and Means chair Jason Smith and Finance chair Ron Wyden. The House passed it on January 31, 2024 by a vote of 357 to 70, a margin that in a less broken Congress would have meant near-certain Senate passage. The bill would have restored immediate expensing for domestic research retroactively to tax years beginning after December 31, 2021, and kept the fifteen-year amortization for foreign research. It also restored 100% bonus depreciation under § 168(k) and expanded the child tax credit.

The Senate took it up in pieces through the spring and summer of 2024 and then, on August 1, 2024, failed a cloture motion 48 to 44, well short of the 60 needed. Senate Republicans, uninterested in handing the Biden administration a policy win in an election year, let the bill die. Taxpayers who had filed extensions and waited for the retroactive fix filed their 2023 returns with capitalized 174 costs and hoped for 2025.

OBBBA, as drafted on June 10

The reconciliation vehicle that emerged after the November 2024 election is being marketed by House Republicans as the One Big Beautiful Bill Act. The House Budget Committee reported out its reconciliation instructions package on May 14, 2025, combining the work product of the eleven committees given reconciliation directives under the FY 2025 budget resolution. The tax title, which came out of Ways and Means, contains two provisions that matter here.

The first restores immediate expensing for domestic research or experimental expenditures, effective retroactively for tax years beginning after December 31, 2021. Taxpayers who capitalized domestic research in 2022, 2023, and 2024 would be allowed to recover the unamortized balance, and the mechanism in the markup text contemplates a one-time adjustment taken over either a single year or a two-year period at the taxpayer's election. Foreign research stays on the fifteen-year schedule, consistent with the 2024 House-passed structure.

The second restores 100% bonus depreciation under § 168(k) for property placed in service after a date tied to enactment, reversing the phased drawdown that took bonus to 60% in 2024 and 40% in 2025 under current law. This is a separate fight but moves on the same vehicle.

The bill as reported is draft text. The Senate Finance Committee has its own ideas, the Byrd rule will knock out anything it touches, and the reconciliation instructions carry a deficit ceiling that forces trade-offs somewhere in the package. The 174 restoration is, by most accounts, one of the provisions with the broadest bipartisan support on the merits and therefore one of the provisions most likely to survive the Senate process intact. That is a low bar.

What a retroactive fix actually does to a 2022 return

Assume a calendar-year C-corp that had $1,000,000 of domestic § 174 expenditures in 2022. Under current law it deducted $100,000 in 2022, $200,000 in 2023, $200,000 in 2024, and would deduct $200,000 in 2025, $200,000 in 2026, and $100,000 in 2027. If OBBBA passes as drafted with retroactive effect to TY 2022, the taxpayer has $700,000 of remaining unamortized basis at the start of the first tax year beginning after enactment. That $700,000 either collapses into a single year of deduction or splits across two years, depending on the election mechanism in the final text and any safe-harbor regulations Treasury issues.

For a company that was profitable in those years, a retroactive fix means amended returns are the cleanest way to capture the refund value. For a company that had NOLs, the mechanism matters less in cash terms and more in the carryforward profile. For a company that was acquired in 2023 or 2024, the question of who captures the refund belongs in the acquisition agreement, and if the agreement is silent, it belongs in arbitration.

The IRS's existing machinery for a change in method of accounting sits under Rev. Proc. 2025-8, which updated the list of automatic changes for 2025, and Form 3115. Any retroactive restoration in OBBBA will either ride on that machinery with a new designated change number, or Treasury will issue a standalone procedure. The Form 3115 is not optional paperwork; a taxpayer that changes its § 174 treatment without filing one does so improperly and invites a § 481(a) adjustment on audit.

Second-order effects, and what to actually do

The four-year distortion has already happened. Software startups with significant salary spend and modest revenue have either raised bridge capital they did not want, cut headcount they did not want to cut, or burned through reserves to pay a tax bill on paper profits. The Section 41 research credit helps at the margin but does not make up the difference; a credit on a return is worth twenty-one cents on the dollar of denied deduction at the current corporate rate, and many small-company filers take the payroll-tax election under § 41(h) instead, which caps the benefit at $500,000 per year against employer FICA.

The practical advice for a taxpayer sitting on capitalized 174 balances as of this month is to do nothing irreversible. Do not amend 2022, 2023, or 2024 returns on the assumption that OBBBA will or will not pass. If the bill passes with retroactive effect, the taxpayer will have a clean procedural path under a new Rev. Proc. or an updated § 481 automatic change. If the bill fails or passes without retroactive effect, the capitalized balances continue to amortize on their current schedule and the company plans accordingly for 2025 estimates. Filing extensions are cheap; panic is expensive.

For founders considering entity choices in the middle of this, the calculus has not fundamentally changed. A C-corp under § 174 suffers from the capitalization rule directly on its own return. An S-corp or partnership passes the capitalization through to its owners, who then pay the individual-rate tax on phantom income; the math at the owner level is usually worse, not better, because individual rates exceed the corporate rate in most cases. The entity form does not fix § 174. Only Congress does.

If the bill does pass in a form close to the May 14 markup, the story of this era of startup tax policy will be that Congress imposed a rule it never intended to take effect, let it run for four years, and unwound it with interest. If it does not pass, the capitalization rule will quietly continue, and the companies that survived the squeeze will be smaller than they should have been.

Sources

Keep reading

More from the journal.