Federal R&D relief does not finish your state tax work

Before adding federal R&D relief to a cash forecast, ask what the company already deducted on its state returns. A remaining federal balance can coexist with no remaining California balance for the same spending. Treating the federal number as a fresh deduction everywhere can materially overstate the benefit.

California also changed its research credit calculation options in 2025. That is a separate development from federal expensing. Finance teams need both a deduction reconciliation and a record of the state credit method actually elected.

This article examines California as a specific example, using official guidance checked October 7, 2026. It is not a fifty-state conformity survey. The worked figures are illustrative; return positions require review of the entity, tax year, elections and underlying activities.

One federal change, several state questions

Federal Section 174A allows current deductions for domestic research or experimental expenditures paid or incurred in tax years beginning after December 31, 2024, subject to applicable rules and elections. Transition provisions permit accelerated recovery of remaining domestic amounts capitalized for tax years beginning after 2021 and before 2025. IRS Revenue Procedure 2025-28, Sections 2.04 and 2.06.

A state review should establish four things: which federal provisions the state follows, any specific exceptions, when the provisions apply, and the company's existing state method. The date a return is filed is not necessarily the year to which the rule applies. Label the period with its actual start and end dates.

California's Franchise Tax Board expressly identifies Section 174A as a provision the state does not conform to. FTB analysis of Public Law 119-21, Section 70302.

That does not mean California newly requires every company to amortize its domestic research. FTB Publication 1001 explains that California continues to apply Section 174 as it read on January 1, 2015. The relevant state framework predates the federal mandatory capitalization regime. Establish the California treatment of each cost rather than inferring it from the phrase “nonconformity.” FTB Publication 1001, research and experimental expenditures.

Reconcile balances before modeling savings

Consider a fictional company preparing its calendar-year 2025 returns. Its adviser has confirmed these opening records for the same earlier domestic research costs:

Assume the company validly elects full federal recovery of the $420,000 remaining balance in 2025. The example does not establish that another election is available today. It describes a filing position already selected and documented.

The state schedule cannot generate another $420,000 deduction from costs already fully deducted there. The preparer must reconcile the federal recovery to California's different history. The table isolates cost recovery; it excludes credit coordination, losses, apportionment and other return adjustments.

This is an important control when moving advisers or accounting systems. A migration can preserve the federal fixed-asset or amortization report while losing the state adjustment workbook. Reconstruct the missing history from filed returns and supporting schedules before copying a federal balance into the state calculation.

For each spending year, retain the original amount, prior deductions, closing balance, method, elections and return reference. Give federal and state schedules distinct columns even when current-year treatment produces the same amount. Agreement in one year does not establish agreement across the life of the cost.

California's new credit option needs its own decision

Senate Bill 711 was chaptered on October 1, 2025. It moved California's general federal conformity date to January 1, 2025, with exceptions, and changed the research credit for tax years beginning on or after January 1, 2025. The old alternative incremental credit was repealed; the alternative simplified credit, or ASC, became available. A previous election for the old method does not automatically become an ASC election. FTB December 2025 Tax News.

Under the California ASC, the ordinary calculation is 3% of qualified research expenses exceeding half the average for the preceding three tax years. If any one of those years has no qualified research expenses, the ASC is 1.3% of current-year qualified research expenses. Make the ASC election on a timely filed original return; it continues in later years unless the FTB consents to revocation. 2025 Form FTB 3523 instructions.

For a fictional eligible company with $500,000 of current California qualified research expenses and $300,000 in each preceding year, the ordinary ASC calculation is ($500,000 − $150,000) × 3% = $10,500, before applicable adjustments and limitations. These assumed qualified amounts have already been reviewed. The example is a credit calculation, not a finding that the company's engineering budget qualifies or a promise of cash received.

Ask the preparer to document the chosen method and filing evidence. An unanswered field in a planning spreadsheet should remain unanswered until someone checks the actual return.

Location evidence and apportionment do different jobs

California's research credit requires research conducted within California. Keep the place of activity in the evidence set rather than deriving it from a headquarters address or a customer's location. FTB 3523 limitations.

Income apportionment answers another question: how business income is divided among jurisdictions. California's Schedule R addresses that calculation. A sales factor is not a substitute for identifying where research occurred. FTB Schedule R instructions.

For distributed teams, a useful record includes the person or contractor, service dates, employing or contracting entity, project, work location and supporting source. If someone moved during the period, preserve the change. Do not resolve conflicting records by choosing whichever location yields a larger credit.

AI infrastructure adds another practical distinction. A cloud billing region, engineer location and vendor address can describe different facts. Keep each in its own field. Let the adviser decide which facts control the specific cost and provision under review.

Build one evidence set with separate conclusions

The efficient workflow shares source records while preserving jurisdiction-specific decisions:

• Link project and payroll records to the relevant people, dates and entities.
• Reconcile current spending and historical balances independently.
• Save each jurisdiction's reviewed treatment, authority and election evidence.
• Record unresolved locations or allocations as questions with named owners.
• Update cash forecasts only after checking whether the reviewed position changes cash taxes in the modeled period.

AI can assist with identifying missing location fields or comparing supplied schedules. Require a source reference for every proposed correction, and retain human approval of the tax conclusion. Rewriting a project narrative cannot repair a missing deduction history.

Proforma's initial evidence review can begin with uploads and manual preparation. Its useful starting deliverable is a record of coverage, cost connections and open questions that a specialist can inspect. That gives the state review a stronger foundation without pretending one federal calculation answers every jurisdiction.

Research checked October 7, 2026. This comparison covers federal and California issues described above; other states, financial-reporting treatment and company-specific filing advice require separate analysis.