Tax Research · Brief · Working level
The substantial authority standard: weighing what you found
Substantial authority under Section 6662 is an objective weighing of listed authorities for and against a return position — roughly a 40% standard. Here is the authority list, the weighing rules, and how credit positions typically score.
"Substantial authority" is the standard that decides whether a return position, if it loses, drags a 20% accuracy-related penalty behind it. Under Treas. Reg. §1.6662-4(d), the test is objective: the weight of authorities supporting the position must be substantial in relation to the weight of those against it — conventionally pegged near a 40% chance of success, stricter than reasonable basis, looser than more-likely-than-not. It is, in other words, the scoring rubric for everything your research file contains, and the reason the file exists; the penalty mechanics themselves are covered in our Section 6662 guide.
The authority list, ranked in practice
The regulation lists what counts but does not rank it; rank emerges from the weighing rules — an authority on point and well reasoned beats one that merely recites conclusions, and a document's pedigree sets its ceiling. In practice the tiers align with the hierarchy in our authority-hierarchy guide:
Everything below is "authority"; weight falls roughly in this order.
| Tier | Authorities | Weighing notes |
|---|---|---|
| Controlling | Code; final and temporary regulations | Nearly conclusive when on point |
| Strong | Court cases (circuit of appeal first); tax treaties | Weight varies with court, posture, reasoning |
| Official IRS | Revenue rulings and procedures; notices | IRS-binding; courts merely respect them |
| Legislative | Committee reports; statement of managers; bluebook | Strong on intent questions |
| Interpretive | Proposed regulations; regulation preambles | Real weight, especially if reliance-authorized |
| Non-precedential | PLRs, TAMs, CCAs, AODs, IRS releases | Diminished if old or issued to others |
Not on the list at all: treatises, law review articles, opinions of counsel, and this website. They can point to authority; they are not authority.
The weighing rules that decide close cases
Four rules from §1.6662-4(d)(3) do most of the work. Relevance and persuasiveness govern: a case with facts materially like your client's outweighs a doctrinally grander case that is distinguishable, and any authority is discounted if later developments undercut it — which is why the currency checks in Citators and currency feed directly into the score. Age matters for the bottom tier: an old PLR interpreting since-amended law approaches zero. Contrary authority must be weighed, not omitted: a conclusion of substantial authority reached by ignoring the adverse Memorandum opinion is professionally indefensible and practically useless. Timing is fixed: authority is measured when the return is filed (or at year-end, at the taxpayer's option), so later-decided cases neither help nor hurt.
Note what the standard is not: it is not "some authority exists." A well-reasoned construction of the statute alone can suffice — the regulation says so — but a hunch plus a treatise cannot.
Scoring typical credit positions
Realistic Section 41 examples, as of mid-2026:
- Fixed-price contract research claimed by the payor, contracts silent on rights: statute and Treas. Reg. §1.41-2(e) support; Fairchild and Dynetics cut both ways on risk; likely substantial authority if the contract terms parallel the favorable cases, but the memo must engage the adverse ones.
- High officer wages as qualified wages with thin time records: Suder supports inclusion of executive wages in principle while cutting the amount as unreasonable; substantial authority for a moderated figure, doubtful for the full number — substantiation cases like Cohan do not rescue Section 41 estimates.
- Routine software configuration as qualified research: published guidance and the process-of-experimentation cases run against; typically fails even reasonable basis. Disclosure on Form 8275 does not help a position with no reasonable basis, and no research file can convert this one.
The honest deliverable is a stated confidence level per issue, not a blended blessing.
Frequently asked questions
- What does substantial authority mean for a tax return position?
- Under Treas. Reg. §1.6662-4(d), substantial authority exists when the weight of authorities supporting a position is substantial relative to the weight of contrary authorities — an objective standard commonly benchmarked around a 40% likelihood of success. It is stricter than reasonable basis (roughly 20%) and looser than more likely than not (over 50%).
- What counts as authority for the substantial authority standard?
- Only items on the regulation's list: the Code and other statutes, regulations (proposed, temporary, final), revenue rulings and procedures, court cases, tax treaties, committee reports and the JCT bluebook, private letter rulings, TAMs, actions on decision, IRS information releases, and notices. Treatises, articles, and practitioner opinions are not authority — though the sources they cite may be.
- Why does substantial authority matter if the position is correct?
- Because it is the penalty shield for undisclosed positions. The Section 6662 substantial-understatement penalty is 20% of the understatement, and substantial authority prevents the item from counting toward the understatement even if the position ultimately loses. Without it, an undisclosed losing position needs reasonable cause to escape the penalty.