What is the difference between a TC opinion, a TC Memo, and a Summary opinion?

Division (TC) opinions decide novel legal issues and are reviewed within the Tax Court; they carry the most precedential weight. Memorandum opinions apply settled law to facts and, while citable, are formally not binding precedent. Summary opinions come from small tax case (S case) procedures, cannot be appealed, and cannot be cited as precedent at all.

What is the Golsen rule?

Under Golsen v. Commissioner, 54 T.C. 742 (1970), the Tax Court follows the precedent of the court of appeals to which a particular case is appealable, even when the Tax Court disagrees. So the same issue can come out differently for taxpayers in different circuits, and you must know your client's circuit before weighing Tax Court authority.

What is the difference between a holding and dicta?

The holding is the legal conclusion necessary to resolve the dispute actually before the court; it binds in future cases within the precedent's reach. Dicta are statements not necessary to the result — observations, hypotheticals, alternative rationales. Dicta can be persuasive but never controlling, and opposing counsel will discount authority that rests on it.

Does it matter whether a tax case was decided on summary judgment or after trial?

Yes. A summary-judgment ruling holds only that no genuine dispute of material fact barred judgment — its factual statements are assumptions viewed favorably to the non-movant, not findings. A post-trial opinion contains actual findings of fact. Citing a summary-judgment order for a factual proposition overstates it.

How to read a tax case: forum, posture, holding, and weight

A tax case is not a unit of authority; it is a bundle of variables — forum, procedural posture, opinion type, circuit, and the line between holding and dicta — and its weight in your research file depends on all of them. A Tax Court Memorandum opinion affirmed by your client's circuit is worth more than a division opinion appealable elsewhere; a Summary opinion is worth nothing as precedent no matter how apt its facts.

Reading a case well means extracting three things: what the court held, what facts drove the holding, and how far the holding reaches. This guide walks the anatomy, using the research-credit and cost-segregation cases briefed in our case-law cluster as working examples.

Which court, and why it matters

Federal tax litigation runs through three trial forums, and the choice — made by the taxpayer — shapes everything downstream. Our Tax Court versus refund forum guide covers the choice from the litigant's perspective; here is the researcher's view.

The forum tells you what the taxpayer paid, who found the facts, and where appeal lies.

Forum	Pay first?	Factfinder	Appeal to	Research significance
U.S. Tax Court	No	Judge (tax specialist)	Taxpayer's circuit	Most tax precedent lives here; Golsen applies
U.S. district court	Yes (refund suit)	Judge or jury	Local circuit	Generalist judges; jury findings unreviewable in practice
Court of Federal Claims	Yes (refund suit)	Judge	Federal Circuit	One appellate court; useful for circuit-split strategy

The Tax Court, established under Article I, hears deficiency cases before payment and decides the overwhelming majority of litigated tax issues — its judges see nothing but tax, and its opinions engage the regulations at a depth generalist courts rarely match. Search its opinions free at ustaxcourt.gov.

Refund forums matter to the researcher because they generate the circuit and Federal Circuit precedent that binds the Tax Court under Golsen. Fairchild Industries v. United States — a Federal Circuit funded-research decision — controls Court of Federal Claims cases like Dynetics but only persuades a Tax Court case appealable to the Ninth Circuit.

TC, TC Memo, and Summary opinions

The Tax Court issues three kinds of opinions, and the label is the first thing to check.

Division (TC) opinions — cited as, e.g., 159 T.C. No. 3 — address novel or important legal questions. They circulate to the full court before release and may be "reviewed by the court" (all judges), the strongest signal. Norwest v. Commissioner and Union Carbide fall in this tier for the research credit.

Memorandum opinions (T.C. Memo. 2019-57 and the like) apply settled law to particular facts. Formally the Tax Court says they are not binding precedent, yet practitioners and the court itself cite them constantly — most of the fact-intensive authority on qualified research, like Suder v. Commissioner and Siemer Milling, is memorandum-level. Treat a Memo as strong evidence of how the court analyzes facts, weaker evidence of pure law.

Summary opinions come from the small tax case ("S case") election available for disputes of $50,000 or less per year. They cannot be appealed and, by statute, cannot be treated as precedent. Read them for a judge's instincts if the facts are on point; never build a position on one.

Relative precedential weight of Tax Court opinion types (scale of 10)

Illustrative weighting for a research memo; actual weight varies with circuit posture and subsequent history.

One more category trips up researchers: orders. Populous Holdings, a favorite funded-research authority, is a summary-judgment order, not an opinion. Orders are not precedent at all, though they may be persuasive. Say so in your memo.

The Golsen rule: precedent depends on your client's zip code

Under Golsen v. Commissioner, 54 T.C. 742 (1970), the Tax Court follows the law of the circuit to which the case before it is appealable, even against its own considered view. Consequences for research:

A Tax Court opinion contrary to your client's circuit's precedent is nearly worthless for that client.
A circuit split means the "same" Tax Court issue has different answers by geography — identify the client's circuit before weighing anything.
Seventh Circuit affirmance of Little Sandy Coal on the substantially-all test, for instance, binds Tax Court cases appealable to the Seventh Circuit with full force, and persuades elsewhere.

Appeals from the Tax Court go to the circuit of the taxpayer's residence or principal place of business when the petition was filed (Section 7482(b)); corporate taxpayers and individuals map differently, so check.

Golsen also explains why the same Tax Court judge can write seemingly inconsistent opinions months apart: each case wears its own circuit's law. When your research surfaces two Tax Court opinions in tension, check their appellate venues before concluding the court changed its mind — often it didn't, and the "conflict" is just geography. A memo that flags this earns credibility; one that cites the favorable opinion without its circuit context invites the examiner to supply it.

Procedural posture: what was actually decided

Before quoting a case, identify how it arrived at decision:

After trial. The court found facts. Factual statements are findings and can be analogized to your client's facts.
Summary judgment. The court held only that no material factual dispute barred judgment, drawing inferences for the non-movant. Populous Holdings held the contracts were unfunded as a matter of law on undisputed terms — a real holding, but about contract language, not about disputed conduct.
Motion to dismiss / stipulated facts. Assumed or agreed facts, respectively; the legal ruling may be crisp but the factual gloss is thin.
Appellate review. Note the standard: legal conclusions reviewed de novo carry full weight; factual findings affirmed under clear-error review tell you less about how another court would find similar facts.

Posture also explains apparent conflicts. A taxpayer can "lose" at summary judgment (issue survives to trial) and win at trial; citing the first opinion without the second misstates the case — which is why subsequent-history checking, covered in Citators and currency, is not optional.

Holding versus dicta

The holding is the legal rule necessary to the result on the facts before the court. Everything else — hypotheticals, alternative grounds, general observations about the statute's purpose — is dicta. The discipline for separating them:

State the precise question the court had to answer to resolve the dispute.
State the answer and the facts that drove it. That is the holding.
Anything quotable that survives steps 1–2 removal is dicta.

Example: Siemer Milling is often quoted for broad statements about documenting a process of experimentation. Its holding is narrower — on the record presented, the taxpayer failed to show experimentation for specific projects. The broad language is guidance about proof, not a documentation requirement imposed by law. Your memo should use it accordingly, and the distinction matters when weighing authorities under the substantial-authority standard discussed in The substantial authority standard.

Reading the opinion efficiently

A repeatable pass through any tax opinion:

Caption and cite. Court, opinion type, year, judge. Docket number for unpublished items.
First and last pages. Tax opinions state the issues up front and the disposition at the end; read those before the middle.
Findings of fact. In Tax Court opinions these are a separate section — skim for the facts that resemble (or distinguish) your client.
The operative analysis. Find where the court applies law to facts; skip the boilerplate restatement of the statute you already know.
Concurrences and dissents. In reviewed opinions, they map where the law is unstable.

Briefing a case for the research file

A case brief in a specialty-tax research file is a half-page, structured so a reviewer (or an IRS agent, or you in three years) gets the weight assessment instantly:

Citation and opinion type (TC / Memo / Summary / order), with the deciding judge.
Circuit posture: where the case was appealable; whether appealed; result on appeal.
Procedural posture: trial, summary judgment, stipulated facts.
Question presented, one sentence.
Holding, one to two sentences, tied to the dispositive facts.
Useful dicta, labeled as such.
Relevance to our facts: analogous / distinguishable, and why.
Currency check: date verified, subsequent history traced.

Filed briefs in this form slot directly into the memo structure described in Building a research file, and they are what turns a stack of PDFs into a defensible position.

A worked contrast makes the discipline concrete. Little Sandy Coal v. Commissioner is frequently summarized as "shipbuilding isn't qualified research." The actual holding is about the substantially-all test of Section 41(d)(1): on the record presented, the taxpayer failed to show that substantially all of each vessel's development constituted elements of a process of experimentation, partly because it did not distinguish experimentation activities from the broader production effort. That holding travels to any industry with large business components; the shipbuilding gloss travels nowhere. Brief the rule, not the industry.

When case law is the wrong tool

Cases decide the disputes taxpayers chose to litigate, on the records they built — a skewed sample. Where a regulation squarely answers the question, the regulation controls and case-hunting adds noise. Where only Memorandum opinions exist and they conflict, the honest memo says the law is unsettled rather than cherry-picking. And for very recent statutes — Section 174A has essentially no case law yet — the research runs through statute, regulations, and legislative history instead. Knowing when the case layer is thin is itself a research finding.