I think Kitces' latest article isn't as rigorous as his usual content. The "step transaction doctrine" is basically a tool of statutory interpretation that a court can use as an interpretive aid to determine whether a given transaction falls within a particular statute. An application of the doctrine necessarily involves a close reading of the actual statutory language, which is notably absent from the Kitces article.
The case Kitces cites,
Gregory v. Helvering, 293 US 465, is illustrative of the above. In that case, the Supreme Court dealt with an alleged corporate "reoganisation" which achieved certain tax benefits. The Court was not impressed with the purported reorganisation and found that "[t]he whole undertaking ... was in fact an elaborate and devious form of conveyance masquerading as a corporate reorganization, and nothing else". In other words, the Court's decision was that the transaction was not a reorganisation within the meaning of the statute. The logic here is that, according to the Court, when Congress used the impugned terms in the statute, it intended to refer to only a sincere business transaction and not a sham transaction. The key thing to understand is that the Court's decision was rooted in an interpretation of the statutory language. In fact, the Court explicitly affirmed that "[
i]t is quite true that if a reorganization in reality was effected within the meaning of [the statute], the ulterior purpose mentioned will be disregarded". The problem for the taxpayer in that case was that his transaction fell outside of the statutory language upon which he had hoped to rely.
By contrast, Kitces does not base his argument on any statutory text. What Kitces is really applying is the "economic substance doctrine", which allegedly allows courts to analyse transactions on the basis of their true economic nature
without looking at the statutory language. The existence of the extra-statutory doctrine has been disputed. See
Grewal (2007).
26 USC § 7701(o)(5)(B) could also be relevant. This statute does not explain the economic substance doctrine, but it does say that this doctrine only applies to "transactions entered into in connection with a trade or business or an activity engaged in for the production of income". The transaction commonly referred to as the backdoor Roth IRA could conceivably fall within that test, since IRA contributions are related to "compensation", which comes from a "trade or business", but there are also arguments that it does not fall within that test.
If we want to apply specifically the "step transaction doctrine", we need an actual argument based in statutory language. Let's try to come up with one. The first attempt might be to argue that, as a matter of statutory interpretation, the backdoor contribution is part of the "aggregate amount of contributions for any taxable year to all Roth IRA" within the meaning of 26 USC § 408A(c)(2). This would cause it to be subject to normal income limits. The problem here is that 26 USC § 408A(c)(6)(B) explicitly says that "[a] qualified rollover contribution shall not be taken into account for purposes of paragraph (2)" (the provision previously cited). That statutory language cannot just be ignored when applying the step transaction doctrine (as opposed to the alleged extra-statutory economic substance doctrine). The next attempt might be to argue that the alleged rollover involved in the backdoor Roth IRA conversion is not a "qualified rollover contribution", but that term also has a statutory definition and it is clearly satisfied.
I cannot see any obvious argument based in the statutory language that would allow the "step transaction doctrine" to apply. That only leaves extra-statutory arguments. Professor Grewal says that these arguments have no basis in law, but even if we assume they do, there is at least a plausible argument that the backdoor contribution is a personal non-business transaction, exempted from the application of the economic substance doctrine by 26 USC § 7701(o)(5)(B).
I express no opinion on any of these issues, but the Kitces article would have been better if he had carefully addressed these matters.