Section 338 and the Step Transaction Doctrine

Abstract

Section 338 overrides ordinary step transaction principles for a large but not clearly defined class of stock purchases. This Article addresses two questions about the scope of the override: whether it applies to a larger set of cases than those in which the acquiring corporation has made a qualified stock purchase, and whether the Treasury Department has authority to adjust the scope of the override by regulation. The Article concludes that the class of override cases is in fact larger than the class of qualified stock purchases, but that the Treasury Department lacks the authority under current law to adjust the scope of the override.

I. Introduction

Much ink already has been spilled over the interaction of Kimbell-Diamond Milling Co. v. Commissioner (K-D) and section 338, but significant open questions remain. K-D stands for the proposition that the economic substance of a stock purchase followed promptly by the liquidation of the purchased corporation (“Target,” or T) is an asset purchase, assuming the two transactions are part of a single plan. Section 338 generally overrides the application of K-D by providing for an election in certain circumstances to treat the taxable acquisition of the T stock as an asset purchase, without regard to whether the purchasing corporation (P) actually liquidates T. Thus, when section 338 applies, the transaction will be treated as a stock purchase even if its economic substance is an asset purchase, as long as P makes no election under section 338(g). Conversely, if P makes the election, the transaction is treated as an asset purchase even if T survives as a matter of corporate law.

In general, the section 338 override applies to all cases in which a stock acquisition would constitute a “qualified stock purchase” (QSP). For this purpose, a QSP is an acquisition, within a 12-month period, of an amount of T stock constituting control of T as defined in section 1504(a)(2) by means of one or more taxable purchases. Nevertheless, questions persist about the policy rationale that supports section 338 and, therefore, about the proper scope of the section 338 override. This Article addresses two of them: (1) whether and to what extent the override of step transaction principles extends beyond the circumstances in which P has made a QSP, and (2) whether the Treasury Department (“Treasury”) has authority to “turn off” step transaction principles when an acquisition viewed in isolation is a QSP but viewed in context forms part of a larger transaction that otherwise qualifies as a reorganization.

II. K-D and Substance Over Form

K-D was not the first case to analyze the economic substance of an integrated stock purchase and liquidation, but it is the leading case identified with the proposition that such a transaction is in substance an asset purchase. In K-D, Kimbell-Diamond bought all the stock of Whaley Mill & Elevator Company and, pursuant to a pre-existing plan, promptly liquidated Whaley. The prospect of significant tax benefits drove the transaction structure. For unusual reasons, Kimbell-Diamond’s basis in the Whaley stock was low (indeed less than its purchase price for the stock), but Whaley’s basis in its assets was high. If the form of the transaction were respected, the built-in gain (BIG, or BIG/L where loss also may be realized) in the Whaley stock would disappear and Kimbell-Diamond would inherit Whaley’s asset basis under the rules for the liquidation of a controlled subsidiary. By contrast, the more direct route to the same result—a purchase of Whaley’s assets—would have given Kimbell-Diamond the same low asset basis that it had in the Whaley stock.

The government prevailed in its challenge to Kimbell-Diamond’s treatment on a step transaction theory according to which the stock purchase should be disregarded and Kimbell-Diamond treated for tax purposes as having purchased Whaley’s assets. The argument for disregarding the stock purchase was that Kimbell-Diamond had no interest in owning the Whaley stock and had inserted that step solely for the anticipated tax benefits. Under substance-over-form principles, such a stock purchase was in substance no purchase at all, a meaningless step in what was in economic reality an asset acquisition.

The K-D court accepted the government’s argument, but, as is often the case in tax matters, a shield for one side became a sword for the other on different facts, and the decision provoked both creative tax planning and plenty of litigation as taxpayers sought to invoke or not invoke its rationale in various factual settings. Without rehearsing the full history here, suffice it to say that the factual question of buyer intent made the doctrine unwieldy if not unworkable. P might cherry-pick high-basis assets for direct purchase from T and then purchase the T stock in an assertedly separate transaction that eliminated corporate-level gain. Or, more simply, P might seek the same treatment as did Kimbell-Diamond but, forewarned by K-D itself, purchase the stock and then age it until the subsequent liquidation would not be stepped together with the asset purchase. Other potentially abusive variations also were possible.

III. Section 338

Following an earlier, largely unsuccessful attempt at a statutory fix, Congress sought a definitive end to the three-way tug-of-war among taxpayers, the government, and the courts with the enactment of section 338 as part of the Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA). Rather than try to develop a statutory (or regulatory) mechanism for distinguishing taxable asset purchases from taxable stock purchases followed by liquidations, Congress made the status of the transaction mostly elective, at least as regards the tax treatment of P. If and only if P meets the requirements of section 338, P may elect (or will be deemed to have elected) to treat the stock purchase as an asset purchase; if no election is made (or deemed), the transaction is treated as a stock purchase. In this respect, section 338 reflects in part Congress’s decision to throw in the towel on policing tax-motivated stock purchases. By making the treatment elective, it conceded that taxpayer efforts to avoid application of the K-D rule had been largely successful. At the same time, however, the section 338 regime limits the ability of taxpayers to pursue arguably inconsistent paths leading to a better tax outcome under prior law. Among other things, it restricts P’s power to select certain tax attributes by eliminating, or at least dramatically curbing, the opportunity for tax electivity with respect to the basis P takes in T’s assets.

A. Outline of Section 338

Section 338 applies when P makes a QSP. In general, a QSP is one or more taxable purchases of T stock over a period of not more than 12 consecutive months in which P acquires in the aggregate at least 80% of the vote and 80% of the value of T. If P makes a QSP, then P can elect under section 338(g) to treat the transaction as an asset purchase, indeed even without regard to whether it liquidates T thereafter. If no election is made, the liquidation (as well as any other post-acquisition transactions involving T) is accorded separate significance for tax purposes. Thus, if section 338 were applied to the facts of K-D itself, the transaction would be respected according to its form as a stock purchase followed by a controlled subsidiary liquidation, assuming (as would be the case given that Kimbell-Diamond wanted the subsidiary’s high asset basis) that Kimbell-Diamond did not make the section 338 election.

A corollary to section 338’s elective regime is that the step transaction doctrine has to be “turned off” if P does not make the election. Otherwise, the regime could be defeated by post-acquisition transactions that would again require analysis under step transaction principles to ascertain the overall tax treatment of the stock purchase and subsequent transactions. The point of the regime was to replace that substance-over-form analysis with an election.

A second corollary, however, seems to be that step transaction principles would remain in effect when the steps constituted a tax-deferred transaction, assuming that the steps were part of an integrated plan. The logic is that section 338 represents an override to the usual principles of analysis, which include application of the step transaction doctrine, in the narrow situation in which P makes a purchase of T that would result in P’s taking a cost basis in T’s assets under the step transaction principles of K-D. If a taxable asset-qua-stock purchase is a predicate to the application of section 338, the threshold question of whether section 338 applies cannot be answered by turning off the step transaction doctrine because it is only section 338 that turns it off. In the case of a series of steps that constitute a tax-deferred transaction under step transaction principles, no K-D transaction resulting in a cost basis to P in the T assets has taken place, even if one or more steps taken in isolation would constitute a QSP; therefore, step transaction principles remain in effect.

As developed below, this second corollary suggests a third, largely unnoticed corollary: section 338 must be understood to “turn off” nonrecognition rules in certain limited cases. One question raised thereby, and also addressed below, is just how widely applicable this third corollary may be.

B. Applicable Rulings

It took the government some time to recognize the full implications of the enactment of section 338 on the K-D doctrine. This Subpart traces the development of the government’s approach through several key revenue rulings.

1. Revenue Ruling 90-95—Application of Step Transaction Principles to Determine Whether a QSP Has Occurred

In Revenue Ruling 90-95, the Internal Revenue Service (the Service) clarified that the existence or not of a QSP depends on application of the substance-over-form and step transaction principles that ordinarily apply in analyzing the tax consequences of multi-step transactions. Under the facts of that ruling, P acquired T in a reverse subsidiary cash merger pursuant to which P’s subsidiary, S, was formed solely for the purpose of effectuating the acquisition, with S then merging into T and going out of existence. Under step transaction principles, the existence of S was properly disregarded, which meant that the economic substance of the merger was P’s acquisition of the T stock in a purchase transaction and therefore a QSP. The subsequent planned merger of T into P consequently was treated as a liquidation under section 332, not as part of the original acquisition of T, because the QSP turned off the step transaction doctrine and P did not make a section 338 election. The ruling stands for the proposition that the existence of a QSP depends on not only the characteristics of the transaction in which P acquires the T stock but also whether the overall transaction is in economic substance a QSP, applying the usual step transaction principles. Notably, because the merger was all cash, even when integrated with the subsequent liquidation it would not qualify as a reorganization.

2. Revenue Ruling 2001-46—Section 338 and Reorganizations

In Revenue Ruling 2001-46, the Service turned to the interaction of section 338 and step transaction principles in the reorganization setting. Pursuant to an integrated plan, P acquired all of the T stock by forming a merger subsidiary and causing the subsidiary to merge into T. P then caused T to merge into P pursuant to state law. The T shareholders received consideration consisting of 70% P stock and 30% cash. Viewed in isolation, P’s acquisition of T constituted a QSP because it did not satisfy the requirements for a tax-deferred stock acquisition. Specifically, a Type B reorganization permits only voting stock to be used as consideration for T stock, and a reverse triangular reorganization permits only up to 20% non-voting stock of P as consideration. However, because the merger of T into P was part of the same plan as the stock acquisition, step transaction principles, if they applied, would treat the overall transaction as a “Type A” asset acquisition. The Service so ruled, citing Revenue Ruling 67-274 (a pre-section 338 ruling). In that ruling, the Service treated P’s acquisition of the T stock for P voting stock, followed by the pre-planned liquidation of T into P, as a Type C reorganization because that was the overall effect of the two transactions. In Revenue Ruling 2001-46, the Service also ruled, citing Revenue Ruling 90-95, that step transaction principles should similarly apply to evaluate the overall transaction and concluded that the QSP requirement had not been satisfied under the principles of Revenue Ruling 67-274.

3. Revenue Rulings 2004-83 and 77-427—The Dog That Didn’t Bark

Revenue Ruling 2004-83 implicitly refined the analysis and, I will argue, confirmed that the reach of the policy of section 338 extends beyond the case in which P is eligible to make a section 338 election. In Situation 2 of the ruling, P owned all the stock of S and T but the corporations were not members of a consolidated group. Pursuant to a plan, P sold all of the T stock to S for cash, and then T liquidated into S. If treated as two separate transactions, the first step would be subject to section 304 and the second would be a section 332 liquidation. Section 304 treats certain sales of stock between parties under common control as redemptions that may be characterized as distributions under section 301. If the two steps were integrated, they would constitute an all-cash acquisitive Type D reorganization not subject to section 304. In ruling that the transaction was a Type D reorganization, the Service confirmed the approach of Revenue Ruling 2001-46 that step transaction principles apply in making the threshold determination whether a transaction is subject to section 338. This conclusion was a predicate to the holding that in the case of a Type D overlap with section 304, Type D would prevail.

For present purposes, the primary relevance of the ruling does not lie in that holding but in a seeming aside: “Authorities that reject the application of the step transaction doctrine based on the policy of § 338 . . . are not relevant in this case because there is no purchase of stock within the meaning of § 338(h)(3)(A) and § 1.338-3(b).” What is notable about this language is that it does not say simply that section 338 does not apply, a point arguably obvious enough from the face of the statute to have gone without mention; rather it states that the policy of section 338 is not relevant, suggesting that that policy extends beyond the situation in which the stock acquisition is a QSP. The ruling also does not cite to the definition of a QSP but to the definition of the term “purchase” set forth in section 338. In other words, there may be purchases that fall within the policy of section 338 even though they are not QSPs. In that case, step transaction principles might be turned off even though no section 338 election would be available.

At least two circumstances support this analysis. The first is akin to the case of “the dog that didn’t bark.” Revenue Ruling 2004-83 makes no mention of another ruling, Revenue Ruling 77-427, that is closely similar on its facts but comes to an apparently contrary conclusion. In that ruling, the T shareholders sold the T stock to P, which was under common control with T for the purposes of section 304 but not for the purposes of section 368(a)(1)(D) so that the integrated transaction was not a reorganization. P then caused T to liquidate into P. The question was whether the acquisition of the T stock was eligible for asset purchase treatment under the then-in-force statutory override of K-D, former section 334(b)(2), or instead was subject to section 304. The ruling holds that section 304, and not section 334(b)(2), applied because section 334(b)(2) applied to purchase acquisitions, which were defined (similarly as in section 338) not to include carryover or exchanged basis transactions. Section 304 provided at the time that the sale of stock in a transaction described in section 304(a)(1) was treated as a contribution of the stock to the capital of the acquiring corporation—a carryover-basis transaction. Thus, the Service declined to apply the step transaction doctrine to treat the stock purchase as an asset purchase.

The import of this holding was in the conclusion that the carryover basis rule of section 304(a)(1) took the transaction out of former section 334(b)(2). On its face, the reasoning is circular, for if the Service had applied the step transaction doctrine, the transactions would have constituted a taxable stock purchase, which in turn (under former section 334(b)(2)) would have been treated as an asset purchase so that section 304 did not apply. To avoid the charge of circularity, one is forced to the conclusion that the Service was distinguishing between the policy of old section 334 and the circumstances in which that section would actually treat a stock purchase as an asset purchase. This policy is very similar to what is hinted at in the Service’s interpretation of section 338 in Revenue Ruling 2004-83 and is stated directly in the legislative history to section 338: The provision is intended to replace any stock purchase qua asset purchase under K-D in which P takes a cost basis in the T assets.

Under this interpretation, it appears that there are two parts to the application of the statutory replacement of K-D (be it old section 334(b)(2) or current section 338). The first examines whether, under K-D step transaction principles, the acquisition of the stock is, or is part of, a transaction in which P’s basis in the T assets is cost assuming no special override to K-D applies. The reason this is the threshold inquiry is that it is in fact the K-D case, and the statute is an override to K-D principles in that (taxable) case. Thus, the statute will apply if K-D otherwise would apply so as to cause a taxable asset purchase.

If it would, then the second part applies. Under the second part, the transactions are not stepped together under K-D for the simple reason that the statutory override is an override of K-D in the taxable asset acquisition case. Instead, the steps are analyzed as steps and then treated as an asset purchase only if the specific criteria for so treating them are satisfied. That is, step transaction principles do not apply unless the narrower, explicitly specified circumstances of the statute under which the stock purchase is treated as an asset purchase are present. Under old section 334(b), those narrower circumstances included a timely liquidation of T; under section 338, those narrower circumstances generally include an explicit election under section 338(g) (although, as noted, in some cases the election is deemed to occur). More succinctly, if the transaction taken as a whole would be a taxable transaction by reason of K-D principles, then it is analyzed in steps and K-D will not apply—any stepping together depends on whether the steps satisfy the replacement for K-D. If the transaction taken as a whole would not be a taxable transaction, section 338 does not apply to override K-D principles.

When the Service published Revenue Ruling 77-427, the statutory override was not an election regime but an objective test according to which the stock purchase would be stepped together with the liquidation and treated as an asset purchase only if certain criteria were satisfied, including that the stock acquisition, taken by itself, was a “purchase.” In the ruling, step one was satisfied in that, under step transaction principles, P would have taken a cost basis in the T assets under K-D. Consequently, the override of K-D applied and the steps were analyzed separately. Under that separate analysis, the stock acquisition was a section 304 transaction. A section 304 transaction produced a carryover basis to P in the T stock, not a cost basis. Hence, the stock-purchase-as-asset-purchase treatment of old section 334(b)(2) did not apply under the express limitations of old section 334(b)(2). Therefore, the transaction was not treated as an asset purchase.

By both stating in Revenue Ruling 2004-83 that the policy of section 338 did not apply and failing to mention Revenue Ruling 77-427 in that ruling (i.e., not barking), the Service appears to have tacitly endorsed the approach taken in Revenue Ruling 77-427. Specifically, because the transaction taken as a whole in Revenue Ruling 2004-83 was a reorganization, the policy of section 338 (and not merely its rules with respect to QSPs) did not apply—that is, the first step was not satisfied. Again, that policy is to turn off step transaction principles when K-D would otherwise apply to result in a cost basis in the T assets, and to replace the K-D result with a stock-purchase-as-asset-purchase result only in the more limited circumstances in which the requirement of a QSP is satisfied and an election is made. Because the transaction was a reorganization and P would not otherwise take a cost basis in the T assets, the policy of section 338 was not implicated. This situation is distinguishable from that in Revenue Ruling 77-427, in which the policy of the counterpart to section 338 under prior law (old section 334(b)(2)) did apply and in which, under that policy, the transaction was not stepped together, for the reasons discussed above. Further, qualification for K-D treatment (per old section 334(b)(2)) was not met because the stock purchase, when separated from the liquidation, was a section 304 transaction in which basis was not cost.

This analysis also explains why Revenue Ruling 2004-83 did not cite Revenue Ruling 77-427, despite its factual similarity and despite the fact that the later ruling could have been taken to overrule the earlier one. Revenue Ruling 77-427 simply was not relevant to the issue that Revenue Ruling 2004-83 addresses. Thus, Revenue Ruling 2004-83 did not reverse or for that matter affirm Revenue Ruling 77-427, except to affirm indirectly that its general approach was still in force; Revenue Ruling 2004-83 simply deals with a separate question.

The second circumstance supporting the two-stage analysis of section 338 (and of former section 334(b)(2) as well) is the obvious policy difficulties that a more literal reading of section 338 produces. If the override of step transaction principles is confined to those situations in which a QSP has occurred—that is, if there is no daylight between application of the “policy” of section 338 and the circumstances that trigger a QSP—the application of the rule becomes mechanical and easily circumvented. For example, P can acquire just under 80% by vote and value of T, or P can acquire the requisite stock but during a period lasting slightly more than one year; other avoidance tactics are also possible. In these situations, the mechanical application of the K-D override would fail to effectuate the policy of making a QSP the only means by which a stock purchase is treated as an asset purchase that produces a cost basis to P. By instead retaining the substance-over-form approach that underlies the policy of section 338 (step 1) in the broader set of cases in which P would take a cost basis in the T assets under step transaction principles, the difficulties that were present under prior law are mostly addressed. Thus, the acquisition of 79% of T during the acquisition period is simply a stock acquisition, even if T is later liquidated, regardless of whether the liquidation would otherwise be stepped together with the stock purchase under K-D principles. The only difference from a QSP would be that the election is unavailable.

4. Revenue Ruling 2008-25—Circular (Non-)Application of Step Transaction Principles

The Service rounded out its analysis in Revenue Ruling 2008-25, which in a sense presents the mirror image of the situation addressed in Revenue Ruling 2001-46. P acquired all of the T stock in a reverse subsidiary merger in which T’s sole shareholder received consideration amounting to $90x of P stock and $10x of cash. P then caused T to liquidate into P as part of the same transaction. In the liquidation, P assumed $50x of T’s liabilities. Under the same step transaction principles that the Service had applied in the earlier rulings, the overall transaction constituted an asset acquisition. However, the acquisition did not qualify as a reorganization. Specifically, it was not a Type A reorganization because T did not merge into P; it was not a Type C reorganization because P’s assumption of T’s liabilities, together with the cash consideration P provided in the acquisition of the T stock, exceeded the amount permitted under the boot relaxation rule of section 368(a)(2)(C); and it was not an acquisitive Type D reorganization because the former T shareholders did not control P after T liquidated.

In what has been described as “a feat of Orwellian Doublespeak,” the Service concluded that because the overall transaction was not a reorganization, the acquisition of T constituted a QSP—this despite the fact that the stock acquisition, taken by itself, was plainly a reorganization described in sections 368(a)(1)(A) and (a)(2)(E) and that it was only by virtue of applying (now turned-off) step transaction principles that it was not so described. The conundrum results from the fact that invoking step transaction principles seems to cause the stock acquisition to be a QSP, but that its being a QSP means that step transaction principles do not apply and, as a consequence, the stock acquisition seems to be a reorganization. In short, it seems that if the transaction is a QSP, then it is not a QSP.

Whatever the merits of this analysis as a logical matter, it also seems that the Service drew the right conclusion on policy grounds. Indeed, the holding is consistent with the Service’s approach in the rulings just discussed. The threshold question to the application of section 338 is whether the transaction would otherwise fall under the K-D doctrine resulting in a cost basis in the T assets. Since section 338 adjusts step transaction principles only when such a transaction has occurred, the enactment of the provision itself did nothing to the usual method of analysis, just as in the rulings previously discussed. Under that method, the transaction in Revenue Ruling 2008-25 is not a good nonrecognition transaction (whether tested under the reorganization provisions or under section 351). Because it does not merit nonrecognition treatment, the individual steps must be analyzed to determine whether it will be treated as an asset acquisition or a stock acquisition. Section 338 forecloses the former unless there is a QSP and the section 338 election is made (or deemed made). The very point of enacting section 338 was to remove substance-over-form considerations if the acquisition was a taxable purchase of T—whether in substance of its stock or of its assets.

One can equally reformulate the rule of Revenue Ruling 2008-25 as providing that nonrecognition principles must be turned off once the policy of section 338 is implicated. Again, that policy is implicated only when the transaction under normal tax principles would result in a cost basis to P (or its affiliate) in the T assets—that is, when nonrecognition rules would not apply to the transaction. The policy of turning off nonrecognition principles in such a situation makes perfect sense because it is only the artificially imposed requirements of section 338 that would turn an otherwise taxable transaction into a nonrecognition transaction, as was the case in Revenue Ruling 2008-25. Because the purpose of section 338 is to apply precisely when the nonrecognition principles do not, allowing the nonrecognition principles to operate only because section 338 applies would be inappropriate.

5. Revenue Ruling 77-427 Revisited—A Loose End

All of this raises a final question. If Revenue Ruling 2008-25 confirms that the policy of section 338 requires turning off both nonrecognition and step transaction principles, why did that rule not apply in Revenue Ruling 77-427 to turn off the deemed nontaxable contribution to capital of former section 304, thereby making the stock purchase qualify under former section 334(b)(2) after all? Recall that the Service concluded that section 334(b)(2) did not apply because the stock acquisition constituted a deemed nonrecognition transaction under former section 304(a)(1).

To sharpen the point, it may be useful to reconstruct more explicitly what I have argued was the Service’s chain of reasoning in Revenue Ruling 77-427:

1. The transaction implicated the policy of former section 334(b)(2) because, under step transaction principles, P would have taken a cost basis in the T assets under K-D;
2. because the transaction implicated the policy of former section 334(b)(2), both step transaction and nonrecognition principles were turned off for the purpose of analyzing the stock acquisition;
3. the stock acquisition, taken by itself, was described in section 304 because the control requirement of section 304(c) was satisfied;
4. under section 304(a)(1) (as then in effect), P was treated as acquiring the T stock in a nontaxable contribution to capital; and
5. because the stock purchase taken in isolation was a nonrecognition transaction (per section 304(a)(1)), P did not acquire the stock in a “purchase” transaction and therefore could not obtain a cost basis in the T assets under former section 334(b)(2) by timely liquidating T.

If the policy of a statutory override of K-D requires turning off both step transaction principles and nonrecognition rules, as Revenue Ruling 2008-25 seems to suggest (step (2) above), why didn’t the Service disregard the nonrecognition treatment stipulated in former section 304 (step (5) above) for the purpose of analyzing the stock purchase (taken on its own) in Revenue Ruling 77-427? Had the Service done that, the stock acquisition would have qualified as a cost-basis transaction after all, so that the timely liquidation of T would have resulted in a cost basis in the T assets under former section 334(b)(2).

Perhaps the most obvious answer is that the Service did not fully appreciate the implications of section 338 until 40 years later. But while that may be true, it seems that the Service nevertheless came to the right conclusion in Revenue Ruling 77-427, which is to say that the articulation of the principle of Revenue Ruling 2008-25 requires some refinement.

As expressed above, the analysis is guided by the specific purpose of the section 338 regime to characterize certain transactions as taxable. The deemed capital contribution rule of former section 304, however, did not implement a tax deferral regime. Instead, it formed part of a larger regime (section 304 generally) that recharacterizes a transaction that is a taxable stock acquisition in form as a taxable redemption in substance. Consistent with that approach, Congress specified that a stock purchase that occurs in connection with the redemption, because it is formal, will be recharacterized according to its substance. A shareholder’s sale of property to its controlled corporation more closely resembles a capital contribution than it does a genuine sale.

This situation differs from that in Revenue Ruling 2008-25, in which the Service concluded that it would be unreasonable to give effect to reorganization principles that would apply only because an underlying policy that they should not apply was in effect (namely, that the transaction was taxable under K-D principles and therefore subject to section 338). In short, the policy rationales in the two rulings pointed in opposite directions. Giving effect to the carryover basis rule of former section 304(a)(1) in Revenue Ruling 77-427 was consistent with the concern that motivated that provision in the first place (itself a substance-over-form provision). Giving effect to the reorganization provisions in Revenue Ruling 2008-25 would have been at odds with the concern that motivated section 338—to treat what was in substance a cost-basis asset purchase as a taxable stock acquisition for tax purposes.

Therefore, it makes sense to characterize the nonrecognition principle of Revenue Ruling 2008-25 more precisely as follows: If the application of step transaction principles at the first stage of analysis yields the conclusion that the transaction taken as a whole is taxable, then any provision that would make the transaction in which the stock purchase took place nontaxable will not apply because doing so would defeat the purpose of applying a K-D override to cost-basis asset acquisitions in the first place. This principle would not apply to Revenue Ruling 77-427 because the transaction in which the (nominal) stock purchase occurred was taxable under former section 304. It was a taxable redemption of P stock under former section 304.

C. Conclusion

It is sometimes said that section 338 turns off the step transaction doctrine with respect to P in the case of a QSP. The Service’s rulings in the area, coupled with the legislative history of section 338, indicate it is probably more accurate to say that section 338 turns off the step transaction doctrine when the policy of section 338 is implicated—a principle that embraces a somewhat broader set of situations than those involving a QSP. That policy is that where a series of transactions in which, under K-D principles, P purchases T stock and, because of the subsequent integrated transactions, would end up owning T’s assets with a cost basis, the transactions are not automatically treated as an integrated series of transactions with respect to P. Instead, the stock acquisition and whatever follows are analyzed separately, and if, but only if, the transaction constitutes a QSP, P may elect under section 338(g) (or in some cases, may jointly elect with the controlling T shareholder under section 338(h)(10)) to treat the stock purchase as an asset purchase. Otherwise, the acquisition is treated as a stock purchase.

IV. The Problem of Regulation Section 1.338(h)(10)-1(c)(2)

The just-proposed summation of the effect and operation of section 338 appears to harmonize both the legislative history of the section and the Service authorities. It explains each of the post-section 338 enactment revenue rulings examined and is also consistent with the pre-section 338 rulings examined.

We now arrive, however, at the second question raised in the Introduction: whether, in light of the principles just explained, Treasury has authority to turn off the step transaction doctrine when the overall transaction of which the stock acquisition is a part constitutes a reorganization. Treasury has invoked this authority in one important instance, Regulation section 1.338(h)(10)-1(c)(2) ((h)(10) reorganization override). The (h)(10) reorganization override permits an election under section 338(h)(10) if P’s acquisition of T, when taken in isolation, would be a QSP regardless of whether it would be a step in a reorganization under step transaction principles. In what follows, I argue that Treasury has exceeded its authority in adopting the (h)(10) reorganization override.

A. Background

At the time of the enactment of section 338 in 1982, an election under section 338(g) was much more consequential for most purchasers than it has been since 1986. Before 1986, the General Utilities doctrine applied to exclude most of the gain realized by a corporation in a non-subsidiary liquidation. Further, former section 337 typically eliminated tax on most of the gain or loss realized on a liquidating corporation’s sale of its assets to a third party prior to and in connection with such a liquidation. Thus, before former section 334(b)(2) was enacted, it mattered a great deal whether the acquisition of the T assets was treated as part of a liquidation of T. If P bought the assets from T in connection with the subsequent liquidation of T (or if P bought the assets from the T shareholders following the liquidation of T), there would be just one level of tax—on the BIG/L that the T shareholders realized in their T stock. If P bought the T stock and then caused T to liquidate in what counted for tax purposes as a separate transaction (in other words, assuming K-D did not apply), there would be two levels: one on the T shareholders’ BIG/L in their T stock (just as under the pre-liquidation asset sale), and a second level, albeit deferred under the subsidiary liquidation rules, for P on the ultimate disposition of the T assets. Importantly, however, the treatment of the T shareholders was the same in both instances. The shareholders recognized the BIG/L on their stock. The difference between asset-sale and stock-sale-plus-liquidation treatment mattered only for P. Relatedly, the result under old section 334(b)(2), the first statutory replacement of K-D, as well as under the section 338(g) election, was entirely under P’s control (and still is in the case of section 338(g)). If P made the election (or liquidated T under old section 334(b)(2)), T would be treated as selling all of its assets under the rule of old section 337, thereby eliminating corporate-level tax that it would otherwise bear (as a subsidiary of P) if the transaction were structured as a stock purchase followed by a non-integrated liquidation. If P did not make the election, both levels of BIG/L were ultimately recognized.

With the repeal of the General Utilities doctrine (GU repeal) as part of the Tax Reform Act of 1986, the significance of the section 338(g) election largely disappeared; except in the case of a T that was a controlled subsidiary of the seller, the main difference was the deferral on T-level gain or loss when P did not make the election. In 1986, Congress also repealed old section 337, so that two levels of tax still apply even if the transaction is structured as an asset purchase from T. Relatedly, it was no longer possible to achieve one level of tax by having P purchase the T assets rather than the T stock.

Things have always been somewhat different with respect to section 338(h)(10). Before GU repeal, the difference from the election under section 338(g) involved the level of gain or loss that was taxed (T assets rather than T stock); after, the difference is that the section 338(g) election generally provides no relief from tax, while the section 338(h)(10) election continues to eliminate tax on the T stock. Because the section 338(h)(10) election is available only when a pre-sale liquidation of T followed by a sale of the T assets would generate the same basic result, it operates in an analogous fashion to section 338(g) in the pre-GU repeal world: It permits the parties to achieve one level of tax without requiring that the form of the transaction be a liquidation of T followed by sale of the T assets to P by the controlling T shareholder.

The significance of the section 338(h)(10) election is partly accidental and largely the result of GU repeal. The problem that motivated its enactment was narrow and technical. Even before GU repeal, some items of corporate-level gain were taxed to T on its liquidation (or on the sale of its assets in connection with its liquidation) when it was not a controlled subsidiary. These included LIFO inventory recapture and, in the case of a pre-liquidation sale of the T assets, other gain or loss realized on T’s inventory. In a situation in which T had been a (non-parent) member of a consolidated group, it had been unclear whether these various items realized at the T level would be treated as arising while T was still a member of the selling group if the election under section 338(g) were made. The Technical Corrections Act of 1982 added what is now section 338(h)(10) to clarify the treatment of T in this case. Under the amendment, the items would not be treated as arising when T was still held by the seller if no (h)(10) election were made in connection with the QSP. This would mean that the same rules would apply as in any other QSP. If, however, the (h)(10) election were made (which must be done by both P and the controlling seller), then the normal rules for liquidating subsidiaries (which generally continue under current law) would apply: T would be treated as making a fully taxable disposition of its assets that would be taxed under the consolidated return regulations, but the controlling T shareholder would recognize no gain or loss on its sale of the T stock.

The overall effect (before GU repeal) for Ts that were members of a consolidated group was to make largely elective the level at which gain or loss would be recognized. If P made the regular 338(g) election, the consolidated seller and the non-consolidated seller were treated similarly. The seller would recognize gain or loss on the T stock, but T would recognize only the limited items of gain or loss realized that were required to be recognized under old section 337—essentially, gain or loss on certain inventory items. Further, the (generally limited) associated tax liability on the deemed sale of the T assets would in effect be P’s (as the owner of T at the time of the deemed sale of T’s assets). If, instead, P and the seller made a joint section 338(h)(10) election, the same treatment that applies in a regular section 332 liquidation would apply to the deemed liquidation of T, which occurred on the seller’s watch. In this case, the BIG/L on the T stock escaped tax, but the seller paid tax on the deemed sale of the T assets.

B. Policy of the Section 338(h)(10) Reorganization Override

The Tax Reform Act of 1986 (TRA) expanded the availability of the (h)(10) election, pursuant to regulations, to the case in which T was a member of an affiliated group of corporations that did not file a consolidated return. TRA also indirectly made the (h)(10) election more consequential, because it now became the only avenue by which a stock sale followed by a liquidation would generate one level of tax. From a practical perspective, its value lay in permitting the parties to achieve that result (when T was a controlled subsidiary of the seller) without requiring that the seller first liquidate T and then sell the T assets to P. Since that time, Treasury has further extended the (h)(10) election to selling S corporation shareholders.

Neither the legislative history of section 338(h)(10) nor the administrative history to Regulation section 1.338(h)(10)-1 offers any specific discussion of the basic rationale for the (h)(10) election beyond that identified above. Perhaps the most that can be inferred is that Congress recognized that the tax benefit available if T were liquidated pre-sale rather than post-sale would result in economically undesirable liquidations, and, much as it had with old section 337, Congress made the tax benefit available without the requirement to liquidate T in the seller’s hands. Beyond providing additional flexibility to taxpayers (as all elections do), there appears to be no clear policy underlying the (h)(10) election.

C. Authority for the Section 338(h)(10) Reorganization Override

Two features of the (h)(10) reorganization override merit attention. The first is that it permits the parties to make the (h)(10) election even when, under step transaction principles, the overall transaction qualifies as a reorganization. The second is that Treasury does not consider its authorization to extend to the case in which both the stock acquisition on its own and the overall transaction qualify as a reorganization. These two features of the rule suggest an interpretation of section 338 that is inconsistent with both the Service’s approach in its own guidance as explicated above and section 338’s legislative history. Specifically, Treasury appears to hold the view that section 338(h)(10) applies as a threshold matter to individualize the steps of an integrated transaction, regardless of whether the policy of section 338 is relevant. Nevertheless, if those steps do not constitute a taxable acquisition (either individually or jointly in the case of integrated transactions), Treasury evidently considers the policy of section 338 not relevant, and the election is unavailable. In short, it is not so much that Treasury considers the availability of the (h)(10) election to be a matter purely of policy; rather it appears that Treasury considers something about the (h)(10) setting to expand the scope of the policy of section 338 but only to a limited degree.

It remains entirely unclear what that something would be. As detailed in Part II, the policy of section 338 is to replace the K-D analysis of an integrated, taxable stock acquisition leading to a cost basis in the T assets with a regime that requires stock purchase treatment unless the specific requirements of section 338 are satisfied (QSP, timely election and so forth). Practically, this means beginning with step transaction principles for the purpose of determining whether in effect a taxable asset acquisition has occurred, and only then turning them off when it has. By contrast, under the (h)(10) reorganization override, the focus begins with the stock acquisition viewed in isolation, presumably on the basis that an otherwise taxable step brings the policy of section 338 into play. The problem with this approach is that the notion of “an otherwise taxable step” is not particularly coherent. The tax law is predicated on application of substance-over-form principles (among other doctrines) to determine whether a “step” is or is not taxable in the first place. We cannot know whether a step is taxable without knowing what that step is, and that requires, in turn, the application of relevant tax doctrines. For this reason, a focus ab initio on individual steps does not appear to have any meaning at all. Assuming that Congress has not overridden the usual mode of analysis, the individual steps in isolation are a proper unit of analysis only if the economic substance of the overall transaction warrants a focus on them taken in isolation. To be sure, Congress has overridden that principle when the problems associated with an integrated approach create the K-D issue that Congress decided to address under section 338. But the (h)(10) reorganization override extends beyond that situation with no obvious authorization. In effect, it assumes that K-D principles are partly turned off for the purpose of determining whether they are fully turned off.

Commentators generally have not questioned Treasury’s authority to regulate the application of judicially created doctrines in the section 338 area even though it is not obvious that Treasury has that authority. For example, the Tax Section of the New York State Bar Association (the NYSBA Tax Section) argues on policy grounds that Treasury retain the regime it has currently adopted for purposes of section 338(h)(10). Similarly, the ABA Tax Section endorses Treasury’s use of a broad authority to suspend step transactions principles on policy grounds. More generally, a search of the commentary on section 338(h)(10) uncovered no suggestion that the (h)(10) reorganization override lies outside of Treasury’s authority. Yet the assumption that Treasury has acted within its authority seems inconsistent with the requirement that Treasury follow both Congressional intent and, in the absence of a statutory override, substantive legal doctrines, which include judicially created doctrines such as the step transaction doctrine and the business purpose requirement.

In what follows, I argue that no such override exists. Even the expansive deference granted to agencies under the so-called Chevron doctrine does not extend to an agency’s policy choices that cannot be tethered to the statute in some meaningful way. Further, the (h)(10) reorganization override conflicts with the reorganization provisions, which do not make reorganization treatment elective; if a transaction is described in section 368(a)(1), it is a reorganization. Nothing in the legislative history of section 368 suggests a Congressional intent to turn off step transaction principles in the analysis of multi-step potential reorganizations, and nothing in the policy of section 338 would suggest an extension to Treasury of authority on this point. Recall that section 338’s policy relates specifically to taxable acquisitions, not reorganizations, as both its legislative history (and that of its predecessor) and the Service’s own rulings make amply clear. That is why authorities begin the analysis with step transaction principles in effect to determine whether section 338 is in play.

Consider K-D itself. The government invoked a step transaction theory on the basis that the theory reflected the economic substance of the stock purchase together with the pre-planned liquidation, and the court agreed, citing Commissioner v. Court Holding Company. That case, along with countless others, requires courts to look to the economic substance of a transaction or arrangement to determine its tax consequences. Indeed, the proposition that the operation of step transaction principles is not optional can be inferred from the facts that taxpayers were able to invoke it successfully thereafter, as well as from Congress’s having concluded that only a statutory override to K-D would turn off step transaction principles. The K-D court itself made this point.

One might argue that there are other bases on which Treasury has authority to turn off step transaction principles. For example, it may exist as a matter of the general delegation under section 7805 or, in the case of the K-D rule, by virtue of the express delegation under section 338(i) to Treasury to prescribe regulations “as may be necessary or appropriate to carry out the purposes of this section.” But as argued below, neither these provisions nor any equitable doctrine of taxation would seem to extend as far as a delegation to “adopt good tax policy,” which effectively is the support needed for the (h)(10) reorganization override. Even then that authority would have to override the rules otherwise applicable to reorganizations. Moreover, even if some such authority were located, Treasury would have difficulty explaining why that authority should be limited to situations in which the stock purchase taken on its own is taxable.

1. Possible Authority Under Section 338

Section 338(i)(1) authorizes Treasury to prescribe regulations to ensure asset and stock purchase consistency. The legislative history to section 338 makes clear that the concern on this issue related to the selective use under prior law of the features of taxable stock and asset purchases when the overall substance of an acquisition was an asset purchase; substantively, the language also predates the addition of section 338(h)(10), suggesting that it has no specific application to that provision. As an example of the basic worry, P might purchase high-basis assets of T and T stock in transactions that intentionally did not qualify for K-D treatment. Under prior law, P would obtain the benefit of a high overall stock-or-asset basis in what was in substance an asset acquisition. Although the substantive rules of section 338 require asset and stock consistency, Congress evidently felt the need to backstop those rules with regulatory authority. This problem, however, has nothing to do with the (h)(10) reorganization override, which focuses on the parties’ joint ability to override step transaction principles in the reorganization setting to treat the transaction as a taxable asset purchase.

2. General Authority to Implement the Policies of Section 338

Another possibility is that Treasury has inherent or otherwise generally delegated authority to adopt the (h)(10) reorganization override. The rule was initially promulgated in temporary regulations that were finalized three years later without change. The preambles to neither the temporary nor the final regulations articulate any rationale for the rule. Rather they merely note that the Service had requested comments on the proposed rule in Revenue Ruling 2001-46 and that all of the comments received were (unsurprisingly) in favor. Revenue Ruling 2001-46 likewise contains no discussion of Treasury’s authority to adopt the rule. In light of the fact that Treasury concluded in that ruling that step transaction principles generally do apply to would-be QSPs that form part of reorganizations, it appears that Treasury simply assumed it had the authority and the only question was whether the adoption of the rule would be sound policy.

Section 7805(a) provides Treasury with authority to prescribe “all needful rules and regulations for the enforcement of [Title 26].” In addition, under so-called Chevron deference, courts will defer to reasonable agency interpretations of statutes that they administer when the statute is ambiguous or contains a gap. Under either of these dispensations, Treasury has concededly broad authority to promulgate regulations that effectuate provisions of the Internal Revenue Code.

As noted earlier, the problems that gave rise to section 338 arose from the undue flexibility that stock purchases provided to acquirers under K-D and its statutory successor, former section 334(b)(2). According to the legislative history to section 338, the Congressional intent was to replace any non-statutory treatment of a stock purchase as an asset purchase with an elective regime. Apart from the (h)(10) reorganization override, Treasury, the Service and the courts have fairly read this language, especially in light of the definition of “purchase” in section 338(h)(3), to apply to taxable acquisitions of T stock, not to all such acquisitions. That reading is appropriate in light of the problem that K-D created (especially in a pre-GU repeal world) for taxable acquisitions of T and that persist when T is a controlled subsidiary of the seller. The problem does not arise when T is acquired in a nontaxable transaction. That, indeed, is why the threshold inquiry of whether a QSP has occurred applies step transaction principles.

There are a number of instances in which Treasury under its general regulatory authority has, or may, set aside judicial doctrines or otherwise exercise expansive power that might initially appear consistent with an authority to promulgate the (h)(10) reorganization override. On closer inspection, however, these instances do not support Treasury’s authority to promulgate the (h)(10) reorganization override. This Part discusses three cases: the liberalization of the “remote continuity” rules under Regulation section 1.368-2(k); the possible exemption of Type D reorganizations from the continuity of interest (COI) requirement that generally applies to reorganizations; and the adoption of the entity classification rules under which an eligible business entity can choose its classification for federal tax purposes.

a. Liberalization of the Remote Continuity Doctrine. The early history of the reorganization provisions featured judicial development of the “remote continuity,” or Groman/Bashford, doctrine. Under Groman/Bashford, a transaction otherwise qualifying as a reorganization would fail to qualify if P transferred the T assets or stock to another entity (typically another corporation) in connection with the putative reorganization, even if that entity was under P’s direct or indirect control. The basis for the disqualification was that the T shareholders’ interest in what were formerly T-owned assets was too “remote” from their ownership in P (or the corporation that controlled P when the T shareholders had received equity in that corporation) for the transaction to reflect a continuing interest in the same assets in corporate form.

Over time, Congress liberalized the remote continuity doctrine, most notably with section 368(a)(2)(C), which permits P to drop the T stock or assets to a controlled corporation without causing an otherwise valid Type A, B, or C (as well as certain Type G reorganizations) to fail to qualify. Notably, however, section 368(a)(2)(C) speaks only of a one-tier drop-down, and it speaks not at all about post-acquisition distributions by P or a controlled subsidiary of P. The other relaxations of remote continuity have been still more targeted.

Despite the limited statutory relaxation of remote continuity, Treasury issued Proposed Regulation section 1.368-2(k) in 1997, which it finalized in 1998 (the -2(k) regulation). The -2(k) regulation permits essentially unlimited dropdowns of T assets or stock to controlled entities of P following an otherwise qualifying Type A, B, or C reorganization. It also permits distributions of T assets or stock in connection with such reorganizations as long as the assets or stock remain within the same affiliated group as P and certain other conditions are satisfied.

Superficially, the -2(k) regulation would seem to offer a sound precedent for the (h)(10) reorganization override. Like the (h)(10) reorganization override, it abrogates a judicially created doctrine, and like the (h)(10) reorganization override, it does so without any specific Congressional grant of authority or Supreme Court ruling invalidating the remote continuity doctrine. Nevertheless, if anything, the circumstances of its adoption actually highlight deficiencies in the (h)(10) reorganization override rather than provide support for it. Notably, Treasury was careful in its explanation for the -2(k) regulation to recount the back-and-forth between Treasury, the Service, and courts on one hand, and Congress on the other, on remote continuity. In the case of reorganizations to which the -2(k) regulation applies, that history uniformly featured Congressional relaxation of the doctrine in the face of judicial or administrative extensions of it. Conversely, it also featured Congressional silence in the face of administrative relaxation of the remote continuity requirement. Thus, a fair reading of the history was, in general, growing Congressional dissatisfaction with the doctrine, as evidenced both by Congressional pushback on its extension to novel factual situations and the general legislative trend to relax the requirements for remote continuity.

The other distinguishing feature goes to the substance of the -2(k) rule. Although Congress’s decision to make incremental changes to the remote continuity doctrine could be taken as evidence that it did not consider appropriate the kind of sweeping adjustment that the -2(k) regulation adopted, the better inference is the opposite. Given the lack of sound policy basis for distinguishing between single-tier dropdowns on one hand and the movement of T assets or stock within P’s affiliated group on the other, it is hard to see why a policy that permits movement of the T stock or assets one tier does not extend to moving the T assets anywhere within P’s controlled group. From an economic perspective there is no real difference between the two. Further, the practice of moving assets within a larger controlled group is broadly consistent with the development of the corporate enterprise over the course of the last 70 years or so, under which parent corporations commonly operate numerous subsidiaries at numerous levels in ways that do nothing to change the business control of the parent over the T assets or stock following an acquisition.

The (h)(10) reorganization override has no analogous policy justification. Nor does it appear to reflect a response to evolving business practice that would justify its use as a means to safeguard existing policy under changed conditions. This would be the case even if the (h)(10) reorganization override permitted the (h)(10) election without regard to whether the stock acquisition was a QSP—arguably a more consistent application of policy principles than the actual (h)(10) reorganization override. However, because the (h)(10) reorganization override does not rest on a policy basis, as evidenced by the fact that it does not permit the election if the stock acquisition qualifies as a reorganization, any putative justification of the rule on policy grounds seems particularly weak.

b. Possible COI Exemption for Type D Reorganizations. A second possible precedent for the (h)(10) reorganization override could be Treasury’s decision to limit the COI requirement in certain settings. The courts developed the COI doctrine to ensure that transactions meeting the technical definition of a reorganization did not qualify for reorganization status when the overall effect of the transaction was for the T shareholders to cash out their investment in T. The clearest example of the problem is a Type A merger in which the T shareholders receive little or no equity in P (or in a corporation that controls P). Such a transaction satisfies the statutory definition of a Type A reorganization, but if permitted to qualify would exempt T from tax on an exchange that is indistinguishable from a taxable sale. To avoid this result, the COI doctrine, which has long been administratively codified, requires that “a substantial part of the value of the proprietary interests in the target corporation be preserved in the reorganization.”

Consensus appears to exist that Treasury can waive the COI requirement in acquisitive Type D reorganizations in certain circumstances. In an acquisitive Type D reorganization, T transfers substantially all of its assets to P and then distributes to its shareholders whatever it receives (or is deemed to receive) from P, together with any remaining assets, in liquidation. The relevant difference from a Type C is that the T shareholders must be in control of P (directly or indirectly) immediately after the exchange. For this purpose, “control” means ownership of at least 50% of P by vote or value, for which constructive ownership rules apply.

Under the usual COI rules, the transaction would fail as a reorganization if the T shareholders did not receive sufficient equity in P. Treasury has generally abandoned that requirement in the acquisitive Type D case because the statutory requirement that T or its shareholders be in control of P (whether directly or indirectly) immediately after the reorganization exchange means that as a practical matter the T shareholders usually have an equity stake in the T assets after the transaction, regardless of whether they receive P stock as a result of the reorganization exchange. Nevertheless, the exchange itself may involve no formal transfer of P equity or only a very small transfer of it in comparison to cash or other property. In these cases, the imposition of a COI requirement would have the effect of enabling the parties to choose their tax treatment on transactions that in substance reflect the idea that the T shareholders are preserving their investment in corporate equity.

An all-cash Type D reorganization illustrates the principle. In this transaction, shareholders are in control of T and P in the same proportions. T transfers all of its assets to P for cash and liquidates, distributing the cash ratably to the T shareholders. The exchange qualifies as an acquisitive Type D reorganization even though the former shareholders of T receive literally no equity in the transaction and it otherwise is identical to a simple sale of the T assets. COI need not apply because the T shareholders continue to own a controlling equity stake in P (directly or indirectly) such that the transaction is equivalent to a cash-out of earnings of T and P viewed as a single enterprise, with the same ownership of T’s assets that existed before the exchange continuing after it. A less extreme version of the same transaction occurs where the T shareholders, or some of them, receive a small amount of equity in P and a large quantity of cash or other property in a Type D. In both cases, requiring a formal exchange of equity in P (or in a corporation that controls P) would defeat the basic purpose of treating a merely formal change in ownership of T’s assets (through P rather than T) as a nontaxable transaction. Thus, the suspension of COI in these cases furthers rather than impairs a clearly established policy objective of the reorganization provisions.

No such larger statutory purpose underlies the (h)(10) reorganization override. Both the history of application of K-D and the express statements in the committee reports accompanying the enactment of section 338 make clear that the problems Congress was addressing arise in the taxable transaction setting. The (h)(10) reorganization override by its terms applies to the nontaxable setting. In the absence of statutory language or legislative history expressly overriding step transaction principles beyond the taxable K-D case, there is no basis for concluding that Congress extended K-D principles to reorganizations.

Related to this point is that the difficulties associated with K-D transactions had to do specifically with their taxable features, including, primarily, P’s inconsistency in its treatment of T stock and T asset acquisitions (whether directly or through the use of affiliates). The Senate Report to TEFRA (in which the original version of current section 338 was enacted) provides a number of illustrations. It notes in particular the problem under former section 334(b)(2) of P’s selective acquisition of T assets with a cost basis and a basis determined by reference to the stock purchase price. In this scenario, P, pursuant to a single plan, might purchase the stock of T and the high-basis assets of T affiliates, later liquidating T under former section 334(b)(2) to obtain an asset basis determined with reference to the cost of the T stock. The overall effect is a permanent reduction in total tax paid as compared to either a unitary stock purchase with an integrated liquidation of T or such a purchase without the integrated liquidation. By contrast, if the transactions when stepped together for tax purposes qualify as a reorganization, P takes a carryover basis in all of the T assets. Further, even if P selectively purchased T assets (or assets of a T affiliate) before the tax-free acquisition of the T stock, the overall effect would be deferral of tax, not elimination of tax gain.

A second problem identified in the Senate Report is the opportunity under prior law to have a carryover of T’s tax attributes for up to five years following a stock acquisition even though the acquisition remained eligible for asset purchase treatment. Thus, P might make a stock purchase eligible for the 334(b)(2) election but enjoy the benefit of favorable T tax attributes (such as net operating losses or accelerated depreciation) for an extended period of time before liquidating T. This problem likewise would not materialize in a reorganization or a section 351 transaction, a point implicitly recognized by the Service in Revenue Ruling 67-274 and Revenue Ruling 2001-46.

c. Check-the-Box Rules as Precedent. A final possible precedent for the (h)(10) reorganization override is the choice-of-entity regulations, or so-called “check-the-box” (CTB) rules, which were adopted in 1996. Under the CTB rules, an “eligible business entity” may choose whether it will be classified as an association, or as either a disregarded entity (in the case of an entity having one owner) or a partnership (in the case of an entity having more than one owner) for tax purposes. An association, in turn, is taxed as a corporation. At least one commentator has contended that the CTB rules exceed Treasury’s authority under section 7701(a)(3), which defines “corporation” for purposes of the Internal Revenue Code, but the courts, in a series of cases, have consistently ruled that Treasury acted within its power in promulgating the CTB regime. Given the CTB rules’ dramatic departure from the earlier regime in the absence of any statutory change, one might conclude that the (h)(10) reorganization override likewise falls within Treasury’s broad regulatory authority to implement the internal revenue laws.

The CTB rules replaced the so-called Kintner regulations under which a business entity’s status for federal tax purposes depended on the corporate resemblance test, with partnership status for any entity that did not have at least three of four identified “corporate” attributes. The Kintner regulations were partly motivated by Treasury’s concern that personal service businesses were availing themselves of profit-sharing benefits intended to be available only to traditional corporations; underlying state law typically prohibited personal service businesses from adopting corporate status. The rules also were designed to address the opposite problem involving taxpayers who sought limited liability but pass-through tax treatment. In both situations, the challenge that taxpayers faced was achieving the optimal regulatory framework: corporate tax status for personal service businesses where state law did not permit personal service corporations or, conversely, limited liability and flow-through tax treatment. The latter considerations were especially important given that dividends generally were taxable at ordinary rates before 2003. Contesting the cases in which the Service disagreed with the taxpayer’s position proved time-consuming and costly to the government.

The already high costs of policing the taxable/pass-through line rose with the advent of limited liability companies (LLCs) in the 1980s. LLCs generally enjoy limited liability but are not classified as corporations for state-law purposes. In Revenue Ruling 88-76, the Service ruled that entities established under the first LLC statute, in Wyoming, did not qualify as corporations under the Kintner regulations because they lacked continuity of life and free transferability of interests even though they had centralized management and limited liability. Because the entity had more than one owner, it therefore qualified as a partnership for tax purposes.

Revenue Ruling 88-76 opened the door to the establishment of entities that had significant corporate law features, including limited liability, but enjoyed pass-through tax treatment. Not surprisingly, other states soon followed Wyoming’s lead in adopting LLC statutes, effectively expanding elective classification nationwide, and thereby rendering the satisfaction of the requirements of the Kintner regulations largely an exercise in document drafting rather than the allocation of substantive legal rights. Meanwhile, the enactment in 1987 of section 7704 had lowered to some extent the tax stakes in entity classification anyway. Under section 7704(a), most publicly traded partnerships are classified as corporations for federal tax purposes, regardless of state-law classification. The overall effect of these developments was to make it easy for nonpublicly traded entities to choose their status, but to require that publicly traded ones be taxed as corporations. In short, the tax line between corporations and other entities had in effect shifted but also become clearer: Publicly traded entities were corporations; the status of other entities was a matter of taxpayer choice through the mechanism of entity choice (LLC, corporation, etc.) and minor adjustments to the entity’s state-law characteristics. Under these circumstances, Treasury concluded that it made sense to make the classification rules explicitly elective for business entities not otherwise required to be corporations for federal tax purposes.

Some taxpayers argued that the CTB rules went beyond Treasury’s authority, but it took time for the cases to percolate through the courts (probably because the rules are almost always taxpayer-favorable). To date, at least four cases have been decided, all holding that Treasury did not exceed its authority in adopting the new regime. The cases generally raise some variation on the claim in Littriello v. United States. Littriello was the sole owner of an LLC for which no election to be classified as an association taxable as a corporation had been made. The Service brought an action against Littriello for the LLC’s unpaid employment taxes, and Littriello argued, among other things, that the CTB rules were invalid as an exercise of Treasury’s authority under section 7701 and therefore that the LLC was a corporation for tax purposes under the old Kintner regulations. As a corporation, only the LLC, and not Littriello himself, would bear legal liability for any unpaid employment taxes. Affirming the district court, the court of appeals held that because the definition of “corporation” in section 7701(a)(3) was ambiguous and also said nothing about the status of LLCs, the CTB regime was a reasonable exercise of Treasury’s power under the Chevron doctrine. The court also noted that Treasury had the authority to respond to changing circumstances with new rules.

The CTB cases might be thought to establish an expansive Treasury authority to adopt regulations based largely on its determination of sound policy. If so, the principle might apply to the (h)(10) reorganization override as a similar exercise. But the CTB cases do not warrant such an inference. Rather they frame the CTB rules as an instance of statutory ambiguity and gap-filling, a characterization not reasonably applicable to the (h)(10) reorganization override. In particular, in the CTB cases, the courts noted that the Code contains no definition of an “LLC” and concluded that Treasury was authorized to fill the gap under Chevron, taking into account the policies of the entity classification regime. The (h)(10) reorganization override, however, does not fill a gap. (In point of fact, if anything it creates one inasmuch as the treatment of minority T shareholders becomes unclear if P and T jointly make the election under section 338(h)(10) for a transaction that under step transaction principles qualifies as a reorganization.)

A second difference from the (h)(10) reorganization override is Treasury’s articulation of a policy in support of the CTB rules, noting in particular that the entity classification regime had become largely elective in all but name for nonpublicly traded business entities. Given the costs of policing the old, “formalistic” regime, adoption of an explicit election regime represented a relatively minor substantive change that was consistent with sound administration and changing facts on the ground, one of the rationales that supports the Chevron doctrine. The (h)(10) reorganization override, by contrast, seems generally inconsistent with the underlying statutory scheme of section 338, as well as with the long-settled rule that section 368 is not elective.

3. Conclusion on Treasury’s Authority

The preceding discussion suggests that Treasury lacks authority for the (h)(10) reorganization override, the logic of which is inconsistent with Congress’s approach to the K-D problem in taxable acquisitions, and the basis for which finds no support in the legislative or regulatory history of section 338 or any independent grant of authority.

As a final point, it is notable that Congress has sometimes granted the kind of regulatory power that would authorize the (h)(10) reorganization override in statutes far more explicit than section 338(i) or section 7805(a). Section 1502 is perhaps the most familiar example. It directs Treasury to promulgate regulations covering returns of consolidated groups of corporations and states in pertinent part: “In carrying out the [directive to promulgate such regulations], the Secretary may prescribe rules that are different from the provisions of chapter 1 that would apply if such corporations filed separate returns.” Short of an express statutory fix that would authorize the (h)(10) reorganization override, it appears that statutory language comparable in breadth to that of section 1502 would be required.

V. Conclusion

The Service’s interpretation of section 338’s override of step transaction principles is generally correct. The override applies whenever the policy of section 338 is implicated to supplant K-D principles. That policy is implicated, in turn, when K-D principles would operate to treat a taxable stock purchase as an asset purchase yielding a cost basis to P, which occurs in a broader class of cases than those in which P has made a QSP. Analysis under section 338 therefore is a two-step affair. First is the question whether a series of steps implicates the policy. If, but only if, it does, the stock purchase is analyzed separately from what comes after. If under that separate analysis the stock purchase constitutes a QSP, it is eligible for an election under section 338, which, if made, will cause the stock purchase to be treated as an asset purchase. If the policy is implicated and no election is made (either because the transaction is eligible but the choice is declined or because the transaction is ineligible for the election), the post-stock acquisition transactions are treated separately from the stock acquisition. Further, because nothing in section 338 or its legislative history qualifies the usual rules of construction that apply to determine whether a taxable purchase transaction has occurred, one must apply those rules to the threshold question of whether the policy of section 338 is implicated. Therefore, in determining whether a stock “purchase” has happened, one applies the same step transaction principles that ordinarily apply in other contexts.

Unfortunately, these considerations indicate that Treasury has erred with the (h)(10) reorganization override, which does not conform to the policy of section 338 and has no independent statutory or judicial basis. Rather, it applies a kind of “bootstrapping” principle to section 338, effectively permitting section 338 to apply to the question of whether it applies. Treasury should revoke the rule and, if it believes there are sound policies underlying it, suggest appropriate legislation to Congress.