|* Partner, Cravath, Swaine & Moore LLP. All views expressed herein are solely those of the author.
I do not believe that codification of the economic substance doctrine (ESD) resulted in significant change to substantive law. Codification assured that the Supreme Court would not reject the existence of ESD based on a literal reading of the Code, but such rejection was unlikely. Codification also adopted various pro-government judicial interpretations of ESD, but court decisions were tending in this direction anyway, and taxpayers had not won many (if any) ESD cases solely because a court applied a test weaker than the codified test.
The principal substantive change made by codification is the no-fault penalty. Even there, the 40% “headline” penalty rate can be reduced to 20% by adequate disclosure. The taxpayer will normally know that ESD is an issue when it files its tax return. Therefore, the 40% rate only applies to a taxpayer trying to hide a questionable transaction from the IRS by not disclosing.
Many commentators have raised concerns that the risk of the no-fault penalty will discourage both “normal” tax planning and normal commercial transactions. This is reminiscent of concerns that a “nuclear winter” would result from the adoption of the partnership anti-abuse rule in 1994. In my experience, codification has had no adverse effect on normal commercial transactions, and has had a chilling effect only on transactions that were already very questionable. This is consistent with the experience after adoption of the partnership anti-abuse rule and with repeated recent statements by government officials that if a business transaction worked before ESD codification, it still works now.
ESD clearly prevents transactions that solely generate noneconomic tax losses. The result is less clear when a transaction allows the recognition of loss in depreciated assets without materially changing the taxpayer’s economic interest in the assets, as in the Shell Petroleum case.
The most significant issue, however, arises when the overall transaction clearly has economic substance, but tax benefits arise from steps that, standing alone, do not. An example is Gregory v. Helvering, where the underlying transaction was a taxable sale of assets, but an initial spinoff was done solely to reduce tax on the disposition. If ESD does not prevent transactions such as this, it would be almost useless in preventing abusive transactions.
Codification clearly contemplates, consistent with prior case law, that ESD can apply to disallow tax benefits arising from individual steps in a larger transaction. Yet applying ESD separately to each step of every transaction would go too far. For example, in Dover, the taxpayer obtained a tax benefit by making a check-the-box tax election. There was no economic substance to the election, yet the Tax Court upheld the transaction. Most tax lawyers, including myself, believe that the Dover transaction does not violate ESD either before or after codification.
So how can you tell whether a particular step in a transaction is Dover-style “legitimate tax planning” or Gregory-style “abusive tax avoidance”? I believe the key is the sentence in the legislative history of codification stating that ESD does not apply if the tax result was clearly intended by Congress. Courts in substance also applied this test before codification. The test obviously does not provide definitive answers, but it focuses on the correct question. The answer is ultimately a matter of judgment, but I believe most tax lawyers have a good sense of where to draw the line, and would reach similar conclusions concerning most transactions. The uncertainties arise largely in heavily tax-structured transactions.
The government has indicated that it will only provide very limited guidance on ESD, despite numerous requests. This is understandable from the government’s perspective. Taxpayers could potentially exploit safe harbors or other pro-taxpayer rules in ways not intended by the government, and a presumption against bifurcation of transactions (as some have requested) would greatly weaken ESD. As a result, ESD could end up weaker than before codification — hardly the intent of Congress. Additional guidance might not even help taxpayers, since it would raise its own interpretative issues, and favorable rules would likely be so narrow (for the reason stated above) that they could create negative implications for transactions not squarely covered.
The must justifiable objection to codification is to the no-fault 20% penalty that arises even with disclosure. Aside from revenue considerations, Congress may have believed an in terrorem penalty for highly tax-motivated transactions is justified. On balance, the penalty seems unfair. Also, it might discourage courts from applying ESD to questionable transactions, could put taxpayers in the peculiar position of arguing that if their transaction does not work, it is for reasons other than ESD, and could require courts to decide on the applicability of ESD even after disallowing the claimed tax benefits for another reason.
Finally, as noted, ESD does not apply if the tax results are clearly intended by Congress. So ESD, like other anti-abuse rules, stops transactions with tax results not intended by Congress, but only if the transactions also happen to violate ESD. The Code and regulations contain innumerable anti-abuse rules, each applying only if specified conditions are satisfied.
Every such rule raises numerous questions about its scope, and causes much tax planning to avoid it. More fundamentally, if a transaction provides a tax benefit not intended by Congress, its success should not logically depend upon whether it happens to be covered by a specific anti-abuse rule.
To avoid these results, Congress could adopt a general anti-abuse rule for the whole Code (as many other countries have done). The sole issue would then be whether the tax result of any transaction was intended by the Code and regulations. This would be a much more logical and consistent approach to tax avoidance transactions.