Tax shelters and the theory of the firm

Christopher Borek, Angelo Frattarelli, Oliver Hart 02 July 2013



Over nearly eight decades, decisions by the Supreme Court and federal courts have formulated an economic substance doctrine, which requires a transaction to have a non-tax benefit to qualify for favourable tax treatment.

But a clear definition of economic substance has not emerged. One commentator has described the doctrine as akin to a “smell test” (Lederman 2010). Standards regarding the economic substance test were particularly muddled in 2001 when the Internal Revenue Service announced that it would disallow losses generated by contingent liability tax shelters – corporate transactions that purport to shift the responsibility for managing future expenses to a separate subsidiary (IRS Notice 2001-17, 2001-1 C.B. 730).

In 2010 Congress codified an economic substance test that requires a transaction to have a legitimate business purpose and “present value of the reasonably expected pre-tax profit [that] is substantial in relation to the present value of the claimed net tax benefits” (26 USC. 7701 (0)(2)(A)).

The property-rights theory of the firm

The property-rights theory of the firm can be helpful in evaluating whether a corporate reorganisation has economic substance (the argument is spelled out in more detail in Borek et al. (2013), where all references can be found). This theory takes the view that firm boundaries matter in a world where contracts are incomplete. Under these conditions not every use of every asset can be specified in a contract. According to the property-rights theory of the firm:

  • The owner of each asset gets to decide the uses not specified; a firm is defined by the assets over which it has residual control rights.

For example, the difference between a situation where company A contracts with company B for a service and a situation where company A acquires company B and the service is delivered in-house is that in the former the owner(s) of A has residual control rights over 'A' assets and the owner of company B has residual control rights over 'B' assets; while in the latter the owner of company A has residual control rights over all assets.

Why do residual-control rights matter? Suppose that it is important to incentivise the manager of the B assets. A natural approach is to pay manager B according to the profits of B activities. However, this may be difficult to do if B assets are owned by A. The problem is that profits are hard to measure objectively and A can use its residual control rights to manipulate these. For example, A could charge some of unit A’s expenses to unit B. As a result, if manager B increases unit B’s productivity manager B may find that most of the reward goes to the owner of company A.

This 'integrity of profit' problem does not exist if B is a separate company owned by manager B, since then A cannot use its residual control rights to divert profit from B. Note, however, that the integrity problem does exist if B is a separate entity wholly controlled by A. In other words, according to the theory, if A wants to incentivise manager B, it is not enough to spin off unit B; control over B operations must also be relinquished.

Black and Decker versus US

In 1998, Black and Decker stood to realise a substantial capital gain from the sale of several business divisions. It also faced growing expenses associated with its provision of employee and retiree health benefits. Black and Decker entered into a transaction that shifted the management of its employee and retiree healthcare benefits plans to a subsidiary, 'BDHMI'. BDHMI was responsible for managing the healthcare benefits for Black and Decker employees and retirees over the period 1999-2007. While a retired company executive was added to the BDHMI board, the individuals managing the company’s healthcare expenses remained largely the same. The risk-adjusted net present value of Black and Decker’s healthcare liabilities associated with BDHMI was approximately $500 million.

Black and Decker provided two motivations for the transaction.

  • First, Black and Decker asserted that isolating healthcare liabilities would allow them to be managed more effectively.
  • Second, Black and Decker argued that it was entitled to a double deduction of the healthcare liabilities based on the way the BDHMI transaction was structured.

Under the economic substance doctrine, Black and Decker was entitled to all the tax deductions claimed only if the assertion of non-tax, management efficiencies was merited. The IRS disagreed with that assertion and the dispute ended up in court.

The US (with Hart as a testifying expert, Borek providing research support, and Frattarelli as lead trial counsel) noted that a striking feature of the Black and Decker-BDHMI transaction was that residual control rights did not shift. Black and Decker maintained complete control over the operations of BDHMI. Specifically,(a) Black and Decker had the majority of the votes of BDHMI; (b) Black and Decker could elect the majority of the board members and replace them at any time; and (c) all decisions of the BDHMI board had to be ratified by a Black and Decker senior vice president.

For these reasons, the US was not swayed by Black and Decker’s claim that the BDHMI transaction made management more independent and allowed the use of higher-powered incentive schemes, yielding efficiency gains. Given that there was no change in control, any improvements in healthcare benefits could have been achieved 'inside the firm' by changing the incentive arrangements between Black and Decker and its employees and outside contractors. Moreover, doing things this way would have saved substantial transaction costs amounting to nearly $3 million in legal and other fees.

Prior to trial the district court granted summary judgement in favour of Black and Decker arguing that, even if the transaction was tax-motivated, it was not a sham. The Fourth Circuit reversed on the grounds that the government offered evidence permitting a reasonable trier of fact to conclude that Black and Decker lacked a reasonable expectation of earning a non-tax profit from the BDHMI transaction. The case settled soon after.

WFC Holdings Corporation (Wells Fargo) versus US

As of 1998, Wells Fargo had sublet numerous properties to third parties. Many of these leases were underwater in the sense that Wells Fargo earned less from its subtenants than it was paying under the original leases. Wells Fargo transferred several underwater leases to Charter, a dormant subsidiary. Wells Fargo asserted that it was entitled to what amounted to a double tax deduction. One in the form of a capital loss of about $425 million from the sale of stock and another in the form of ordinary and necessary business expenses associated with the transferred leases.

Wells Fargo’s Business Case highlighted two important economic benefits of creating Charter. First, managers would be incentivised to beat the market. Second, transferring the underwater leases to Charter would allow for improved negotiations with subtenants that had deposit or borrowing relationships with Wells Fargo and had leveraged these arrangements to get more favourable leasing terms.

Similar to Black and Decker, a striking feature of the Wells Fargo-Charter transaction was that Wells Fargo maintained complete control over Charter. Wells Fargo had 99% of the votes of Charter and could elect or replace the board at any time. The US (with Hart as a testifying expert, and Borek providing research support) argued that given this the idea that the Wells Fargo-Charter transaction made management more independent and that this permitted the use of higher-powered incentive schemes, or allowed management to bargain more aggressively with subtenants, was unconvincing. Following a trial in October 2010, the district court found that Wells Fargo lacked a legitimate business purpose for this transaction, and that the transaction lacked economic substance. Wells Fargo’s appeal is pending.


Evaluations of economic substance and/or business purpose often depend on an assessment of whether a particular party’s income claims look more like equity or debt, or on a present value calculation of the efficiency benefits of the transaction. However, characterising a party’s income rights can be challenging, and calculations of net present value are sensitive to assumptions and judgement.

The view advanced here is that the property-theory of the firm provides an alternative and simpler approach. According to this theory, the fundamental question is: did control change?

Of course, a natural question follows. If the economic substance of a transaction is evaluated according to whether control changes, might parties not simply respond by designing their transaction so that control does change? This is possible. However, the costs of structuring a transaction in such a manner are likely to be significant. In the Black and Decker and Wells Fargo cases giving up control of employee and retiree health benefits on the one hand or real-estate operations on the other is not something that either company ever seems to have contemplated. It seems unlikely that they would have easily embraced such a strategy.


Borek C, A Frattarelli and O Hart (2013), “Tax Shelters or Efficient Tax Planning ? A Theory of the Firm Perspective on the Economic Substance Doctrine”, NBER Working Paper No. 19081.

Lederman, L (2010), “W(h)ither Economic Substance?”, Iowa Law Review 95:389-444.



Topics:  Taxation

Tags:  tax shelters, economic substance, theory of the firm

Managing Principal, Analysis Group

Assistant Chief of the Civil Trial Section, Southern Region, US Department of Justice Tax Division

Andrew E. Furer Professor of Economics, Harvard University


CEPR Policy Research