Richard Posner is one of the most influential legal scholars of recent generations. He is perhaps best known as a leading figure in the school of Law and Economics. Complimenting his academic work, he served as a judge on the Seventh Circuit Court of Appeals for 36 years before retiring in 2017. In the field of taxation, one of his more memorable decisions was Exacto Springs Corp. v. Commissioner, 196 F3d 833 (7th Cir. 1999), which concerns the characterization of payments from closely held corporations to individuals who are both shareholders and employees: is the payment properly classified as a salary or as a distribution?
The question of how to characterize payments to shareholders arises whenever shareholders provide services or sell property to the corporation that they control. If a shareholder leases property to a corporation, is the payment that the parties describe as rent truly rent or is it only partly rent and partly a distribution? The issue of classification is particularly significant in the field of international taxation. For example, if a corporation operating in Country A pays what it describes as a royalty to a parent (or otherwise related) corporation in Country B, is the payment actually a deductible royalty or is it a nondeductible distribution? The answer to that question may determine whether Country A can collect tax from the economic activity in its territory.
In principle, the answer to the question of classification is simple. Any amount above the market value of the property or services is a distribution. Thus, if a corporation leases from a controlling shareholder property with a fair market rent of $100 and pays $120, then $100 of the payment is properly classified as rent and $20 is properly classified as a distribution (whether the distribution is a dividend or a distribution of capital will of course depend upon its earnings and profits at the time). However, the practical problem in applying this rule is that it is often difficult to determine the fair market value of the services rendered. In the royalty example, unless the parent corporation habitually grants similar licenses to unrelated licensees, there is no obvious benchmark for determining a fair market royalty. Comparisons with other corporations and other industries suffer from the obvious drawback that the intellectual property licensed in each case is unique.
The same problem arises when determining fair market salaries for controlling shareholders who are also employees. Where shareholders perform fairly routine tasks, it is likely that a market price could be reasonably determined. However, when – as is usually the case – controlling shareholders are hired to manage the corporation, it is not easy to determine what salary they would have been able to negotiate in an arm's length transaction.
In approaching this task, courts have traditionally investigated a number of what they consider to be relevant factors and in so doing have produced what have become known as "multifactor tests." For instance, the Tax Court in Exacto Springs considered the type and extent of services rendered, the scarcity of qualified employees, the qualification and prior earnings of the employee, the contribution of the employee to the business, the prevailing compensation paid to comparable employees, and the particular characteristics of the employer's business. While there is nothing here that allows a clear mathematical computation of the appropriate compensation, considering these factors – perhaps among others – can certainly guide a court's thinking about how much a particular employee should be paid.
Judge Posner rejected the multifactor test as nondirective, arbitrary, and unpredictable. In its stead, he proposed an independent investor test. Under this test, the key question is whether, after paying the employee's salary, shareholders receive a reasonable return on their investment. If they do, then the salary is presumptively reasonable. He went on to clarify that the presumption can be rebutted if it is shown that the return to capital is not due to the CEO's exertions. One problem here is that the extent to which a particular employee contributes to the success of the employer's business is not a simple task. Note that this is one of the multifactor tests and, moreover, possibly the most difficult of them to quantify. Another problem is that Judge Posner gave no indication how to determine the appropriate salary if the presumption is successfully rebutted.
In his article, Professor Listokin points out another significant problem with the independent investor test. Under ordinary circumstances, residual profits belong to equity investors. Other service providers – including employees – typically receive more-or-less fixed compensation for their services. However, under the independent investor test, the tables are turned: shareholders are entitled to a more-or-less fixed return on their capital and any residual profits belong to the employee (that is, to the shareholder/employee is her capacity as an employee). Such a distribution of the risks and rewards of a business does not appear to correlate well with ordinary business practices and it is unreasonable to assume that parties operating at arm's length would agree to such an arrangement. Furthermore, the return to equity can be highly volatile.
Listokin posits that the independent investor test may be more applicable to the problem of distinguishing debt from equity than it is to distinguishing salaries from distributions. His argument is that because debt is less volatile than equity it is appropriate to determine the return that a creditor would reasonably expect and to conclude that anything up to that amount is presumptively interest. To my mind this is a misapplication of the independent investor test. As I noted earlier, in the case of a dual capacity shareholder, the underlying question is always what the services provided are actually worth. Any payment up to that amount is properly considered payment for the service (rent, salary, etc.); anything exceeding that amount is properly considered a distribution. In the debt/equity case, shareholders are also lenders. The service they provide is allowing the corporation temporary use of their money. Therefore, applying the general classification rule that looks to the value of the services provided, we would say that whatever return a third party lender could expect to receive in an arm's length transaction is properly classified as interest. The independent investor test, on the other hand, flips this formula on its head. It would say that any interest paid is reasonable provided that, following the payment of interest, equity owners receive a reasonable return on their investment. Such an argument does not appear attractive. In fact, it may serve further to undermine the independent investor test.
Listokin concludes by noting that judicial tests should focus on economic substance, and that multifactor tests fail in that regard. Substantive accounts, such as the independent investor test are better, and if they are unsuccessful – and he indicates that the independent investor test falls into that category – they should be replaced by other substantive tests. In principle I would agree. In the case of dual capacity shareholders, I would argue that the appropriate substantive test is the value of the services provided. In the case of rent or interest, this is often relatively straightforward. In more complicated cases, such as salaries, the multifactor tests – if understood not as independent substantive tests but as aids in determining what the services are actually worth – may have an important role to play.
In an era of intense partisanship and angry and disrespectful political discourse (as described by the author toward the end of his article), it is sobering to consider current issues in the light of a broader historical context.