For some time now, many states have been moving away from the traditional three-factor apportionment formula to an approach that places more or all of the weight on the sales factor. In our view, this trend is not the result of any guiding legal or economic principles or of a broader effort to achieve a fair and uniform system. Instead, this new trend appears to be the result of politically-motivated, state-level decision-making. The emerging system of widely varied, largely unjustified apportionment formulas is regrettable and not only inconsistent with sound tax policy, but also contrary to the original goals and concerns of the Uniform Division of Income for Tax Purposes Act (“UDITPA”). Moreover, this dysfunctional system leads to serious concerns regarding the constitutionality of its inconsistent and arbitrary apportionment formulas.
Brief History of Apportionment Formulas
The U.S. Supreme Court has long recognized that states may impose an income tax on an apportioned share of a multistate business’ income.1 One way to apportion income is to use an apportionment formula that reflects (at least roughly) the in-state activities required to generate the income ultimately subject to state taxation. Traditionally, the states have used a three-factor (i.e., property, payroll and sales) formula to determine the share of income taxable within the state. The Supreme Court has long recognized the constitutional viability of such a formula.2
In the 1950s, UDITPA was drafted as a model for the division of multistate business income. The UDITPA drafters were primarily interested in achieving the uniform use of fair rules (hopefully tethered to sound economic principles), including apportionment, concerning the division of income. To that end, they advocated the use of a three-factor formula. And, at that time, almost all states with corporate income taxes adopted UDITPA’s formula or one similar to it.3 That uniformity began to splinter in the wake of the decision in Moorman Manufacturing Co. v. Bair.4
Today, apportionment formulas vary widely among the states and include the traditional equally-weighted three-factor apportionment formula, three-factor formulas that double weight the sales factor, single-sales factor formulas and elective or variable systems that defy easy classification in any one category.5 This considerable inconsistency is made worse by the staggering array of special rules applicable to certain taxpayers, diverse definitions, differing sourcing rules and often inconsistent regulatory and judicial interpretations of applicable law.6
Moorman and the Advent of Single-Sales Factor Apportionment
The taxpayer in Moorman was an animal feed manufacturing company, based in Illinois, with sales to customers in Iowa.7 The company was subject to a three-factor apportionment formula in Illinois, but a single-sales factor apportionment formula in Iowa. The taxpayer argued that Iowa’s singlesales apportionment formula violated the Due Process Clause of the U.S. Constitution, in part, because it taxed income earned in Illinois. The taxpayer also argued that Iowa’s formula violated the Commerce Clause, in part, because the interaction between Iowa’s single-sales factor apportionment approach and Illinois’ three-factor formula resulted in double taxation.8
While the Supreme Court upheld Iowa’s single-sales factor formula in the specific circumstances of the case, this legal holding does not mean that single-sales factor apportionment is economically sound or constitutes wise tax policy.9 In fact, the Court has long expressed significant concerns regarding the wisdom of such a formula.10 However, the Court also has recognized the flaws inherent in three-factor apportionment formulas and has acknowledged that, in any specific instance of apportionment, double taxation may result from either formula.11 Accordingly, the Court has refused to force the states to use any particular formula to apportion income. As such, the Supreme Court determined in Moorman that Iowa’s single-sales factor apportionment was constitutional.12 But finding on specific facts and arguments that a particular apportionment formula does not violate the Constitution is not the same as an endorsement of the wisdom or effectiveness of such a formula.
Economic Theory of Apportionment
At its simplest, apportionment is a process (or a step in a process) intended to estimate how much of a multistate business’ income is fairly attributable to sources within a particular state.13 Apportionment formulas are mathematical processes. The inputs of these processes are geographically traceable factors representing sources of income.14 The outputs of these processes are numbers between zero and one that, when multiplied by multistate business income, are intended to approximate the amount of that income taxable in the apportioning state.
Property and labor have long been recognized as significant contributors to the generation of income.15 In addition, a business’ investments in both property and labor (e.g., payroll) can be traced to geographic locations.16 Thus, including property and payroll factors in UDITPA’s three-factor formula was a sensible decision based on an economic guiding principle and sound tax policy.
Inclusion of a sales factor is harder to justify and may depend as much on political reality as economic theory or tax policy. In fact, as UDITPA was being debated, many economists believed that the sales factor should serve no role in apportionment formulas.17 As originally envisioned, the sales factor was intended to represent the contributions that a taxpayer’s market for goods and services in a particular state made to that taxpayer’s income.18 In today’s environment, we seriously question whether this initial purpose for the sales factor still holds true in the very different economic landscape and in light of the new and ever-changing rules governing the calculation of the sales factor.
The sales factor is also problematic because the contributions to income that it reflects are difficult to geographically trace.19 Sales factor sourcing rules have long been one of the most controversial aspects of apportionment.20 The sales factor has been variously sourced to the customer’s location, the delivery location, the seller’s location or a split between multiple locations.21
Despite these problems, the framers of UDITPA concluded that the sales factor was useful in promoting UDITPA’s goal of uniformity because its inclusion protected the interests of states in which products and services were sold.22 As one commentator noted:
Ordinarily, in apportionment formulae, the property and payroll factors tend to favor the state of origin or production in the assignment of income. The basic theory justifying the inclusion of the sales factor in such formulae is that it offsets the effects of the property and payroll factors and protects the interest of the state of destination. This seems appropriate since the state of delivery contributes to the income arising from the sale; it has provided the market.23
By their very nature, all apportionment formulas yield imperfect results and provide only rough approximations of the sources of income. Accordingly, commentators have long recognized that all apportionment formulas are, at some level, arbitrary.24 This arbitrariness is (at least in part) due to the fact that, even assuming all factors of income could be identified, the relationship between those factors is so complex that a single formula is unlikely to ever provide accurate results for all businesses. States have tried to combat this problem by adopting industry-specific formulas, but such formulas are also inadequate to the task of precisely and accurately dividing the income of each unique business, even within a similar industry.25 That said, the inherent shortcomings of formulary apportionment in general should not be used to justify the inadequately narrow and even more arbitrary system of single-sales factor apportionment.
Political Motivations, Not Any Economic Guiding Principle or Sound Tax Policy, Have So Far Driven the Trend Toward Single-Sales Factor Apportionment
One is hard pressed to think of an economic guiding principle or a sound tax policy argument that justifies abandoning the three-factor apportionment formula in favor of a single-sales factor formula. Indeed, single-sales factor apportionment is virtually unjustifiable economically.26 Among other things, such a formula ignores completely the contribution to income of labor and property, which, as discussed above, are two critically important contributors to income.
The truth is that, for the most part, states have adopted singlesales factor apportionment formulas because lawmakers and policymakers perceive that doing so potentially provides economic advantages to the state (e.g., in the form of increased in-state economic activity or increased tax revenue) with little political cost.27 Increasing the relative weight of the sales factor rewards businesses whose relative in-state presence (reflected by the average of their property and payroll factors) exceeds their in-state sales factor.28 Businesses whose relative in-state presence (reflected by the average of their property and payroll factors) is less than their in-state sales factor bear an increased tax burden relative to their income.29
In other words, moving to single-sales factor apportionment generally shifts tax burdens from companies that concentrate substantial property and labor in the state to companies that do not. It seems clear, at least in large part, that the purpose underlying single-sales factor apportionment is to provide tax breaks for companies with substantial in-state property and payroll, while potentially offsetting the decrease in tax revenue by increasing the tax burden on taxpayers that have the majority of their property and payroll outside the state. Indeed, for example, when California moved to single-sales factor apportionment, it projected a revenue increase resulting from the change to a single-sales factor formula from a three-factor formula (even with the double weight given to the sales factor in the apportionment formula that California had been using since the mid-1990s).30 In essence, California managed to pass tax legislation that increased revenue, while at the same time reducing the tax liability of numerous taxpayers with a large in-state presence (in terms of property and payroll). All things being equal, it follows that companies with the majority of their property and payroll outside California will bear the tax burden that supports the increase in state revenue.
As we alluded to above, we cannot resist wondering whether a court may find that a situation like the one in California fits squarely within the Supreme Court’s pronouncement that discrimination against interstate commerce simply means “differential treatment of in-state and out-of-state economic interests that benefits the former and burdens the latter.”31
States do not seem to apply any principled framework when adopting single-sales factor formulas. Rather, they seem to enact these formulas based on political motivations without regard for uniformity, tax policy or sound economic theory. This trend seems to have been sparked by the Moorman decision because, at least on its surface, the decision painted single-sales factor apportionment with a veneer of constitutionality. But even assuming that constitutional challenges to single-sales factor apportionment are unsuccessful, is the ad hoc, politically motivated process by which these formulas are adopted a sensible approach to important tax policy? Even if every state adopts a uniform single-sales factor formula, thereby eliminating the risk of double taxation, is there any reason to believe those rules provide anything more than a largely arbitrary division of income? It seems a reasonable proposition that laws governing the complex system of state taxation in the United States should be the result of deliberate and thoughtful processes, not ad hoc, state-specific decisions based mostly on economic and political motivations. Raising policy concerns and challenging the constitutionality of single-sales factor formulas may at least shed light on, and possibly succeed in curbing, the arbitrary nature of the trend in favor of single-sales factor apportionment.