Understanding uniformity and diversity in state corporate income taxes.

During their first 50 years, until the early 1960s, like Topsy, state corporate income taxes "just growed," the product of state legislation enacted with little or no regard for that of other states or the need for uniformity. (Legislation has sometimes been enacted with an eye toward gaining a competitive advantage over other states, but that came later.) This section reviews efforts to achieve uniformity during this period, culminating with the drafting of UDITPA.

The episodes examined involve five issues, the definition of income, the choice of methods of dividing the tax base among the states, jurisdiction to tax or nexus, the formula to be used to apportion income, and combination of the activities of related entities. In these episodes we will find evidence of all the aforementioned ways taxes can become more or less uniform. The outcomes of some of the episodes have been (or might have been) desirable, but others have not.

The Definition of Income

Although the first states that imposed corporate income taxes provided their own definitions of income, as early as 1917 states began to adopt the federal definition. (7) By the early 1960s the definitions of income employed by most states exhibited "moving conformity," that is, conformity to the federal definition for the current tax year, and some states conformed to the federal definition in effect at a particular time in the past ("static conformity"). See U.S. House of Representatives (1964, p. 109). Today, "the outstanding characteristic of state corporate net income taxes is their broad conformity to the federal corporate income tax." (8) Conformity with the federal definition was not the result of either harmonization or federal action. Hellerstein and Hellerstein (2007, [paragraph] 7.02) describe the forces producing convergence as follows.

Pressure from taxpayers for easing compliance

and auditing burdens has been the

prime force responsible for the very wide

conformity of the state corporate tax base

to the federal corporate tax base, prior to

allocation or apportionment of the base

among states in which the corporation

is taxable.

Changes in federal tax policy sometimes make continued conformity difficult for the states. The 1980s saw many states "decouple" from the federal definition, rather than endure the revenue losses implied by the federal provisions for accelerated depreciation enacted in 1981, and then "recouple" when the 1986 tax reforms eliminated these provisions. Many states again decoupled in response to the corporate tax cuts enacted under President George W. Bush. This zigging and zagging creates enormous complexity for taxpayers, especially if they operate in many states. See Hellerstein and Hellerstein (2007, [paragraph] 7.02).

Dividing the Tax Base

Since the early days of state corporate income taxes states have used three methods to divide the income of corporations operating in more than one state. Under formula apportionment a state taxes a fraction of the taxpayer's net income equal to the fraction of certain activities (or values) of the taxpayer that occur within the state. Specific allocation attributes to a specific state all the income derived from certain activities that occur in the state and are unrelated to the taxpayer's general multistate business operations. Separate accounting treats the in-state portion of a multistate business as a separate entity and accounts for its income separately. (It is important to distinguish between separate geographic accounting for the in-state income of a single corporation, the topic of the previous sentence, and separate entity accounting for the income of related corporations, which figures prominently in the rest of this paper.)

At first the states seemed to prefer the use of separate accounting, as did committees of the NTA that examined the issue, when separate accounts were being maintained for other reasons. The growth and integration of business during the 1920s made this practice increasingly untenable, and by the beginning of the 1930s formula apportionment was the preferred method used to divide income that was not allocated to a specific state. (9) Again, convergence occurred, without either explicit harmonization or federal action.

The Apportionment Formula

If formula apportionment is to be the preferred method of dividing income, there should ideally be agreement on a common formula. Yet, in 1929 the formulas being used by the states were "all over the map'--a situation that changed little over the next quarter century (see U.S. House of Representatives, 1964, pp. 114, 119). Writing from the vantage point of the early 1960s the Willis Committee (U.S. House of Representatives, 1964, p. 118) observed regarding this method:

[V]ariation appears to be its most significant

historical characteristic. Not only

have there always been wide diversities

among the various formulas employed

by the states, but the composition of

these formula seems to be constantly

changing.

In addressing this issue during the 1920s, the NTA committee took a stance that "has colored all subsequent efforts to achieve uniformity of income division." (U.S. House of Representatives, 1964, p. 130). It placed political expediency over principle:

All methods of apportionment ... are

arbitrary.... IT]here is not one right rule

for apportionment.... [T]he only right

rule ... is a rule on which the several states

can and will get together as a matter of

comity. (Proceedings of the National Tax

Association, 1922, pp. 198, 201, quoted

in U.S. House of Representatives, 1964,

p. 118.)

By the late 1930s the NTA committee came to support the "Massachusetts formula, which accorded equal weight to payroll, property, and sales, because it would require the least accommodation of state statutes. But whether sales should be attributed to the state of origin or the state of destination remained controversial (U.S. House of Representatives, 1964, p. 131). (10)

UDITPA

Almost from the beginning of state taxation of corporate income there have been sporadic calls for the federal government to promulgate rules for the division of income of multistate corporations. (11) No significant effort was made along these lines until 1965 (in the Willis Report, to be discussed below). In the meantime, in 1957, the National Conference of Commissioners on Uniform State Laws (NCCUSL) adopted UDITPA. In 1964 the Willis Committee (U.S. House of Representatives, 1964, p. 133) observed, "The States have not been very quick in embracing UDITPA" and noted further:

An examination of the enactment of other

uniform acts shows that uniformity is not

quickly or easily achieved through this

method.... The Conclusion is inescapable

that the voluntary adoption by the

States of any kind of uniform system is a

slow and halting process, if not a virtual

impossibility."

In other words, voluntary harmonization was not working and might never work. However, that committee's proposal for federal legislation led to a quick and largely successful, if limited, effort at pre-emptive harmonization. As John Warren (2005, p. 133) has noted, "The states' response was to rush to adopt UDITPA to show that they could solve the uniformity problem without congressional interference." By now one-half of the states that have income taxes have adopted UDITPA and the laws of most of the rest closely resemble UDITPA. This success is marred by the unilateral abandonment of the equal weighting of apportionment factors by many states (see Hellerstein and Hellerstein, 2007, [paragraph] 9.01, which lists states that have adopted UDITPA).

UDITPA distinguishes between nonbusiness income, which is generally allocated to a specific state, and business income, which is apportioned, based on the formula provided therein. Income from intangibles (e.g., interest and dividends) that is unrelated to the taxpayer's unitary business is the most important form of allocated income. UDITPA allocates most non-business income to the taxpayer's place of commercial domicile. (12) The U.S. Supreme Court has ruled, in Mobil Oil Corp. v. Commissioner of Taxes of Vermont, 445 U.S. 425 (1980), that states can apportion income from intangibles that is related to the taxpayer's unitary business, and some do so; see Hellerstein, 1993. UD1TPA foresees little role for geographic separate accounting. Its use (like that of other non-standard methods) is limited to situations where, in the language of section 18 of UDITPA, prescribed methods do not "fairly represent the extent of the taxpayer's business activity in this state."

Most corporate income is probably subject to apportionment, rather than allocation. (13) Under UDITPA, business income is apportioned by the equally weighted three-factor formula. UDITPA also defines each of the three factors. The definition of the payroll factor, based on federal rules for unemployment compensation, has usually been thought to be relatively straightforward and satisfactory, although the growth of outsourcing casts doubt on that judgment. It has the considerable virtue of interstate uniformity.