The federal role in state taxation: a normative approach.

INTRODUCTION

States have considerable flexibility under the U.S. Constitution for imposing the taxes they choose. Generally, states are free to levy taxes without consideration of how other states are affected or without constraints imposed by the federal government, with two major exceptions. First, state taxes cannot run afoul of the dormant (or negative) Commerce Clause, as developed and interpreted by the U.S. Supreme Court. Second, Congress is constitutionally granted affirmative authority under the Commerce Clause to regulate interstate commerce, and it obtains most of its capacity to influence state taxes directly through this lever. (1)

This arrangement has allowed states access to the resources necessary to finance their service responsibilities and the autonomy to tax based on local factors. Federal intervention has been relatively modest, as the dormant Commerce Clause has been interpreted to give states wide latitude in developing tax policy, and Congress has generally been reluctant to exercise its affirmative Commerce Clause powers in the area of state taxation. The question arises, therefore, as to whether the very modest level of federal restraints and interventions is sufficient to prevent subnational taxing arrangements from generating perverse economic effects, such as might occur with tax havens and some other forms of tax competition. (2) If these federal limits have not been effective, or have been counterproductive, then the advisability of additional and/or modified federal interventions merits consideration.

This paper seeks to analyze the case for federal intervention in state taxation and to identify specific instances where interventions may be appropriate. We also evaluate several existing federal interventions to determine whether they are consistent with the criteria we identify for federal government involvement and whether the specific policies are likely to achieve the expectations.

The role that the federal government should play in controlling/enabling state/local government revenue generation depends on the goals that are chosen. A number of goals could be envisioned, including such esoteric values as states rights, autonomy and local control. Without reaching any conclusion on these or other goals, we take the simple view that the federal role is to increase national economic efficiency and, thus, federal intervention may be justified in cases where efficiency could be enhanced. We also acknowledge, however, that there may be sound political or constitutional reasons why the federal government should not intervene even if there are possible efficiency gains from the intervention.

Although we do not focus on it in this paper, we do note that the federal government is not without other tools that could be used to offset negative state tax externalities. The grant system, for example, is an option for offsetting negative horizontal (state-to-state) effects, as described by Inman and Rubinfeld (1996). (3) Significant restructuring of grants would be necessary, however, and the overall magnitude of grants would likely need to be much larger than those currently imposed to offset some of the effects. The federal government could also structure national taxes to offset vertical externalities that can arise between the federal and state governments. Additional tools include deductibility and credits against federal taxes for state and local taxes paid, and so forth.

A very legitimate question arises: Will the federal government act to make state/ local tax behavior more efficient or simply follow political pressures that are no more likely to make state/local taxes more efficient than if the governments were left free to design their policies independently (including pursuing cooperative measures independent of federal intervention). Certainly a case can be made that enhanced efficiency has seldom been the driving force behind federal legislation related to state/local taxation. Accordingly, we briefly examine whether federal policies are likely to encourage efficiency, as well as consider implementation strategies that might increase the probability of achieving salutary outcomes. Our primary focus, however, is on the narrower question of how appropriately set policy would make taxation in the federal context more efficient.

WHY FISCAL FEDERALISM?

The Economic Functions of Government

An extensive literature has developed regarding the economic justifications for a devolved/decentralized system of government. Richard Musgrave's (1959) seminal work is often cited as the foundation for thinking on the relative roles that different levels of government should play in a federal system. He divided government into the stabilization, distribution and allocation functions, and argued that the first two should be primarily at the national level. (4) This leaves allocative efficiency as the primary economic justification for a federal system of government. Even when allocative efficiency is considered, certain services, such as national defense, are best delivered at the national level because the benefits of uniform national consumption and internalizing consumption and political externalities exceed the costs (see Epple and Nechyba (2004)). (5) Provision of many other services by subnational governments is frequently efficiency enhancing because service levels can be differentiated according to local demands and diseconomies of large scale can be avoided. This is consistent with what is often termed the subsidiarity principle, which argues that services should be delivered at the closest possible level to service recipients. Of course, decisions on what is the closest possible level must be made after consideration of various factors, including the economies of scale/size (see Fox and Gurley (2006)) and externalities associated with service delivery.

Other reasons for devolved systems of government have also been asserted, though often not in a tight conceptual framework. For example, subnational governments have been justified as laboratories for experimentation on good government and based on a belief that government at local levels better hears the citizenry.

Tiebout and Taxes

The justifications for federalism and subnational governments have generally been developed on the service delivery/ expenditure side of government. Seldom has a case been made that state/local tax collection is more efficient than national collection, although the Tiebout model merits exploration in this regard.

Tiebout (1956) spawned a significant and continuing literature on the efficiency gains that result if a system of local governments proxies private market operation. The literature identifies the conditions under which efficiency can be attained as local consumers obtain the services they demand at minimum cost by voting with their feet. Service demands in these models are generally financed with a form of user fee that charges consumers a marginal price for services.

The Tiebout model has been extended and evaluated broadly, as discussed in Epple and Nechyba (2004). While the Tiebout framework offers many keen insights and helpful approaches to understanding local service delivery, it is difficult to make the case that state/local governments rely heavily on benefit charges. Explicit benefit charges generate a relatively small share of state and local government funding in the U.S. For example, charges and fees raised 14.4 percent of total state and 26.2 percent of total local government own-source revenue in 2004. (6) Other taxes, such as the property tax, are sometimes seen as benefit charges, but there is considerable disagreement about the extent to which this is true. (7) Further, no counterpart system of benefit taxes exists at the state level (with a few possible exceptions, such as higher education tuition). Thus, in the absence of a system of user fees, the efficiency gains that can be obtained through delivery of services through a federal system of government might be at least partly offset by the extra costs of generating revenues at the subnational level. As we explore in the following section, generating the resources to finance state and local service delivery is likely, in fact, to impose additional costs and lessen the efficiency gains from decentralization.

BASIS FOR FEDERAL INTERVENTION

Subnational taxation engenders a series of horizontal and vertical externalities that can impose additional burdens on the economy. Horizontal externalities refer to the effects that taxation by one government has on other governments at the same level, and vertical externalities refer to the effects that taxation by a government at one level has on governments above or below it. Additionally, the geographic or economic size of states may be insufficient to achieve scale economies in tax collection, and so federal participation or cooperative arrangements between states may allow lower administrative costs. Further, there may be economic savings on the compliance side if some degree of uniformity can be achieved. In short, the potential for lowering the costs of assessing, administering and complying with taxes offers the primary justification for federal intervention in state and local taxation. This section addresses each area where federal Intervention could lower the national costs of taxation.

Efficiency Costs of Subnational Taxation