The federal system in the United States consists of the federal government, 50 state governments, and more than 80,000 individual local government units. In other words, this multi-tiered structure of state and local levels is one of the most important features of the U.S. federal system. Local governments are creatures of states and are granted certain levels of autonomy from the state. Fiscal decentralization, which involves the devolution of fiscal authority and/or service delivery responsibility to lower levels of government, has been discussed in the literature of public finance and management in the U.S. From the theoretical perspective of economic efficiency, fiscally-decentralized state and local relations are the bedrock assumptions on which both the Tiebout model1 and the theory of fiscal federalism2 are built (Oates, 1972, 1977; Tiebout, 1956). Both assume that local governments under the condition of fiscal decentralization can be more responsive to the needs and preferences of local residents and enhance the efficiency of government. A large body of empirical and theoretical work examines the effects of decentralized systems of revenue generation, spending, and revenue sharing between levels of government on the spending and revenue burden of all governments in the system (Bahl, 1984; Gramlich, 1997; Hendrick, Jimenez, & Lal, 2011; Oates, 1972, 1977; Sjoquist, 2003).
Despite widespread recognition of the contribution of fiscal decentralization to government size in theoretical works within the U.S. federalist system, few researchers have studied decentralized government systems in the context of state and local government long-term debt outstanding, an important indicator of the relative magnitude of government borrowing. Along with spending and revenue, debt is a crucial factor in a government's financial equation, as much of the U.S. government's current expenditure on infrastructure is financed by borrowing. Furthermore, long-term debt outstanding is usually serviced by governments' own-source revenues. Additionally, many states are now issuing debt to cover operating deficits (Bilfuco, Bunch, Duncombe, Robbins, & Simonsen, 2012).
Given that local governments exist in different levels of decentralized systems that vary from one state to the next, we examine how fiscal decentralization links to state and local government own-source revenue capacities to service government long-term debt outstanding (capacity to service debt outstanding) as well as state and local government own-source revenue capacities to service government long-term debt issuance (capacity to service debt issuance). We estimate models of debt servicing capacities of all state and local governments aggregated at the state level from 1962 to 2012 (at five-year intervals). For comparison, we estimate models for state governments alone, as well as local governments alone for the same years by using the state as the unit of analysis. Because local governments also vary systematically in the extent to which state-local fiscal and institutional structures constrain their abilities to generate revenue, divest spending responsibilities, and share revenue with them (Hendrick et al., 2011; Pagano & Hoene, 2010), we also control for other fiscal, institutional, political, and socio-economic factors at the state level.
This study offers insight into the link between fiscal decentralization and both flow and stock measures of debt. Because we find a tendency toward greater fiscal centralization at the state level through descriptive analysis, it becomes more important to understand how fiscal decentralization relates to capacity to service debt among various levels of government. Our regression analyses show that fiscal decentralization is positively linked to state and local government capacity to service long-term outstanding debt, whereas it has little impact on their capacity to service long-term debt issuance. The conclusion is that the allocation of expenditure responsibilities to both state and local governments is closely linked to government capacity to service long-term debt outstanding held by state and local governments, and by local governments alone. Moreover, we add an interactive variable between fiscal decentralization and tax and expenditure limits (TELs) to explore the interactive effect on government capacity to service debt across states. We find that the positive effect of fiscal decentralization on capacity to service debt increases with the restrictiveness of TELs. Thus, it provides researchers and practitioners with a better understanding of own-source revenue capacity to service debt among various levels of government within the intergovernmental context based on the literature of fiscal federalism in the U.S.
We next provide a review of the literature on decentralized governing structure primarily in the context of U.S. fiscal federalism and studies linking these issues to debt. In the third section, we present the operationalization of variables and introduce data sources and the methods of analysis. The estimation results are then presented and discussed. This paper concludes with a summary of the results, their policy implications, and directions for future research.
(2.) LITERATURE REVIEW
Research on fiscal federalism documents the actual effects of different fiscal governing structures on providing and producing public goods and services. Conclusions about the level and type of government that make different functions more efficient or equitable may subsequently be drawn. State and local governing relations have been described as a continuum of state control versus local autonomy (Stephens, 1974), and also encompass the extent to which fiscal authority and/or service delivery responsibility is concentrated at the top or bottom of a system (Hendrick, 2011; Hendrick et al., 2011). That is, one can determine which U.S. states are more fiscally-decentralized and have given greater discretion or authority to local governments over their finance, structure, functions, and fiscal responsibility of delivering services and goods. On the other hand, state-local relations that are centralized limit local discretion and assume greater responsibility for delivering local services. Centralized state-local relations may result in a large Leviathan government3 and increase the size of the public sector (Brennan & Buchanan, 1980). In many cases, centralized statelocal systems are also more paternalistic and intrusive towards local government (Stephens & Wikstrom, 2000).
Both international and U.S. studies on fiscal decentralization have focused on examining the consequences of the sharing of fiscal responsibilities between the central government and subnational governments or between state and local governments on government size and financial deficit. Research supports the notion that fiscal decentralization reduces the size of national governments and expands that of subnational governments in 32 industrialized and developing countries (Jin & Zou, 2002), Switzerland (Feld, Kirchgassner, & Schaltegger, 2010), and European countries (Cassette & Paty, 2010). However, fiscal decentralization does not lead to an increase in the size of aggregate governments, as fiscally decentralized countries tend to experience slower economic growth (Baskaran & Feld, 2013; Gemmel, Kneller, & Sanz, 2013; Rodriguez-Pose & Ezcurra, 2011).
The majority of empirical research in the U.S. supports federalists' claims that a fiscally decentralized state-local sector is associated with reduced government expenditures or revenues, suggesting that fiscal decentralization improves service delivery efficiency. There are others who argue, however, that having many differentiated and specialized local governments in a fiscally decentralized state does not by itself ensure there will be allocative and productive efficiency (Brennan & Buchanan, 1980). This also suggests that fiscal decentralization could be associated with higher debt burden, ceteris paribus, particularly at the local level. When local governments are observed separately, those in states with higher levels of fiscal decentralization will have greater spending and more debt, as responsibility for services is devolved to the lower level of governments. In addition, the loss of economies of scale may account for greater spending or debt in decentralized systems (Boyne, 1992; Feld et al., 2010; Forbes & Zampelli, 1989; Oakerson 1999; Zax, 1989). As demonstrated by these different theoretical perspectives, whether fiscal decentralization results in higher or lower levels of government debt issuance is unclear.
The limited empirical research on decentralized governing structures suggests that decentralized systems issue higher levels of debt. Bahl and Duncombe (1993) hypothesize that states with a more expansive view of the role of governments choose higher debt levels. They also find that state governments with a tradition of spending at high levels are more likely to take on higher debt levels. Trautman (1995) finds that states with more decentralized management structureshave higher levels of outstanding debt than centralized states. The findings support the assumption that less central oversight, especially for public authorities, is associated with increased levels of borrowing.
In De Mello's (2001) research, fiscal decentralization at the national level is associated with higher subnational borrowing costs. He argues that less central government control over local government spending and reduced incentives for prudence in the management of debt increases the costs of debt. Fiscal decentralization may also deepen vertical imbalances between levels of government in the form of transfers of resources from the national to subnational levels of government (Rodden, 2002). Greater vertical imbalance, in turn, gives local governments incentive to overspend and issue...