IV. FISCAL MANAGEMENT AND STABILIZATION POLICY IN FEDERAL SYSTEMS: THE EXPERIENCE OF ARGENTINA AND MEXICO Bradford G. Reid [ From the book, Fiscal Relations in Four Countries: Four Essays, Paul Boothe, ed. Ottawa: Forum of Federations, 2003 ] PART 1: FISCAL MANAGEMENT Prudent fiscal management requires a clear delineation between the long-run and short-run behaviour of the government’s budget position. The long-run objective of the fiscal authority should be to achieve a sustainable fiscal position in which its budget is balanced. This requires a matching of expenditure responsibility with adequate revenue resources so that the “structural” budget position is one of balance. Long-run balanced budgets allow a separation of the fiscal authority from the monetary authority and give the monetary authority the necessary autonomy to pursue long-run price stabilization as its policy objective. Balanced budgets reduce or eliminate fiscal authority reliance on the inflation tax as a revenue source, and this in turn provides the monetary authority with the needed commitment and credibility to achieve an environment of long-run price stability. If government revenues and/or expenditures are sensitive to cyclical fluctuations in economic activity then the budget position of the fiscal authority should respond to these fluctuations in the short run. Surplus budget positions should occur in business cycle upturns and deficit budget positions should occur in business cycle downturns. These short-run cyclical movements in the fiscal authority’s budget are optimal both from the perspective of traditional Keynesian aggregate demand management models and from the perspective of minimizing dead-weight losses through expenditure and tax rate smoothing (Barro 1979). Optimal fiscal management is thus composed of budgets that are structurally balanced in the long run but are also cyclically sensitive to short-run economic conditions. This requires an ability to borrow to finance fiscal deficits during business cycle downturns and a willingness to save fiscal surpluses during business cycle upturns. Achieving optimal short-run fiscal management objectives, in particular sustaining the ability to borrow during cyclical downturns, is ultimately contingent on achieving structural budget balance in the long-run. If the fiscal authority’s budget is one of structural deficit then long-run borrowing and debt accumulation will occur. The market will eventually become reluctant to absorb additional debt, which will force the fiscal authority to abandon its structural deficit policy. However, this market-imposed borrowing constraint will also limit the fiscal authority’s ability to finance cyclically occurring deficits and curtail its ability to follow an optimal short-run fiscal management policy. Gavin and Perotti (1997) have found that Latin American fiscal outcomes, in contrast to the fiscal outcomes of industrial economies, have been too “pro-cyclical” and that the lack of counter-cyclical response has been due, at least in part, to the existence of market-imposed borrowing constraints. Fiscal management becomes a more complicated issue in federal systems of government where there is likely to be a substantial amount of interdependence between the national and subnational governments. This interdependence arises from shared revenue sources, from shared expenditure responsibilities and from intergovernmental transfers among different levels of government. Dillinger and Webb (1999) argue that the sustainability of both national and subnational government fiscal management practices in a federal system depends upon the existence of: (1) revenue autonomy and expenditure autonomy between different levels of government, and (2) market-imposed borrowing constraints, particularly on subnational government debt issue. The amount of revenue and expenditure autonomy within a federal system will determine the amount of control that the various levels of government exert over their own budget positions. Unsustainable public sector budgets are less likely to occur when each level of government is allocated its own revenue sources with which to finance its own expenditures, and when these revenue sources are sufficient to meet the expenditure responsibility. With appropriate autonomy, each level of government can match the benefit stream associated with expenditures to the cost of funds required to finance that benefit stream. Dillinger and Webb also argue that independence among different levels of government with respect to their borrowing decisions is an important element in achieving fiscal management sustainability in federal systems. Sustainable fiscal positions are more likely to occur when the national government does not bail out subnational governments and forces subnational governments to service their own debts. Without national government bailouts, creditors will be forced to accept any losses caused by the failure of subnational governments to service their debts and market discipline will be imposed on subnational debt issue. PART 2: FISCAL MANAGEMENT CASE STUDIES a) Argentina Since the mid-1970s, Argentina has been characterized by a significant amount of decentralization of expenditure responsibility from the national government to provincial and local governments. Revenue-raising power has remained concentrated at the national level so that there is considerable reliance by provincial and local governments on transfers from the central government to finance subnational government expenditures. A clear imbalance exists between expenditures at the provincial and local government levels and the availability of own-source revenues at those levels. Constitutional expenditure assignment provides the central government with exclusive responsibility for defence, foreign affairs, international trade, the regulation of interstate trade, monetary policy, immigration policy and the provision of unemployment insurance. Responsibility is shared between the federal and provincial governments in the areas of social welfare, police and highways and between the federal, provincial and local governments in the areas of health and education. The Argentine constitution also assigns tax sources to the federal and provincial governments but, in practice, the provinces have delegated much of their responsibility for the legislation, administration and collection of their tax sources to the federal government. This delegation of taxing authority has been accompanied by the creation of a system of transfers from the national government to the provinces. There are three basic transfer mechanisms: (1) co-participation transfers that provide automatic, non-discretionary transfers to the provinces from income taxes, excise taxes and the value-added tax; (2) other automatic transfers that provide a sharing of revenue collected from the fuel tax, energy tax and wage tax; and (3) discretionary transfers. The co-participation transfers account for about two-thirds of all transfers from the federal government to the provincial governments. While provincial dependence on transfers from the federal government is large, this in itself has not weakened provincial power relative to the federal government as much of the transferred amount is mandated and not at the discretion of the federal government. However, the federal government ultimately determines the total pool of funds available for transfer by setting the tax rates on those tax sources mandated for revenue sharing. The decentralization of expenditure responsibility with continued centralized revenue collection has created severe vertical imbalances between the national and subnational levels of government. Many of the provincial governments have encountered continuing fiscal management difficulties, giving rise to long-run structural deficits in their budget positions. In Argentina all levels of government are allowed to borrow both domestically and internationally. Considerable amounts of subnational borrowing have occurred as the provinces have at times not been able to control the relationship between their expenditures and revenues. During the 1980s provincial borrowing was financed largely from two sources: central government loans to the provinces and loans from provincial government-owned banks to the provincial governments. The first of these lending relationships created a relationship between national and subnational fiscal positions and the second created a relationship between subnational fiscal positions and national monetary policy. Both of these relationships allowed provincial governments to avoid market-imposed discipline on their borrowing activities and encouraged subnational structural budget deficits. Lending from banks owned by provincial governments to those provincial governments led to a monetary policy linkage as these loans were rediscounted by the central bank of Argentina prior to 1991. Thus provincial government debt issue influenced the rate of monetary expansion and the rediscounting gave provincial governments access to a share of national seignorage and inflation tax revenue. This in turn reduced the ability of the central bank to control the money supply and the rate of price inflation. In 1991 this linkage was ended by the convertibility law, which prohibited central bank rediscounting of provincial bank loans. b) Brazil The democratization process that occurred in Brazil during the 1980s was also accompanied by a process of decentralization in its federal system of government. In particular, decentralization on the revenue side has resulted in a greater degree of control over revenue sources by state and local governments than previously existed. The 1988 constitution assigns relatively few expenditure responsibilities exclusively to particular levels of government. Most expenditure functions are shared responsibilities. With respect to actual spending, the federal government allocates most of its program spending to social security and social assistance. State program spending occurs primarily in the areas of education, social assistance and health. More exclusivity is found in the assignment of revenue sources. The national government collects its revenue from the personal and corporate income taxes, a selective value-added tax, payroll taxes associated with social security provision and taxes on foreign trade. The constitution assigns to the state governments revenue from a broad-based value-added tax, motor vehicle taxes and estate taxes. In recent years, growth has been stronger in those revenue sources assigned to the state governments. The degree of revenue decentralization is understated by the constitutional assignment of revenue sources between levels of government in Brazil. There is an extensive array of intergovernmental grants from the national government to subnational governments. Among these grants are mandated transfers from the federal government to the states that result in the sharing of federal VAT revenue and federal income tax revenue. While the proportion of sharing is mandated, the federal government controls the pool of funds available for sharing by setting the rates of taxation on the shared tax bases. In recent years there has been a movement by the federal government away from raising funds using shared tax sources and toward raising funds from non-shared sources (like payroll taxes). This has created some concern about too heavy a reliance on taxes, like the payroll tax, that are more distortionary than other taxes. The substantial decentralization of revenue to the state governments contained in the constitution of 1988 led to initial concern that this could create significant fiscal management problems for the federal government. In fact, the state governments have experienced the most difficulty in achieving long-run sustainability in their budget positions. With no clear boundary between federal and state expenditure responsibility, the federal government was able to adjust to the process of revenue decentralization through a combination of ad hoc expenditure off-loading to the state governments and a reduction in discretionary transfers to the states. This downloaded a structural deficit problem to the state level, which has not been successfully addressed by the state governments. The state-level structural deficit problem has been exacerbated by an environment in which states have not faced an appropriate amount of market discipline in their borrowing to finance these deficits. States have been allowed to own commercial banks and to borrow from these banks. The federal government and the central bank have demonstrated a willingness to provide bailouts to these commercial banks in the event of default, so that state borrowing has been able to occur in a relatively unconstrained fashion. c) Mexico The Mexican federation is characterized by a relatively high degree of expenditure centralization combined with limited revenue-raising powers assigned to the state and municipal governments. The only sources of own-revenue for the state and local levels of government are property taxes, fees and user charges, with all other revenue sources assigned to the national government. The constitution mandates that state governments share in the revenues collected from specific tax sources but the size of that revenue sharing is determined by the federal government. Lower levels of government in Mexico rely heavily on this revenue sharing with the federal government to finance their expenditures. State and local governments have the authority to borrow in domestic capital markets. The ability to borrow internationally is limited to the central government. During the 1990s state and local governments experienced problems with structural deficits. State governments have frequently borrowed in the domestic capital markets and have often “guaranteed” these loans by pledging future expected flows from revenue sharing with the federal government as collateral to the lending banks. This practice has generated concern at the federal level about the need for central government bailouts in the event that states become unable or unwilling to service their debts, both from the perspective of maintaining stability of the domestic banking system and from the perspective of maintaining the flow of public services. PART 3: STABILIZATION POLICY The traditional view of the public finance literature was that responsibility for the conduct of stabilization policy should be assigned to the national level of government. Clearly, if the federation is to be a monetary union, monetary policy must be conducted at a centralized level. However, fiscal policy could be conducted under either centralized or decentralized decision-making in a federation composed of higher and lower levels of government. Several arguments have been advanced in favour of centralized fiscal policy for stabilization purposes. First, it has been argued that the primary sources of short-run cyclical fluctuations are shocks common to all regions of a country, rather than region-specific shocks, so that a national fiscal policy response is the appropriate policy tool. Second, when regional economies are more open than the national economy, regionally implemented fiscal policy will be subject to a higher leakage effect. This will reduce the incentive for lower levels of government to engage in counter-cyclical policies and increase the incentive to free-ride on the fiscal policies of others in the federation, ultimately resulting in a suboptimal amount of stabilization. Third, stabilization policy may be more expensive to conduct in a decentralization environment if the debt issued by lower levels of government is viewed by capital markets as being riskier than that issued by the national government. Finally, the co-ordination of policy may be easier to achieve when that policy originates from the national government rather than when that policy originates from several decentralized government units. In contrast to the traditional view, it has more recently been argued that decentralized fiscal policy may be optimal. When the regions of a country are highly specialized in production and trade, the impact of relative price movements will fall differently on each of the regions. In this environment, regionally based shocks that have asymmetric effects across the regions will often dominate common nationally based shocks as a source of short-run business cycle fluctuations. The optimal policy response to these types of fluctuations is a regional response, allowing the decentralized policy to dominate centralized policy. Also, when regions are different in economic structure, the spillover and leakage effects among the regions will be smaller. The incentive effects for free-riding in the conduct of stabilization policy at the lower levels of government are then reduced. While it is possible to debate the issue of whether to centralize or decentralize responsibility for stabilization policy from a theoretical perspective, in most federations that responsibility is shared. With shared expenditure responsibilities, shared revenue sources and the ability to issue debt, both higher and lower levels of government typically have the capacity to direct fiscal instruments to the conduct of stabilization policy. Thus, in most federations the issue of policy co-ordination is more important than that of policy assignment. The policy co-ordination problem arises most strongly in federations where there is a high degree of interdependence between the national and subnational levels of government. When there is not a clear delineation and exclusivity of responsibility, the policy initiatives of one government level will spill over onto the budget positions of other government levels. This spillover effect could potentially offset or undo the impact of the original policy initiative. For example, in Argentina, Brazil and Mexico there exists mandated revenue sharing from the national to lower levels of government. Should any of these countries’ national governments decide that it is appropriate to increase tax rates on shared-revenue tax bases in order to dampen aggregate demand for stabilization purposes, lower levels of government will experience a revenue “windfall” through the revenue-sharing process. If the lower levels of government decide to simply spend these additional revenues, the restrictiveness of the national government’s policy initiative will be reduced. Co-ordination of the policy between the national and lower levels of government may be required to ensure that it conveys the desired impact on the economy. References Allsopp, C., G. Davies and D. Vines (1995) “Regional Macroeconomic Policy, Fiscal Federalism, and European Integration,” Oxford Review of Economic Policy, 11: 126-144. Amieva-Huerta, J. (1997) “Mexico,” in Fiscal Federalism in Theory and Practice, T. Ter-Minassian, ed. (Washington, DC: International Monetary Fund), pp. 570-597. Barro, R. (1979) “On the Determination of Public Debt,” Journal of Political Economy, 87: 940-971. Boothe, P. and J. Petchey (1996) “Assigning Responsibility for Regional Stabilization: Evidence From Canada and Australia,” in Reforming Fiscal Federalism for Global Competition, P. Boothe, ed. (Edmonton: University of Alberta Press), pp. 141-162. Case, A., J. Hines and H. Rosen (1993) “Budget Spillovers and Fiscal Policy Interdependence: Evidence from the States,” Journal of Public Economics, 52: 285-307. Dillinger, W. and S. Webb (1999) “Fiscal Management in Federal Democracies: Argentina and Brazil,” Working Paper #2121, World Bank, Washington, DC. Gavin, M. and R. Perotti (1997) “Fiscal Policy in Latin America,” NBER Macroeconomics Annual 1997, pp. 11-61. Paloni, A. (1994) “The External Constraint and Short-Run Macroeconomic Policy: Theory with an Application to Argentina, Brazil and Mexico,” Scottish Journal of Political Economy, 41: 256-277. Schwartz, G. and C. Liuksila (1997) “Argentina,” in Fiscal Federalism in Theory and Practice, T. Ter-Minassian, ed. (Washington, DC: International Monetary Fund), pp. 387-422. Shah, A. (1998) “Fiscal Federalism and Macroeconomic Governance: For Better or Worse?”, Working Paper #2005, World Bank, Washington, DC. Ter-Minassian, T. (1997) “Brazil,” in Fiscal Federalism in Theory and Practice, T. Ter-Minassian, ed. (Washington, DC: International Monetary Fund), pp. 438-456.