¹ Riker, , “Six Books in Search of a Subject or Does Federalism Exist and Does It Matter?” Comparative Politics 2, no. 2 (1969), 324CrossRefGoogle Scholar.

² Oates, Wallace, Fiscal Federalism (New York: Harcourt Brace Jovanovich, 1972)Google Scholar.

³ Tiebout, Charles, “A Pure Theory of Local Government Expenditures,” Journal of Political Economy 64 (October 1956)CrossRefGoogle Scholar; Brennan, Geoffrey and Buchanan, James, The Power to Tax: Analytical Foundations of a Fiscal Constitution (New York: Cambridge, 1980)Google Scholar.

⁴ Weingast, Barry, “The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development,” Journal of Law, Economics, and Organization 11 (April 1995)Google Scholar; Montinola, Gabriella, Qian, Yingyi, and Weingast, Barry, “Federalism, Chinese Style: The Political Basis for Economic Success in China,” World Politics 48 (October 1994)Google Scholar.

⁵ Rodden, Jonathan, “The Dilemma of Fiscal Federalism: Grants and Fiscal Performance around the World,” American Journal of Political Science 46 (July 2002)CrossRefGoogle Scholar; Triesman, Daniel, “Decentralization and Inflation: Commitment, Collective Action, or Continuity?” American Political Science Review 94 (December 2000)Google Scholar; Wibbels, Erik, “Federalism and the Politics of Macroeconomic Policy and Performance,” American Journal of Political Science 44 (October 2000)CrossRefGoogle Scholar.

⁶ Woodruff, David, Money Unmade: Barter and the Fate of Russian Capitalism (Ithaca, N.Y.: Cornell University Press, 1999)Google Scholar.

⁷ Armijo, Leslie Elliott and Jha, Prem Shankar, “Centre-State Relations in India and Brazil: Privatisation of Electricity and Banking,” in Kahkonen, Satu and Lanyi, Anthony, eds., Institutions, Incentives and Economic Reforms in India (New Delhi: Sage Publications, 2000)Google Scholar.

⁸ Weingast (fh. 4); Montinola, Qian, and Weingast (fn. 4).

⁹ For a concise and spirited defense of this perspective, see Buchanan, James, “Federalism as an Ideal Political Order and an Objective for Constitutional Reform,” Publius 25 (Spring 1995)Google Scholar.

¹⁰ On information and agency, the most important contribution is still Oates (fn. 2), in which the efficiencies of decentralization are greatest in the presence of heterogeneous regions and small externalities. For a similar conclusion from a political economy perspective, see Lockwood, Ben, “Distributive Politics and the Costs of Centralization,” Review of Economic Studies 69 (August 2002)CrossRefGoogle Scholar. Tiebout (fn. 3) offers the classic statement on the virtues of mobility and sorting.

¹¹ For examples, see Inman, Robert and Rubinfeld, Daniel, “The Political Economy of Federalism,” in Mueller, Dennis, ed., Perspectives on Public Choice: A Handbook (Cambridge: Cambridge University Press, 1997)Google Scholar; Lyons, W. E., Lowery, David, and Dehoog, Ruth Hoogland, The Politics of Dissatisfaction: Citizens, Services, and Urban Institutions (Armonk: M. E. Sharpe, 1992)Google Scholar; Rodden, Jonathan and Rose-Ackerman, Susan, “Does Federalism Preserve Markets?” Virginia Law Review 83, no. 7 (1997)CrossRefGoogle Scholar.

¹² In addition to the principal-agent, externality, and coordination problems introduced below, critics of decentralization point out that these theories underestimate the importance of economies of scale in centralized provision of public goods and that some policies, most notably redistribution, cannot be efficiently conducted at the local level. Moreover, a key problem with informational arguments in favor of decentralization is that information-constrained voters might be more inclined to monitor the central government than local governments and that shared or overlapping authority might make accountability more difficult.

¹³ Weingast (fn. 4); Montinola, Qian, and Weingast (fn. 4).

¹⁴ Watts, Ronald, Comparing Federal Systems (Kingston, Ontario: Institute of Intergovernmental Relations, 1999)Google Scholar; Newhouse, John, “Europe's Rising Regionalism,” Foreign Affairs 76 (February 1997)CrossRefGoogle Scholar; Doner, Richard and Hershberg, Eric, “Flexible Production and Political Decentralization in the Developing World: Elective Affinities in the Pursuit of Competitiveness?” Studies in Comparative International Development 34 (Spring 1999)CrossRefGoogle Scholar.

¹⁵ Weingast (fn. 4).

¹⁶ Qian, Yingyi and Roland, Gerard, “Federalism and the Soft Budget Constraint,” American Economic Review 88, no. 5 (1999)Google Scholar.

¹⁷ Lohmann, Susanne, “Federalism and Central Bank Independence: The Politics of German Monetary Policy, 1957–92,” World Politics 50 (April 1998)CrossRefGoogle Scholar.

¹⁸ Qian and Roland (fn. 16).

¹⁹ Lohmann (fn. 17).

²⁰ Brennan and Buchanan (fn. 3).

²¹ Prud'homme, Rémy, “The Dangers of Decentralization,” World Bank Research Observer 10 (August 1995)CrossRefGoogle Scholar; Triesman, Daniel, “Political Decentralization and Economic Reform: A Game-Theoretic Analysis,” American Journal of Political Science 43 (April 1999)Google Scholar; Wibbels (fn. 5).

²² William Dillinger and Steven Webb, “Fiscal Management in Federal Democracies: Argentina and Brazil,” Policy Research Working Paper no. 2121 (Washington, D.C.: World Bank, 1999).

²³ In a sample of developing countries, Wibbels (fn. 5) finds higher and more volatile deficits and inflation rates among federations than among unitary systems. Using a larger sample, Treisman (fn. 5) finds that federations do not demonstrate higher inflation rates than unitary systems, but if inflation problems develop, federations are less likely to resolve them.

²⁴ Rodden(fn.5).

²⁵ Francesca Fornasari, Steven B. Webb, and Heng-Fu Zou, “Decentralized Spending and Central Government Deficits: International Evidence,” World Bank Working Paper (Washington D.C.: World Bank, 1998).

²⁶ Triesman (fn. 5).

²⁷ Samuels, David and Snyder, Richard, “The Value of a Vote: Malapportionment in Comparative Perspective,” British Journal of Political Science 31 (October 2001)CrossRefGoogle Scholar; Stepan, Alfred, “Federalism and Democracy: Beyond the U.S. Model,” Journal of Democracy 10 (October 1999)CrossRefGoogle Scholar.

²⁸ An exception is Canada, where the Senate is weak and appointed, and regional bargaining takes place primarily in other forums: within the cabinet, directly between Ottawa and provincial governments, and in the premiers' conferences.

²⁹ Alesina, Alberto and Drazen, Allan, “Why Are Stabilizations Delayed?” in Sturzenegger, Federico and Tommasi, Mariano, eds., The Political Economy of Reform (Cambridge: MIT Press, 1998)Google Scholar.

³⁰ Fornasari, Webb, and Zou (fn. 25); Triesman (fn. 5).

³¹ This literature is too large to review here. Highlights include Roubini, Nouriel and Sachs, Jeffrey, “Political and Economic Determinants of Budget Deficits in the Industrial Democracies,” European Economic Review 33 (May 1989)CrossRefGoogle Scholar; Poterba, James and Hagen, Jürgen von, eds., Fiscal Institutions and Fiscal Performance (Chicago: University of Chicago Press, 1999)CrossRefGoogle Scholar; Torsten Persson and Guido Tabellini, “Political Institutions and Policy Outcomes: What Are the Stylized Facts?” (Manuscript, Department of Economics, Bocconi University, 2001); Robert Franzese, “The Political Economy of Public Debt: An Empirical Examination of the OECD Postwar Experience” (Manuscript, Department of Political Science, University of Michigan, 2001).

³² Relatively low values for the United States and Switzerland may be surprising—this is because local and municipal governments are not included in either the numerator or the denominator. Local data were unavailable for several countries, and furthermore our arguments are focused on constituent units in federations.

³³ Courchene, Thomas, “Preserving and Promoting the Internal Economic Union: Australia and Canada,” in Boothe, Paul, ed., Reforming Fiscal Federalismfor Global Competition: A Canada-Australia Comparison (Edmonton: University of Alberta Press, 1996)Google Scholar.

³⁴ Dillinger and Webb (fn. 22).

³⁵ McCarten, William, “The Challenge of Fiscal Discipline in the Indian States,” in Rodden, Jonathan, Eskeland, Gunnar, and Litvack, Jennie, eds., Decentralization and the Challenge of Hard Budget Constraints (Cambridge: MIT Press, forthcoming)Google Scholar.

³⁶ Bomfim, Antúlo and Shah, Anwar, “Macroeconomic Management and the Division of Powers in Brazil,” World Development 22 (April 1994)CrossRefGoogle Scholar; Jonathan Rodden, “Federalism and Bailouts in Brazil,” in Rodden, Eskeland, and Litvack (fn. 35).

³⁷ Given data shortcomings, we are unable to differentiate between provincial spending for which allocation decisions are made solely by provincial officials and that which is controlled indirectly by federal fiat (unfunded mandates, for instance). H1 makes the plausible assumption that a shift from central to provincial spending implies some loss of direct central control over the consolidated public sector budget.

³⁸ For an overview of concepts and measurements of fiscal illusion and a literature review, see Oates, Wallace, “On the Nature and Measurement of Fiscal Illusion: A Survey,” in Oates, ed., Studies in Fiscal Federalism (Brookfield, Vt: Edward Elgar, 1991)Google Scholar. For a theoretical application to intergovernmental grants in particular, see Oates, “Lump-Sum Intergovernmental Grants Have Price Effects,” in Peter Mieszkowski and William Oakland, eds., Fiscal Federalism and Grants-in-Aid (Washington, D.C.: Urban Institute, 1979).

³⁹ See Weingast, Barry, Shepsle, Kenneth, and Johnsen, Christopher, “The Political Economy of Benefits and Costs: A Neoclassical Approach to Distributive Politics,” Journal of Political Economy 89 (August 1981)CrossRefGoogle Scholar.

⁴⁰ E.g., Winer, Stanley, “Some Evidence on the Effect of the Separation of Spending and Taxing Decisions,” Journal of Political Economy 91 (February 1983)CrossRefGoogle Scholar; Stein, Ernesto, “Fiscal Decentralization and Government Size in Latin America,” in Fukasaku, Kiichiro and Hausmann, Ricardo, eds., Democracy, Decentralization and Deficits in Latin America (Washington, D.C.: Inter-American Development Bank and OECD, 1998)Google Scholar.

⁴¹ Rodden(fn.5).

⁴² Dillinger and Webb (fn. 22) make a plausible argument to the contrary: they suggest that transfer dependence sometimes provides the central government with valuable leverage that can be used to impose reforms and tighter fiscal discipline on the subnational units. However, such strong conditionality is the exception rather than the rule in most intergovernmental transfer schemes in federations, and even if central governments make such proclamations, they may not be credible in the long run.

⁴³ See, e.g., Kiichiro Fukasaku and Luiz de Mello, “Fiscal Decentralization and Macroeconomic Stability: The Experience of Large Developing and Transition Economies,” in Fukasaku and Hausmann (fn. 40).

⁴⁴ For cases without certain kinds of revenue-sharing programs, the GFS data and government data are identical, since the GFS is based on country sources.

⁴⁵ For example, Brennan and Buchanan (fn. 3).

⁴⁶ Riker and Schaps, “Disharmony in Federal Government,” Behavioral Science 2 (1957).

⁴⁷ Dillinger and Webb (fn. 22); Garman, Christopher, Haggard, Stephan, and Willis, Eliza, “Fiscal Decentralization: A Political Theory with Latin American Cases,” World Politics 53 (January 2001)CrossRefGoogle Scholar; Ordeshook, Peter and Shvetsova, Olga, “Federalism and Constitutional Design,” Journal of Democracy 8 (January 1997)CrossRefGoogle Scholar.

⁴⁸ McCubbins, Matthew D., “Party Governance and U.S. Budget Deficits: Divided Government and Fiscal Stalemate,” in Alesina, Alberto and Carliner, Geoffrey, eds., Politics and Economics in the Eighties (Chicago: University of Chicago Press, 1991)Google Scholar; Cox, Gary and McCubbins, Matthew, “Divided Control of Fiscal Policy,” in Cox, Gary and Kernell, Samuel, eds., The Politics of Divided Government (Boulder, Colo.: Westview Press, 1992)Google Scholar; Alt, James and Lowry, Robert, “Divided Government, Fiscal Institutions, and Budget Deficits: Evidence from the States,” American Political Science Review 88 (December 1994)CrossRefGoogle Scholar.

⁴⁹ Jonathan Rodden, “Creating a More Perfect Union: Political Parties and the Reform of Federal Systems” (Manuscript, Department of Political Science, MIT, 2001).

⁵⁰ Jones, Mark, Sanguinetti, Pablo, and Tommasi, Mariano, “Politics, Institutions and Fiscal Performance in a Federal System: An Analysis of the Argentine Provinces,” Journal of Development Economics 61 (April 2000)CrossRefGoogle Scholar.

⁵¹ An interesting contrary hypothesis in the Indian context is presented in Stuti Khemani, who argues that since the deficits of the Indian states are funded primarily by loans over which the central government has discretion, state deficits are essentially pork manipulated by the central government. As a result, deficits are higher in the states controlled by the center, though it is unclear whether this would have any effect on overall public sector deficits. See Khemani, “Partisan Politics and Subnational Fiscal Deficits in India: What Does It Imply for the National Budget Constraint?” (Manuscript, World Bank, 2001).

⁵² Coalition governments at the center complicate the collection of this data for Switzerland, Brazil, and Austria. In fact, we are unable to calculate a sensible measure for Switzerland, where the federal executive is a collegial body that represents (by convention) all of the major parties. In Brazil, where the party system is highly fractionalized, national executives must rely on unstable legislative coalitions. It is plausible that members of such coalitions would be able to discipline their copartisans at the state level in a manner consistent with the theoretical propositions outlined above. Nevertheless, the variable presented in Figure 1 (and used in subsequent regressions) counts only those states run by the same party as the chief executive, for the simple reason that where coalition governments are prevalent, chief executives have had little success at disciplining states governed by other coalition members. To deal with the concern, we have also constructed a variable that codes states controlled byjunior members of the federal coalition as controlled by the center. This variable is different for a small number of years only in Brazil and Austria and does not affect the results reported below. In the case of subnational coalition governments (prevalent in Austria, Germany, and India), we code based on the senior member of the coalition that occupies the office of chief minister, prime minister, president, and so on.

⁵³ Note that for country-years characterized by authoritarianism we have coded this variable as 1, indicating that the central government controlled all state governments.

⁵⁴ David Wildasin, “Externalities and Bailouts: Hard and Soft Budget Constraints in Intergovernmental Fiscal Relations” (Manuscript, Washington, D.C., World Bank, 1997).

⁵⁵ See May, R. J., Federalism and Fiscal Adjustment (Oxford: Oxford University Press, 1969)Google Scholar. With reference to defense, a similar argument was made by Alexander Hamilton about the Dutch Confederation: “In this confederacy, one large province, by its superior wealth and influence, is commonly a match for all the rest; and when they to not comply, the province of Holland is obliged to compel them”; cited in Frisch, Morton J., ed., Selected Writings and Speeches of Alexander Hamilton (Washington, D.C.: American Enterprise Institute, 1985), 200Google Scholar.

⁵⁶ GDP share would perhaps be preferable to expenditure share as a measure of ajurisdiction's ability to impose negative fiscal externalities on others, but we are unable to obtain provincial-level GDP data for the full sample. Given our interest in fiscal policy, expenditure concentration is preferable to a measure of population concentration.

⁵⁷ Wildasin (fn. 54).

⁵⁸ Elazar, Daniel J., “From Statism to Federalism: A Paradigm Shift,” Publius 25 (Spring 1995)Google Scholar; Watts (fn. 14);Treisman (fn. 5).

⁵⁹ Suberu, Rotimi T., Federalism and Ethnic Conflict in Nigeria (Washington, D.C.: United States Institute of Peace Press, 2001)Google Scholar.

⁶⁰ It is important to avoid double counting grants that are included in the center's expenditures and the provinces' revenues. In the numerator the two cancel out when calculating the combined central-provincial surplus. However, to accurately measure the denominator—total expenditures—it is necessary to subtract grants to avoid double-counting the grants when they are “spent” at the central level and then again at the local level.

⁶¹ Treisman (fn. 5); Wibbels (fn. 5).

⁶² We use the log, as the inflation data are skewed.

⁶³ These data are from the World Bank, World Development Indicators (information available at www.worldbank.org).

⁶⁴ Alesina, Alberto and Roubini, Nouriel, Political Cycles and the Macroeconomy (Cambridge: MIT Press, 1997)Google Scholar.

⁶⁵ This variable is coded 1 only for elections for the national-level parliament or chief executive. Data are taken from World Bank, Database of PoliticalInstitutions (www.worldbank.org).

⁶⁶ Haggard, Stephan and Kaufman, Robert R., The Political Economy of Democratic Transitions (Princeton: Princeton University Press, 1995)Google Scholar.

⁶⁷ Data taken from World Bank (fn. 65). As described below, we have experimented with several ather measures of horizontal political fragmentation as well.

⁶⁸ Buchanan, James and Wagner, Richard, Democracy in Deficit: The Political Legacy of Lord Keynes (New York: Academic Press, 1977)Google Scholar.

⁶⁹ Data taken from the Polity 98 data set (www.cidcm.umd.ed/inser/polity/indev.htm).

⁷⁰ We have also experimented with a range of additional demographic control variables: area, population, population/number of states, urbanization, population density, ethnic fractionalization, and percentage of the population above and below the working age. None of these attained statistical significance, and none affected the substance or significance of the results reported herein.

⁷¹ Data taken from World Bank (fn. 63).

⁷² Data taken from World Bank (fn. 63). We have also estimated models that address these possibilities by differentiating between expected GDP and shocks, but this estimation technique does not affect the results presented below.

⁷³ Data taken from World Bank (fn. 63).

⁷⁴ For recent overviews of tradeoffs, see Maddala, G. S., “Recent Developments in Dynamic Econometric Modelling: A Personal Viewpoint,” in Mebane, Walter, ed., Political Analysis, vol. 7 (Ann Arbor: University of Michigan Press, 1998)Google Scholar; Judson, Ruth and Owen, Ann, “Estimating Dynamic Panel Data Models: A Guide for Macroeconomists,” Economic Letters 65 (October 1999)CrossRefGoogle Scholar; Baltagi, Badi, Econometric Analysis of Panel Data (New York: Wiley, 2001)Google Scholar; Bernhard Kittel and Hannes Winner, “How Reliable Is Pooled Analysis in Political Economy? The Globalization-Welfare State Nexus Revisited,” Discussion Paper 2–3 (Cologne: Max-Plank-Institut für Gesellschaftsforschung, May 2002).

⁷⁵ Beck, Nathaniel and Katz, Jonathan, “What to Do (and Not to Do) with Time Series Cross-ection Data,” American Political Science Review 89 (September 1995)CrossRefGoogle Scholar.

⁷⁶ The bias also increases with the magnitude of the autoregressive coefficient. See Judson and Kven (fn. 74).

⁷⁷ See Robert Franzese, Cindy Kam, and Amaney Jamal, “Modeling and Interpreting Interactive Hypotheses in Regression Analysis” (Manuscript, University of Michigan, 1999).

⁷⁸ In models withoutfixedeffects, the interaction term and its components are jointly significant at the 1 percent level in thefiscalperformance equation (model 5) but do not reach significance in the inflation equation (model 6).

⁷⁹ The arguments presented above suggest not only that deficits and inflation might be higher in the absence of vertical copartisanship but also that central and provincial governments might attempt to shift theirfiscalburdens onto one another in a “vertical war of attrition” instead of taking painful adjustment measures. To examine this possibility, we have also estimated a dynamic model in which the copartisanship variable is interacted with the lagged dependent variable in order to test whether high levels of copartisanship are associated with faster adjustment to large deficits. The results suggest that, indeed, high levels of copartisanship are associated with faster adjustment.

⁸⁰ See Garrett, Geoffrey and Rodden, Jonathan, “Globalization and Decentralization,” in Kahler, Miles and Lake, David, eds., Globalizing Authority (Princeton: Princeton University Press, forthcoming)Google Scholar.

⁸¹ It is difficult to know how to interpret the effect of fractionalization in the central legislature on combined central-provincial budget balance. We have estimated models using other measures of political fractionalization suggested by Roubini and Sachs (fn. 31) and by Tsebelis, George, “Decision-Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism and Multjpartism,” British Journal of Political Science 25 (July 1995)CrossRefGoogle Scholar. We also estimated dynamic models that interact these measures with the lagged dependent variable to capture delayed adjustment, but none of these variables attains significance or affects the main results.

⁸² Specifically, we used Im-Pesaran-Shin and Levin-Lin tests. See Maddala, G. S. and Kim, In-Moo, Unit Roots, Cointegration, and Structural Change (Cambridge: Cambridge University Press, 1998)Google Scholar; Baltagi (fn. 74).

⁸³ For a defense of the error correction specification for this type of data, see Beck, Nathaniel, “Comparing Dynamic Specifications: The Case of Presidential Approval,” in Stimson, James, ed., PoliticalAnalysis, vol. 3 (Ann Arbor: University of Michigan Press, 1991)Google Scholar.

⁸⁴ Arellano, Manuel and Bond, Stephen, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application in Employment Equations,” Review of Economic Studies 58 (April 1991)CrossRefGoogle Scholar.

⁸⁵ Note that the results reported above do not include Switzerland because of our inability to calculate the copartisanship variable. When Switzerland is included and the copartisanship variable is dropped, the substance and the significance of all other variables are unchanged.

⁸⁶ All of the results are available from the authors upon request.

⁸⁷ Jones, Mark, Sanguinetti, Pablo, and Tommasi, Mariano, “Politics, Institutions and Fiscal Performance in a Federal System: An Analysis of the Argentine Provinces,” Journal of Development Economics 61 (April 2000)CrossRefGoogle Scholar.

⁸⁸ Khemani (fn. 51).